IBM PPC => No Altivec?

[arn]

The following was published to a public Apple mailing list:

I attended a briefing today on IBM's high performance computing technology, which is hinged on their Power4 CPU (this CPU has awesome performance in the various real-world benchmarks I have seen). After the briefing, I asked the presenter (a chief engineering manager from IBM) about the Power4 derivative for desktops and low end servers to be announced in October. You may recall that there has been speculation that this CPU would find its way into PowerMacs in the future. Well, it sounds like this CPU is not in Apple's future -- the "over 160" vector instructions are not AltiVec (even though AltiVec has 162 instructions), and there are technical issues that would prevent AltiVec from ever marrying with Power4 or its successors. Furthermore, the guy came right out and said that they have pitched the desktop Power4 to Apple, but Apple was not interested.

So, although Power4-based PowerMacs seemed like a promising (and likely) possibility, it looks like it won't happen. I guess we will have to wait and see what Apple has in store for the future . . . .

[reyesmac]

The powermac machines have been falling behind more and more not because Intel is making faster chips quicker, it is because Apple is not keeping up with them at a fast enough rate. Who is Apple making their machines compete with? The iMac, with an even slower chip, at least can impress windows users. The iMacs designs are always better than the PC's that it competes with. But the powermac is a different story, it in the last few years has not been designed to compete head to head with PC's. You see that if you make a PC part for part with an iMac they will both cost almost the same. But the Powermac keeps getting more expensive for what you get. I truly wonder what Apple has up their sleeve if anything to combat this. Will they just become a consumer company? If a truly fast Mac will come along, how long will they have planned it? It seems to me they just started worrying about speed when theirs machines couldnt go past 500mhz. If there was one thing I do not like about Apple it is that they do not have any competition on its own platform. Any other Mac clone company would have made faster machines than Apples by now. They would just look bad, but at least they would be cheap and the OS still pretty. Even if Apple comes out with a fast miracle chip, I can bet you it will only be the $3000 dollar one that it will be in. They always make you pay thru the nose for the machine that can actually compete with the PC in terms of speed. I have been waiting for the G5 ever since the G4 came out. I guess by the time that it does, I will be able to actually afford it.....cool!:rolleyes:

[BongHits]

Originally posted by arn

So, although Power4-based PowerMacs seemed like a promising (and likely) possibility, it looks like it won't happen. I guess we will have to wait and see what Apple has in store for the future . . . .
[/i]

well if it continues to be motorola's BS upgrades....then i won't be in apple's future for a long time...i think my 933 will be competent enough to last me until Apple get's their **** together.

[dongmin]

Yeah the last few reports on this have tended to refute any notions of a future IBM-Apple marriage. I'm sad. It seemed so promising when the Power4-derivative rumors first came out.

[TechLarry]

Marklar.

PLEASE!

Motorola is a lost cause, and if this news is correct then Apple has no choice.

It's time to move on...

TL

[DeepDish]

Not knowing what Apple has up it's sleeve, I am in not in a position to speculate.

Nonetheless, Apple, please don't make some stupid decision here. Apple, like most companies have made stratigical errors throughout it's histroy. However, Apple seems to make more than others. Apple creates great software and hardware, but they do make some major short sighted decissions also in terms of long term planning.

If Apple thinks it can stick with MOTO, then this is another major example of a idiot decision!

[Uragon]

guess this is good news for upgrade companies (powerlogix and sonnet).... and MAKLAR is becoming more of a reality. Why do they always make it difficult for Apple (Moto/IBM).. assuming that Apple wasn't interested with IBM due to "not to compatible with Altivec....

[pilotgi]

The following was published to a public Apple mailing list:

So does that mean this info is reliable?

Anyway, as far as Apple's future plans for faster processors, _anything_ is possible.

Maybe they're waiting for Moto's 9nm chips they claim they will be producing "6 months before the competition."

Maybe they'll go back to the G3 with the rumored 200Mhz bus and speeds up to 2Ghz.

Maybe AMD, maybe Sun.

Maybe_anything_

Steve sure loves to keep his secrets.

[wchamlet]

I think you people are getting a little ahead of yourselves. Stand back and look at Apple as a whole company. How much money have they made since the first iMac? A few billion in profit? From the info I have seen on the net Apple has something like 4 billion dollars cash. Not tied up in paperware and other investments like Microsoft and other companies have done. In my opinion I think they are doing and amazing job acquiring profits, especially for being such a small company.

Now with that in mind, think what is Apple trying to do? Maximize profits? Yup. Create something new to aquire more profits? Yup. Fall further and further behind the current crop of PC's? Nope.

Apple is probably one of the few forward thinking companies in the world. For my needs, the DP 1 GHZ nonDDR G4 is more than enough for me. I could really care less if there was a 1.7 Ghz G4 available next year. And I really doubt that is where Apple is headed into the future. Remember the G4 when it came out. Outstanding chip, wasn't it? Did anyone predict it coming? Not that I heard.

Apple will provide a next generation chip, but I really don't think it will be based on this genarations technology. I wouldn't be surprsised though if Apple focuses on more specific technology in the future, like embedded graphics on the motherboard, or something that will really stand out from PC's. Because, like someone said earlier, the Powermacs are now competing with PC's, and losing. I think that Apple will change that, but I cannot guess how.

[Timothy]

That SJ doesn't care about competing in the real world; and, as long as we Mac devotees don't leave, then I guess he's right.

We are in a tough spot, those of us who've chosen the Mac platform; we have no leverage in terms of competition. We've invested so much into the platform in terms of software, peripherals, and training that we hesitate to switch.

But, for how much longer? It is clear that with the money and resources that the Intel/AMD group can throw at processors, we will continue to slide further behind. I think the only option for Apple will be to make the switch to Intel/AMD processors in their boxes.

I am becoming increasingly annoyed at the silence from Cupertino regarding this very real, and growing...growing...growing problem. I've been waiting for some hint as to a possible solution for about 4 years, and nothing has come; in fact, SJ refuses to discuss the issue directly.

I remain discouraged...

[chubakka]

This report isn't even talking about the next IBM PPC chip... it's talking about the Power 4... Apple isn't interested in the Power 4. This "report" is a dead end. Skip it.

[gandalf55]

i am hanging my head low here... Apple has been moving to sway Hollywood types to the Mac platform via software acquisitions, etc. That activity requires Pro level machines. Fast ones. Ones that scare the neighbors. Fans? I could care less... throw 3 in there if the processors scream. apple should be looking to who can provide the best in an acceptable time frame. period. price? well, for most Pro users, its not as big an issue when you have a company signing the check (and getting a return on the investment by more productive workers.)

Things have been trickling for a while now. Looking for someone to turn the faucet a little bit. We're thirsty :)

[djkut]

Originally posted by arn
The following was published to a public Apple mailing list:

I After the briefing, I asked the presenter... about the Power4 derivative for desktops and low end servers to be announced in October...the "over 160" vector instructions are not AltiVec (even though AltiVec has 162 instructions), and there are technical issues that would prevent AltiVec from ever marrying with Power4 or its successors. Furthermore, the guy came right out and said that they have pitched the desktop Power4 to Apple, but Apple was not interested.
So, although Power4-based PowerMacs seemed like a promising (and likely) possibility, it looks like it won't happen. I guess we will have to wait and see what Apple has in store for the future . . . .


chubakka:

I think it's pretty clear that this person wasn't talking about the actual Power4.


Once again, though, why are we believing this person so easily, why is this so reliable?

[Timothy]

With the 1.25 ghz chips not even really entering the picture until November, really, it is probable that there will not be an "update" for another 6 months at the latest, putting us in the range of MWNY 2003. If Apple gives us another 20% speed upgrade then (announced, not shipping until Nov. 2003) that would mean that we won't even cross the 1.5 ghz barrier for another year.

This is getting bleak; and Apple continues to provide NO ROADMAP for any alternate hope. Why aren't the major shareholders raising more crap about this?

This latest update of the Powermac was a joke; in reality, it was no update. The announced high-end chips still don't exist, but in theory only. The case is perhaps the ugliest case since some of the pizza box performa cases. If we imagine that Motorola is completely taxed just to make 1.25 ghz chips by November, it is highly unlikely that there is anything else waiting in the wings. It makes you wonder if they even look forward with any seriousness, or if they (Apple/Motorola) just wait until the night before an "update" and decide, then, to start working on a faster chip? I was under the naive impression that real work and possible solutions were being worked on now? Apparently not.

Did I mention I am feeling a bit down on Apple today? ;-)

[Kid Red]

Just because this one guy said Apple wasn't interested means 3 things-

1) He's covering for Apple
2) He doesn't know jack
3) He's telling the truth

If he is telling the truth then thaty could mean 3 things as well-

1) This chip isn't the one IBM is making for Apple tho they are still making one
2) Apple wasn't interested because the G4 still has plenty of legs left
3) X on Intel is a reality (it is, but I mean that it's the future)

So don't hang your heads to low.

[Thirteenva]

Originally posted by reyesmac
The powermac machines have been falling behind more and more not because Intel is making faster chips quicker, it is because Apple is not keeping up with them at a fast enough rate. Who is Apple making their machines compete with?

Motorola is not "keeping up" not apple, apple is innovative and finds creative solutions to problems. Motorola has dropped the ball for R&D for desktop chips. Apple's marketing strategy has never been to sell the fastest chip, it's been to sell the best package. There selling a whole unit, that as a whole has more than you would get with other systems. Isn't this why we all love apples so much? Why else would we be here when we could jump on the wintel 2ghz bandwagon... Obviously speed is not the only factor in selling computers.



The iMac, with an even slower chip, at least can impress windows users. The iMacs designs are always better than the PC's that it competes with.

Case in point, the imac does well despite its "even slower chip" because apple is selling a hardware/software package that appeals to people.



But the powermac is a different story, it in the last few years has not been designed to compete head to head with PC's. You see that if you make a PC part for part with an iMac they will both cost almost the same. But the Powermac keeps getting more expensive for what you get.

How can you justify that the powermac is "getting more expensive for what you get". The lower end imac price is virtually the same yet they just added a second processor. Seems to me your getting more for you money than 3 months ago. From what i hear the low end PMac is selling very well because people find it to be a great deal for that kind of performance.



I truly wonder what Apple has up their sleeve if anything to combat this.

I also wonder whats in store for apple. They'll have to move to a new chip eventually how far can they stretch the g4, I doubt it has life above 1.6ghz.


It seems to me they just started worrying about speed when theirs machines couldnt go past 500mhz.


I don't recall this problem.


If there was one thing I do not like about Apple it is that they do not have any competition on its own platform.


Thats the beauty of apple, they provide both the hardware and software solution so it's integrated seamlessly.


Any other Mac clone company would have made faster machines than Apples by now.


And how would they do this having access to the same chips apple has access to???


Even if Apple comes out with a fast miracle chip, I can bet you it will only be the $3000 dollar one that it will be in. They always make you pay thru the nose for the machine that can actually compete with the PC in terms of speed.


I doubt if there was a new chip it would only be in the top power mac. It would probably appear in all level power macs in single and dual versions at different power levels.


I have been waiting for the G5 ever since the G4 came out. I guess by the time that it does, I will be able to actually afford it.....cool!:rolleyes:

If every time something new comes out you hope for something better you'll always be dissappointed. When the g5 comes out you'll want a G6. Use whats available or seek other solutions to your problem.

I hear dell is having a sale...

[GetSome681]

Originally posted by Uragon
guess this is good news for upgrade companies (powerlogix and sonnet).... and MAKLAR is becoming more of a reality. Why do they always make it difficult for Apple (Moto/IBM).. assuming that Apple wasn't interested with IBM due to "not to compatible with Altivec....

This makes no sense to me. Let's say that this desktop Power4 doesn't have Altivec capabilities. Ok. So we decide to switch to some x86 chip. Those won't have the Altivec unit either. So why would Apple ever turn down this IBM chip and go to Intel or such? Doesn't make sense. They lose Altivec going either way, yet going Intel means a HUGE pain in the ARSE in terms of programs, and switching everything over.

Even without Altivec, it would seem silly to me for Apple to switch to Intel, etc., rather than take this chip.

[bertinman]

Originally posted by BongHits
i think my 933 will be competent enough to last me until Apple get's their **** together.

Sorry, but doesn't that concede the fact that the motorola chips are good enough and can stand their own? If you are able to do what you need on your "933" why do you even need a "G5?"

This is just like the fast car situation. Why do you need a fast car when every car made now can go at least 65 mph (or 100kph)? The faster cars are a luxury and that is it. Same goes for faster computers. Sure people wouldn't mind having more power but really how much is the gain from the difference?

Maybe I'm slow minded but I kind of like slight pauses so I can think ahead of the computer (meaning milli-secounds, because that is all I have been given in the last few years).

My two cents are that we all should start worrying about other things anyway. Why not think about something creative/cool/unique that you can do with your mac and do it instead of wishing that there are faster macs. There will always be faster macs.

(I really enjoyed the switch ad contest btw--sad that I didnt get to finish mine in time)

(btw BongHits, this is not an attack, just thought it was a good beginning to my rant for the day :) no hard feelings)

-- bert :cool:

[LethalWolfe]

Originally posted by bertinman


Sorry, but doesn't that concede the fact that the motorola chips are good enough and can stand their own? If you are able to do what you need on your "933" why do you even need a "G5?"

This is just like the fast car situation. Why do you need a fast car when every car made now can go at least 65 mph (or 100kph)? The faster cars are a luxury and that is it. Same goes for faster computers. Sure people wouldn't mind having more power but really how much is the gain from the difference?

Maybe I'm slow minded but I kind of like slight pauses so I can think ahead of the computer (meaning milli-secounds, because that is all I have been given in the last few years).

My two cents are that we all should start worrying about other things anyway. Why not think about something creative/cool/unique that you can do with your mac and do it instead of wishing that there are faster macs. There will always be faster macs.

(I really enjoyed the switch ad contest btw--sad that I didnt get to finish mine in time)

(btw BongHits, this is not an attack, just thought it was a good beginning to my rant for the day :) no hard feelings)

-- bert :cool:

In terms of the consumer market I agree. But in the pro market faster, more powerful computers will always be needed as the tasks computers are given to do will alwasy become more intensive (such as FX heavy films/commercials/TV shows and CG movies).


Lethal

[LethalWolfe]

Originally posted by gandalf55
i am hanging my head low here... Apple has been moving to sway Hollywood types to the Mac platform via software acquisitions, etc. That activity requires Pro level machines. Fast ones. Ones that scare the neighbors. Fans? I could care less... throw 3 in there if the processors scream. apple should be looking to who can provide the best in an acceptable time frame. period. price? well, for most Pro users, its not as big an issue when you have a company signing the check (and getting a return on the investment by more productive workers.)

Things have been trickling for a while now. Looking for someone to turn the faucet a little bit. We're thirsty :)


Hollywood has been using Mac hardware (running Avid software)for years and years and years. Apple is now trying to take over post production on the software side of things too (and they are making a good run at it). I've seen post houses "make the switch" but that was from Avid to FCP. I've never seen a post house edit using a PC. Apple hardware has always been a staple in post production (and will be for a long time) and they are now trying to make their software equally as popular.


Lethal

[bertinman]

Originally posted by Kid Red

3) X on Intel is a reality (it is, but I mean that it's the future)

Umm, why don't people see the "marklar" thing as a joke... as it should be.

I am assuming that by "X" you mean OSX? well, that has NEVER been on an intel machine. why you ask? Altivec is not on a i386 machine.

I think people are confused because there has been something called "X" for quite some time in the UNIX/Linux world. It is an windows server (not MS windows just windows as in a command/graphical/what-not windows), KDE, Gnome and whatever else window managers you like are loaded on top and make quite useful/pretty desktop for Linux/UNIX gurus. Check out www.XFree86.com and you'll see it's not a brand new program. Also, if you want to use it on your Mac, go ahead its available through Apple free of charge (JOIN ADC FREE!).

So, in closing--I have Marklar/Darwin-x86 right here next to me running "X."
It is a Athlon 1gig that is happily saying "Welcome to Macintosh!" "login:" -- WOW.

Oh wait I lied, one more thing. Join any mailing list for darwin and search the archives ther will be at least one message thread about making Aqua (everyone’s favorite operating system which runs with Darwin at it's core) run on an i386 (intel compatible) machine--all saying that it is close to impossible and/or not worth it.

-- bert :cool:

[Faeylyn]

Groan..........

Why is ANYONE taking this report seriously? Is this any different at all from "some guy told a friend of mine that his buddy's roomate's sister had an aunt that said Apple wasn't going to use the new chip..."?

Whether it's true or not, IT DOESN'T MATTER. NO ONE CARES whether Apple uses the new IBM chip. NO ONE CARES whether Apple develops quad or 8x machines. You reading this DO NOT CARE. What you care about is whether Apple's next generation of Macs is significantly faster than the current generation. You care about whether Apple's machines can keep up with (and surpass) the PCs out there. You care about this even if you personally don't need the power because it ensures Apple's survival and thus the Macs survival.

If Apple were to come out with a Motorola 64-bit G5+ in January that beat the tar out of the best PC, you'd be happy. If they came out with Macs with IBM's new chip that beat the tar out of PCs, you'd be happy.

So, the question then is, what is the likelyhood that Apple will come out with a new Mac that will be significantly faster than the current Macs? For your answer, look to the companies Apple has been purchasing lately. Unless Apple's top brass have suddenly become brain-dead, they realize that the current crop of Macs is NO WHERE NEAR up to the task to run these apps and be taken seriously by the market they're going after. And they already have a solution for this problem. They wouldn't spend the millions on those companies otherwise.

There will be new kiss-arse Macs within a year. Do you really care what's inside, as long as they leave the PCs in the dust?

[bertinman]

Originally posted by LethalWolfe


In terms of the consumer market I agree. But in the pro market faster, more powerful computers will always be needed as the tasks computers are given to do will alwasy become more intensive (such as FX heavy films/commercials/TV shows and CG movies).


Lethal

Good point, lost that one.

I believe the Xserve is the way of the future. Everywhere I look people are buying clusters (I have a brand new 80 node Linux one next to me) and doing amazing things that no single computer can do. The pro towers can be fast enough to talk to the servers and run their code there are well. Multi-processing is fun and worth it.

True, the Xserve is not the fastest server, but it IS a baby. already with the large scale amount of time/money put into them the bugs are getting uncovered and erased. Also, the Xserve has the ability to become huge powerhouses if just given the right friends (fast XRaid servers and fast ethernet and FSB (which no doubt will be updated soon)).

I am sure there are plenty of fellow terminal users who would agree that alot can be had from a dinky crap computer that just hooks up to a "mainframe" or "room or networked kick-ass servers." Now if you had a decent computer you could also do most of the work right there, but then send the large number crunching tasks to the server.

Anyway this is getting too long for a guy who is suppose to be crunching numbers himself:D.

-- bert :cool:

[Thirteenva]

Originally posted by LethalWolfe


In terms of the consumer market I agree. But in the pro market faster, more powerful computers will always be needed as the tasks computers are given to do will alwasy become more intensive (such as FX heavy films/commercials/TV shows and CG movies).


Lethal


Apple is catering to these markets with faster machines. Alot of the apps used in these industries are multi processor aware and will run quite well on a dual 1.25 ghz without even a hiccup. Also alot of these industries rely on more than just ghz ratings. Faster hard drives, good graphics cards and expandability are all important factors in computer choice. Power macs have all of this available to them.

Whats makes us loyal apple users is the fact that we realize there is more to a computer than ghz ratings. Isn't this why we by apples over wintel machines. If all we cared about was speed we could just get an AMD in a custom built box....

Why do people forget this as soon as the subject of processor speed comes up...

[Kelesis]

Originally posted by bertinman
I am assuming that by "X" you mean OSX? well, that has NEVER been on an intel machine. why you ask? Altivec is not on a i386 machine.
I just want to point out that this is not a valid argument. OS X runs on the G3 processor, which does not have Altivec.

[DaveGee]

Originally posted by bertinman
Umm, why don't people see the "marklar" thing as a joke... as it should be.

Bzzzzt... sorry but thanks for playing. :D

Do you know where X came from? (X as in 'OS X') - Hint: NeXT/OpenStep
Do you know what CPU that code base was written for? Hint: Not PPC
Do you know why it took so long for X to debut? Hint: New CPU Needed to be supported aka PPC

If you knew the roots of X you wouldn't think 'Marklar' was such a joke. In fact if Apple didn't take time to keep X working on x86 I'd be really shocked since as of 18 months ago X was running on x86.

AltiVec is far from required... iBooks and (older) iMacs are all the proof you need on that.

Dave

[Postal]

What die-size is Motorola's G4 currently being manufactured on? Isn't it 0.18-micron? If so, then moving to 0.13 or 0.09 would let them do a LOT with the clock speed and/or adding features. It'd be nice if we could return to the "high-end-becomes-the-low-end" scheme again. 1.25 GHz low-end chips with 1.6 GHz for the high end? That might happen...

[TyleRomeo]

Originally posted by LethalWolfe



Hollywood has been using Mac hardware (running Avid software)for years and years and years. Apple is now trying to take over post production on the software side of things too (and they are making a good run at it). I've seen post houses "make the switch" but that was from Avid to FCP. I've never seen a post house edit using a PC. Apple hardware has always been a staple in post production (and will be for a long time) and they are now trying to make their software equally as popular.


Lethal

lethal is right on, everyone thats a prosumer and in any form of creative industry (film/graphics/photo/animation/ ect..) works on a mac. If you use a PC you either get laughed at or weird stares. So anyone here that says apple is a consumer company can keep on saying that. Let consumers that want to use a simple OS for using word, IE, and light digital work with apples digital hub iProducts. It doesnt matter how fast PCs are, they are unstabble, you can never edit successfully on a PC. Actually a lot of film editors use 533MHZ G4s to edit along with AVIDs hardware becuase they are very stable. So MHZ or GHZ only matter to those apple people who want the fast GHZ numbers just for the plain fact of having them.

So if want the GHZ go to the dark side if you want ease and comfort go with a MAC. MACs are like lexury cars rather then muscle cars. I'm actually happy about that.

Kamil

[synergy]

Originally posted by TyleRomeo




So if want the GHZ go to the dark side if you want ease and comfort go with a MAC. MACs are like lexury cars rather then muscle cars. I'm actually happy about that.

Kamil

Sure Macs are like Luxury cars. Then there are car makers like BMW who makes the M3 as a sport luxury car. Which is what I want. Both from BMW and from Apple. Fast and comfortable, thats not asking too much.

[alex_ant]

Let's assume this news is true:

After the briefing, I asked the presenter (a chief engineering manager from IBM) about the Power4 derivative for desktops and low end servers to be announced in October. You may recall that there has been speculation that this CPU would find its way into PowerMacs in the future. Well, it sounds like this CPU is not in Apple's future -- the "over 160" vector instructions are not AltiVec (even though AltiVec has 162 instructions), and there are technical issues that would prevent AltiVec from ever marrying with Power4 or its successors.
AltiVec is a Motorola trademark that IBM would not be able to use even if it designed its own fully-AltiVec-compatible vector processor. The fact that the VPU on this Power4 derivative is not called AltiVec means nothing. For all anyone knows, it could be a fully AltiVec-compatible vector unit called PoopyVec. Note that Apple never uses the name "AltiVec" either - it uses the term "Velocity Engine." The "technical issues" mentioned, I'm guessing, are a misinterpretation of what the IBM guy said. This chip could well have a VPU that is not "AltiVec," but that would still run vectorized code designed for AltiVec. I can't imagine two incompatible vector units on two different chips in the same processor family.
Furthermore, the guy came right out and said that they have pitched the desktop Power4 to Apple, but Apple was not interested.
I'll turn conspiracy theory mode on for a second and say that if Apple was interested in this chip, it wouldn't be surprising to hear a middle manager at IBM say otherwise. Apple exhibits complete mastery at keeping secrets, and what's the best way to keep secrets? To tell them only to those who need to know them. Does this IBM guy need to know? If not, then he probably doesn't, or has been misinformed by his superiors in order to prevent speculation. I realize this sounds very elaborate and unlikely, but this is how Apple works. I'm sure IBM is quite used to it.

In my opinion, these Power4 derivatives still look like likely candidates for future Macs - if not the next revision of Power Macs, then the revision after that.

Alex

[bertinman]

Originally posted by DaveGee


Bzzzzt... sorry but thanks for playing. :D

Do you know where X came from? (X as in 'OS X') - Hint: NeXT/OpenStep
Do you know what CPU that code base was written for? Hint: Not PPC
Do you know why it took so long for X to debut? Hint: New CPU Needed to be supported aka PPC

If you knew the roots of X you wouldn't think 'Marklar' was such a joke. In fact if Apple didn't take time to keep X working on x86 I'd be really shocked since as of 18 months ago X was running on x86.

AltiVec is far from required... iBooks and (older) iMacs are all the proof you need on that.

Dave

OUCH. I'll keep my mouth shut on that one.:o

But, NeXT was on the mac type cpu... 68k, there is a NeXt Cube down the hall (yes, Im blessed with almost every type of computer). This stupidity on my part comes from a hectic day... grouped all macs as one for a moment there.

I still feel that intel based osx is not the best idea none-the-less, because it would ruin Apple's control of hardware which is why Apple's systems are so great--They work together.

*puts on flame retardent suit* ok ready for the flames that will undoubtily come soon...

(btw... I do know the roots... my "idiot gear" just kicked in today)

-- bert :cool:

[chubakka]

Apple has us so in the dark we need a flashlight!

They've gotta have something in the works...
and none of us have a clue as to what it is.

I think it's pretty amusing. I guess I'm an optimist.

[Mr. Anderson]

I'm just going to take a wait and see approach to all this. We still have the IBM announcement in October - who knows what they might say.

I just hope what ever happens with the next gen chip for Apple is that it can go head to head with anything else out there.

D

[bertinman]

Originally posted by chubakka
Apple has us so in the dark we need a flashlight!

They've gotta have something in the works...
and none of us have a clue as to what it is.

I think it's pretty amusing. I guess I'm an optimist.

Exactly, I think people should be more happy than mad about all the secrets.

Which would you rather have, lots of secrets or not? If we didn't we would all be bored... just like the Wintel fanatics (they are rare, but when I do run into them they're usually not too happy in general), or we could be... mad because we don't know whats happening. Or, amused at the bored people, mad people, and even the secret holders because we've got a great soup opera to watch :D. I like the secrets.

wow, I had a pocket of pennies today... and a few buttons too(eh, DaveGee and Kelesis? ;)).

ding! 2 cents!


-- bert :cool:

[DakotaGuy]

What would happen if you could put the Alti-Vec technology on the 750fx processor from IBM. Am I right or wrong in saying that the 750fx (current G3 in the iBook) is actually, less Alti-Vec, a far superior CPU compared to the G4? Doesn't IBM claim soon they will have this processor running at 2GHz. Doesn't it have less pipeline stages then the G4 from Motorola, and am I right or wrong in the fact that it is built at a .13 compared to .18 on the G4.

I would think a 750fx with Alti-Vec and true DDR would out perform any current G4 on the market even at the same clock speed. Is this a possibility?

[AmigaMac]

Apple better not even think 'Intel Inside' or I'm outta here! AMD might be tolerable!

[Cappy]

I saw it on the internet so it must be true. :rolleyes:

[eirik]

I think the report is very credible. In another forum, a poster familiar with the individual that reported his encounter with the IBM dude says that the he has a good track record.

The IBM guy would be insane to lie. He is faced with two options that would not cost him dearly: 1) say nothing or 2) say the truth, knowing that your superiors would back you up.

So, assume for a moment that Apple truly is not interested in Junior, would Apple do so if it did not have a better option? Not likely.

Someone made a good point about a non-compatible implementation of VMX by IBM. Most likely, a few tweaks and a simple re-compile would do the trick. Once one has gone to the trouble of programming with vector/matrix operations, the hard part is done.

This leaves the other possibility, there's an equivalent or better CPU out there. I say equivalent because time to market is an important factor too.

So, maybe IBM has another CPU. I doubt it. Most likely, I'd say Motorola has soemthing else to offer. So, is it a Book E chips w/AltiVec and maybe 64 bit...? Or, is it something like the speculated 7470?

I think the latter most likely, though the former I'd prefer. Even so, a 7470, as speculated, would be great. We'd have native DDR!!! If they'd add one more feature to it I'd be thrilled. How about an onboard RIO component? Yeah baby, yeah! (Oh yeah, MERSI too!)

For memory intensive apps, all those Gigaflops of CPU capability are untapped when there's a bottleneck in front of the CPU. A single 7455 at 1 Gig easily saturates the bus on big jobs. With a good implementation of RIO and MERSI, Apple could just keep adding CPU's.

I don't have great interest in a chip that is faster than what we have today where small jobs are concerned. But for big jobs that are memory intensive, two or four CPU's, with the right architecture, would KICK ASS.

So, I think the 74XX still has a lot of life to it. Would I like something with twice the superscalar capability and a true 64 bit arch? Sure.

The superscalar enhancement would help ALL applications but greatly increase the relative size and power draw. The 64 bit arch, well, I doubt I could afford to buy that much DDR RAM anyway. I don't see myself with more than 2 GB anytime soon. Besides, there are other ways to address more memory.

Eirik

PS I wouldn't hold your breath about one of Apple's CPU's being produced in the first 3 to 6 months on the 90 nm facilities. More likely, we'll see static RAM and high-end cell phone chips first. Plus, Motorola's announcement says they'll BEGIN to ramp up in 4Q02. That too indicates low volume for a while. We're stuck with the bottleneck for quite some time.

[Thresher]

We can sit here an argue what type of CPU will be in the next generation Macs all day, but it's pointless.

I can't imagine that the top leadership is not aware of the speed problem. Even if Steve's RDF is on overload, Apple is a publicly traded company, if the stock starts taking a nose-dive because the market share decreases even further as a result of the speed issue, something will be done.

Personally, I think the new Powermacs are a stopgap until some really kickass hardware comes out. My bet is on the IBM chips. My reasoning is that since IBM didn't have the new Power4 chip ready, Apple had to do something and soon. Motorola doesn't even seem to have anything in the pipeline as a replacement for the current G4, so it would seem they would be out for the next revision. I suspect we'll see a new Powermac sooner than most people believe. I suspect it will be sometime in January. It's a hunch, but there you go.

Still, I'd love to see OS X on a proprietary version of the X86-64 platform. It will probably never happen, but it would make Thresher a very happy boy. You get the speed of the AMD platform and Apple doesn't have to sell it's soul to intel.

A couple of notes:

Altivec is not the end all, be all. Most apps don't even use it. AMD has licensed SSE and SSE2 for their next platform. Although they are not as comprehensive, they would do a lot of the things that Apple needs.

BSD and Mach are portable and can be recompiled for any platform to take advantage of the hardware.

Lastly, a die shrink down to .13 or .09 doesn't necessarily translate into faster speeds. It does mean less power consumption and heat. It CAN mean faster speeds, but not automatically, if the reduced heat is enough to let them increase the clock speed.

[Choppaface]

well i hope apple gets its act together somehow....jobs said that he envisioned catching up with intel in late 2003, so he's got a year to do it....

[Catfish_Man]

...if Apple isn't using this IBM chip because it doesn't have Altivec, then they definitely aren't using an x86 chip. Sure x86 has SSE/MMX/etc... but this IBM chip also has a vector unit of some sort (IBM said so). Therefore, if Apple ditched this chip because it doesn't have Altivec, they are planning on using one that does have Altivec. That means Motorola. Before people start bashing Motorola some more, I'd like to point something out: To the best of my knowledge, IBM has never made a faster desktop chip than Motorola. Motorola isn't doing so great now, but they do have a better track record than IBM.

Note: I still hope this guy's wrong. That IBM chip sounds awesome.

[e-coli]

Originally posted by TyleRomeo


lethal is right on, everyone thats a prosumer and in any form of creative industry (film/graphics/photo/animation/ ect..) works on a mac.

This is most certainly untrue. Avid boxes that run uncompressed video only run on NT systems. they're crazy stable, and you're hard-pressed to crash them. This is the system most huge post houses are using. If they're using anything else, it's a flint / flame / inferno on an sgi station. But none of the major post houses are running stations with compression.

[alex_ant]

Originally posted by Catfish_Man
Before people start bashing Motorola some more, I'd like to point something out: To the best of my knowledge, IBM has never made a faster desktop chip than Motorola.
Hmm... that's not saying a whole lot, is it. :)

[TyleRomeo]

Originally posted by e-coli


This is most certainly untrue. Avid boxes that run uncompressed video only run on NT systems. they're crazy stable, and you're hard-pressed to crash them. This is the system most huge post houses are using. If they're using anything else, it's a flint / flame / inferno on an sgi station. But none of the major post houses are running stations with compression.

no silly, I worked for a post house this summer, where MTV did rock and jock 2002 and the blue crush surfer special and not to mention comedy centrals Project Search light. AND THEY ALL USED MAC BASES AVID UNITY EDITING MACHINES. Ohh wait we did get two rented IBM and Compaq Avids and they froze and had so many bugs in with them. The clients said they would always stick with MACs and forget PC editing machines.

With AVID the mac just keeps the other hardware and software happy. The $10,000 Genie PCI cards do most of the work.

AND for home prosumer editing, people use FCP, which only runs on a mac. Don't tell me that Premeire is better than FCP on a PC.

Basically when it comes to editing the work is done on a Mac and not on a PC. End of story.

Tyler

[dongmin]

Originally posted by Thresher

Altivec is not the end all, be all. Most apps don't even use it. AMD has licensed SSE and SSE2 for their next platform. Although they are not as comprehensive, they would do a lot of the things that Apple needs.


You may not think it's that great but Steve Jobs apparent does. Remember the quote: "The roadmap on the PowerPC actually looks pretty good and there are some advantages to it. As an example, the PowerPC has something in it called AltiVec, we call the Velocity Engine -- it's a vector engine -- it dramatically accelerates media, much better than, as an example, the Intel processors or the AMD processors... so we actually eke out a fair amount of performance from these things when all is said and done"

Apple is committed to Altivec. It may not be universally exploited by developers, but Apple definitely makes use of it in OS X. It's about the only thing that keeps the G4 remotely on the radar.

Of course, it's possible that IBM's vector unit is comparable or better in which case it makes sense for Apple to consider switching. But then, what about the development costs? For now, Altivec is a major part of the Apple cpu strategy.

[lenz]

Originally posted by Catfish_Man
...if Apple isn't using this IBM chip because it doesn't have Altivec, then they definitely aren't using an x86 chip. Sure x86 has SSE/MMX/etc... but this IBM chip also has a vector unit of some sort (IBM said so). Therefore, if Apple ditched this chip because it doesn't have Altivec, they are planning on using one that does have Altivec. That means Motorola. Before people start bashing Motorola some more, I'd like to point something out: To the best of my knowledge, IBM has never made a faster desktop chip than Motorola. Motorola isn't doing so great now, but they do have a better track record than IBM

Ummm....No?

What would you say about G3s? When both apple and moto were making them who was producing the higher clocking chips? I hope Apple does use IBM's chip because we all know the amount of power big blue has. Moto is barely staying on it's feet and doesn't really bring in too much profit from selling desktop chips. IBM has always been ahead of the game.

[Cappy]

Originally posted by dongmin


You may not think it's that great but Steve Jobs apparent does. Remember the quote: "The roadmap on the PowerPC actually looks pretty good and there are some advantages to it. As an example, the PowerPC has something in it called AltiVec, we call the Velocity Engine -- it's a vector engine -- it dramatically accelerates media, much better than, as an example, the Intel processors or the AMD processors... so we actually eke out a fair amount of performance from these things when all is said and done"

Apple is committed to Altivec. It may not be universally exploited by developers, but Apple definitely makes use of it in OS X. It's about the only thing that keeps the G4 remotely on the radar.

Of course, it's possible that IBM's vector unit is comparable or better in which case it makes sense for Apple to consider switching. But then, what about the development costs? For now, Altivec is a major part of the Apple cpu strategy.

Jobs' statement means nothing. It was he who declared the death of crts and then a few months later the eMac was introduced and has become a more prominent item in their product line. In addition what else is he to say that wouldn't have stockholders and potential customers totally altering their plans.

People have a hard time accepting this but if Apple announced x86 support tomorrow, their stock would likely take a nasty beating. I actually want to see x86 support for the most part but I don't feel Apple is in any position to pull that off at the moment. Maybe 2-3 years. They've got to keep MS onboard with Office long enough to create a version that would run on x86 as well as PPC. That alone could sell copies of Mac OS X and Office X to businesses out the whazoo. They don't have to throw out their hardware investment then.

[e-coli]

Originally posted by TyleRomeo


no silly, I worked for a post house this summer, where MTV did rock and jock 2002 and the blue crush surfer special and not to mention comedy centrals Project Search light. AND THEY ALL USED MAC BASES AVID UNITY EDITING MACHINES. Ohh wait we did get two rented IBM and Compaq Avids and they froze and had so many bugs in with them. The clients said they would always stick with MACs and forget PC editing machines.

With AVID the mac just keeps the other hardware and software happy. The $10,000 Genie PCI cards do most of the work.

AND for home prosumer editing, people use FCP, which only runs on a mac. Don't tell me that Premeire is better than FCP on a PC.

Basically when it comes to editing the work is done on a Mac and not on a PC. End of story.

Tyler


Funny. I've been doing the Audi spots for years, and they're all done on NT based Avids.

When we bought our uncompressed stations, a decent Mac version wasn't available yet.

Different strokes for different houses, I guess. But you can't make the assumption that ALL high-end editing and finishing is done on Macs, because it isn't. That was the point of my original post. You missed it.

BTW, Unity isn't an editing system. It's a media management tool. :rolleyes:

...and just FYI, MTV is famous for not using uncompressed high-end editing. As a matter of fact, they frequently use the DV or DVC codec. Hardly high-end. ;)

[sentinal]

lethal is right on, everyone thats a prosumer and in any form of creative
industry (film/graphics/photo/animation/ ect..) works on a mac.
If you use a PC you either get laughed at or weird stares.


this is not the case. Film and animation businesses are mac-friendly, but in any seat where performance is important, you'll find PC's running win2k or linux.


So anyone here that says apple is a consumer company can keep on saying
that. Let consumers that want to use a simple OS for using word, IE, and
light digital work with apples digital hub iProducts. It doesnt matter how
fast PCs are, they are unstabble, you can never edit successfully on a PC.
Actually a lot of film editors use 533MHZ G4s to edit along with AVIDs
hardware becuase they are very stable. So MHZ or GHZ only matter to
those apple people who want the fast GHZ numbers just for the plain fact
of having them.

film editing is not the same thing as film and tv visual effects. animators and compositors and render farms need every ounce of performance available. the stability and smoothness of osx is worth a lot, but without raw performance its not enough

[DaveGee]

Originally posted by bertinman
I still feel that intel based osx is not the best idea none-the-less, because it would ruin Apple's control of hardware which is why Apple's systems are so great--They work together.

*puts on flame retardent suit* ok ready for the flames that will undoubtily come soon...

(btw... I do know the roots... my "idiot gear" just kicked in today)

-- bert :cool:

Bert,

Didn't mean to sound harsh sometimes my typing gets a sound of it's own... And as far as X running on x86 yea I'm sure it's true but I also agree with you... I'm not too sure it's a good idea either. :cool:

D

[ffakr]

I'm not saying this report wasn't poorly worded, but if we look at what was actually said, it really does sound like they were talking about Power4s, not PowerPCs.

Fact 1: IBM has announced a 64 bit PowerPC with SIMD TBA at the Microprocessor forum. IBM said PowerPC not Power4... they said the design of this PowerPC is influenced by the Power4 processor, but they didn't say it was a Power4 (they are different, though very similar architectures). Power4s can run PPC code, but PPCs do not have all the instructions of Power4s.

Fact 2: gcc has publicly released changes to their code base to accomodate a PowerPC64 with Altivec support. This isn't a rumor, it's in GCC now. Do a CVS diff to see it.

These are not rumor. They are facts. Facts from the GCC project, facts from IBM themselves.

Now look at what is actually in this report...

I asked the presenter (a chief engineering manager from IBM) about the Power4 derivative for desktops and low end servers
OK, either poorly written, or he asked and got an answer about a Power4 derivative (a new Power4). IBM has told the Microprocessor Report that this is a PowerPC with design cues from the Power4. In my book, A sky scraper with a gradual sloped roof and big flower pots is definately taking cues from Frank Lloyd Wright, but it sure isn't a derivatve of a prarie house. This is, to me, a big difference.

prevent AltiVec from ever marrying with Power4 or its successors
Um, once again, IBM has stated this is not a Power4, it is a PowerPC. Either this is sloppy prose, or the IBM engineer was talking about a different processor. He of all people should know the difference between a PowerPC and a Power4.

they have pitched the desktop Power4 to Apple, but Apple was not interested
Once again, IBM has said the upcomming announcement is not a Power4. The chip we are really talking about is a PowerPC.

Basically, whether or not this person intended to ask about a PowerPC or a Power4, it appears that the IBM engineer responded with answers about's Apple's interest in the Power4.

[Thresher]

You've parsed that statement pretty carefully, but I can't disagree with you at all.

I've found in my brief status as a Mac fan that you HAVE to parse statements regarding Apple VERY carefully. If you don't ask the question precisely, you don't get an accurate answer.

[Kid Red]

Originally posted by bertinman


Umm, why don't people see the "marklar" thing as a joke... as it should be.

I am assuming that by "X" you mean OSX? well, that has NEVER been on an intel machine. why you ask? Altivec is not on a i386 machine.

I think people are confused because there has been something called "X" for quite some time in the UNIX/Linux world. It is an windows server (not MS windows just windows as in a command/graphical/what-not windows), KDE, Gnome and whatever else window managers you like are loaded on top and make quite useful/pretty desktop for Linux/UNIX gurus. Check out www.XFree86.com and you'll see it's not a brand new program. Also, if you want to use it on your Mac, go ahead its available through Apple free of charge (JOIN ADC FREE!).

So, in closing--I have Marklar/Darwin-x86 right here next to me running "X."
It is a Athlon 1gig that is happily saying "Welcome to Macintosh!" "login:" -- WOW.

Oh wait I lied, one more thing. Join any mailing list for darwin and search the archives ther will be at least one message thread about making Aqua (everyone’s favorite operating system which runs with Darwin at it's core) run on an i386 (intel compatible) machine--all saying that it is close to impossible and/or not worth it.

-- bert :cool:

uh, wrong. Apple has had X on Intel for quite sometime now. There is an article at MacNN that discusses this very issue. Also, Apple may be ready to switch to an AMD processor by this time next year if IBM or Moto don't follow thru.

There is no confusion with "X" running on Intel.

[alex_ant]

Originally posted by Cappy
People have a hard time accepting this but if Apple announced x86 support tomorrow, their stock would likely take a nasty beating. I actually want to see x86 support for the most part but I don't feel Apple is in any position to pull that off at the moment. Maybe 2-3 years.
My opinion is just the opposite: I think if Apple were to announce x86 support tomorrow, their stock would go through the roof. Analysts are not the smartest people in the world when it comes to high-tech. They hear the word "Intel" and they think of things like "market leader," "unstoppable juggernaut," "nobody ever got fired for buying Intel," etc., and it makes them all cozy and happy. They understand terms like "core competency" and "value added" and so on, but they do not necessarily understand such concepts as why more gigahertz does not necessarily mean better performance. Therefore, when they hear of Apple announcing that OS X will run on x86, they will think, "Great! Finally Apple's operating system will be married to a CPU with lots of gigahertz! Outlook: Good."

However, that said, I would not like to see Apple have anything to do with x86, ever, and I do realize "ever" is a strong word...
They've got to keep MS onboard with Office long enough to create a version that would run on x86 as well as PPC. That alone could sell copies of Mac OS X and Office X to businesses out the whazoo. They don't have to throw out their hardware investment then.
The x86 architecture is fast enough now that, AltiVec aside, it could probably run such apps as Office.X without a crippling speed penalty, if Apple wanted to put for the necessary effort writing a suitable PPC emulation layer for an x86 OS X.

I also don't think the Office.X-x86-Mac OS X combination would be incredibly appealing to businesses. Sure, they can run OS X on the machines they already have, but why should they?

1) They've already paid for Windows
2) The conception still exists (mostly rightfully so) that Windows is what software that businesses use runs on
3) "Nobody ever got fired for buying Wintel"
4) Lots of people have gotten fired (or at least laughed at) for even mentioning anything having to do with Macs
5) Windows is what people know; OS X would require massive retraining of personnel
6) OS X is unproven and its future is uncertain, whereas everybody knows with absolute certainty that Microsoft and Windows will be around for a long long long time
7) Everybody else uses Windows, so why shouldn't we
8) Windows is faster
9) Win2k is more professional looking; OS X's interface looks like a toy to business managers

All these points have been constantly reinforced for the past decade plus, and they aren't going to go away anytime soon. Basically, I do think OS X on x86 could suit plenty of businesses, but the same thing was said about OS/2, and I don't envision OS X making a big dent in the business world anytime in the near future, x86 or not.

Alex

[alex_ant]

Originally posted by ffakr
Now look at what is actually in this report...
I asked the presenter (a chief engineering manager from IBM) about the Power4 derivative for desktops and low end servers
OK, either poorly written, or he asked and got an answer about a Power4 derivative (a new Power4). IBM has told the Microprocessor Report that this is a PowerPC with design cues from the Power4. In my book, A sky scraper with a gradual sloped roof and big flower pots is definately taking cues from Frank Lloyd Wright, but it sure isn't a derivatve of a prarie house. This is, to me, a big difference.
Well, "derivative" = "derived from." You could say that that skyscraper is a derivative of a prarie house if it actually was derived from a prairie house, even in the smallest way, and even if nothing about it resembles a prairie house. This new desktop PowerPC could be called a Power4 derivative if it was derived even in the slightest from the Power4. It could also be called a Power4 successor if it did in fact come along after the Power4 (although to say that would be kind of misleading).

I don't think this is sloppy prose - it just demands careful reading. This part:
Furthermore, the guy came right out and said that they have pitched the desktop Power4 to Apple, but Apple was not interested.
Makes it very clear which CPU both the author and the marketing guy are referring to. IBM would have never pitched the actual Power4 to Apple.

Once again, IBM has said the upcomming announcement is not a Power4. The chip we are really talking about is a PowerPC.

Basically, whether or not this person intended to ask about a PowerPC or a Power4, it appears that the IBM engineer responded with answers about's Apple's interest in the Power4.
I agree with you here. Something has to be not right about this report. Perhaps the IBM engineering manager misinterpreted the questions, or relayed misinformation he was given, or perhaps he simply talked out of his arse. I was just picking nits. :)

Alex

[reyesmac]

I don't see Apple going x86 if G4's can handle all the products that will come out of all the 3D companies they bought. That being said, I think Apple would do more to please its customers if it would just offer the Macs it sells with only the high end stuff. If they cared about giving us the most speed, why don't they offer all their macs with the best video card and hard drives at $1.7k-2.5k-and 3.3k? Why do we have to settle for 16x speed burners? If we are paying so much, why not give us even a single processor 867mhz mac with Gforce Ti and a 100 or 133 raid and a 40x burner for $1700? I sure would not complaine about the speed if I got all that! But not only do we have to buy Macs that don't run OS X as snappy as PC's run Windows (don't say Quartz Extreme because a PC user could actually upgrade its hardware to use new technologys for the price of a single Mac video card, plus, I shouldnt have to buy a new system just to have a speedy finder, I should have the option of using a less alpha channel using GUI). The most important thing I need is for the windows to move around as fast or faster than they move on my BW G3. That machine what bought at a time when Apple was selling their hardware for 1.5k up to 3k.

I think if Apple does not get more in line with were the hardware industry is going it will be left behind. The industry is making faster computers at lower prices, thats why the price of high end PC's gets lower every few years. Remember, since they make the whole widget, if they are going to be the best platform, they have to compete on the software side and the hardware side. If they don't beat the competition or at least be as good for the same price, then they are not the best, and I refuse to accept that the Mac is not the best for what it does. Heck, if they had G4's that were clocked at 2.4 ghz and prices as cheap as PC's, would there even be any doubt as to who was the best computer maker in this galaxy? There must be something very bad going on in Cupertino for them not to come out with faster machines by now. They know if they impress us Mac users, we would sell our cars to afford their products. What they are offering us is not the best that is possible, something is stopping them? Are they waiting for some sort of contract with Motorola to end? I don't know but if the worst is not behind us then there will be some dark days ahead indeed.

If Apple does not learn that it cannot offer its computers at the price it currently does, it will learn it the hard way, and with our marketshare, that would be disasterous.

[bertinman]

Originally posted by Kid Red


uh, wrong. Apple has had X on Intel for quite sometime now. There is an article at MacNN that discusses this very issue. Also, Apple may be ready to switch to an AMD processor by this time next year if IBM or Moto don't follow thru.

There is no confusion with "X" running on Intel.

Please scroll up about 5 hours... thanks :)

Plus:
All is cool DaveGee :cool:

-- bert :cool:

[bretm]

Originally posted by LethalWolfe



Hollywood has been using Mac hardware (running Avid software)for years and years and years. Apple is now trying to take over post production on the software side of things too (and they are making a good run at it). I've seen post houses "make the switch" but that was from Avid to FCP. I've never seen a post house edit using a PC. Apple hardware has always been a staple in post production (and will be for a long time) and they are now trying to make their software equally as popular.


Lethal

The post production market made the switch to NT 4 years ago. Avid quit development for a time on their mac systems. Today, Avid DS and Symphony still don't operate on macs (those are the high end ones that hollywood uses) and their low end avid xpress DV wasn't released on mac until version 3.5 just a few months ago. The composer line just recently came in sync with the nt versions.

4 years ago the G3s came out without enough slots for Avids so Avid almost ditched 'em completely. I think Apple wanted them to myself so all the loyal mac avid users would switch to FCP (which was waiting to be announced). I think Avid figured this out, backtracked, and changed their mind. They didn't want to look like the bad guy. But the mac loyalty is almost completely dead in the avid world.

The Avid user base is really small anyway. I think Apple could care less. Although they should. There's only 50-100 thousand avids out there. That doesn't include all the other support computers at the post houses and all the computers the editors buy for their homes.

[edenwaith]

Here is a line from the book "Building Cocoa Applications" on page 279:

"Very little of Mach and Mac OS X are written in the PowerPC's native assembly language. As a result, Mac OS X can easily be ported to other microprocessors simply by recompiling the system and rewriting a small number of device drivers."

If this holds true, then, yes, Mac OS X can appear on an Intel-based PC. Also, I've heard that there have been early builds of Darwin on PCs. Even if Mac OS X is ported successfully over to PCs, here are some reasons Apple may or may not want to do it.

In the early 1990's, they had their Star Trek project which ported OS 7 to a PC, and it was successful earlier on. One downside at the time was OS 7 ran faster than the new PowerPC chips that Apple was trying to transition over to at the time. Very bad sign for their hardware company if the OS ran faster on PC hardware. Granted, I don't know of any results which have seen or tested OS X or Darwin on a PC vs. a Mac. What if the same even occurred and the PC hardware ran faster? Once again, this wouldn't be good news to Apple's hardware department. Unless they drastically switch their goals and become more of a software company than a hardware company, porting OS X to a PC would not be good financially.

Also, there would be the problem of successfully porting their OS to PCs, with the multitude of hardware out there. Look at Linux. One challenge is to get as much hardware, new and old, to run with the OS. If Apple decided to license OS X only to particular computer manufacturers who would build systems that were guaranteed to work with OS X, then that might be their start into breaching the PC world. But would that be enough, or should they then spend more development time in building drivers (or catering to 3rd party hardware manufacturers) to create drivers for OS X.

I suppose, what Apple is ideally going for is more PCs users to drop their Dells and their Windows, and buy Mac hardware and Mac software, bringing money into Apple from two sides. But as we've seen in the case of Microsoft, huge profits can be garnered from software sales since the physical costs of software are quite low, especially in comparison to hardware. A few manuals and CDs probably cost a few bucks at most, and considering the high cost of MS's software, the company from Redmond is making a nice little profit with each sale.

[chickenlittle]

...the sky is falling, the sky is falling! Some guy from somewhere said some other guy said that he talked to a guy that said Apple wasn't gonna use the Power4 derivative!

:rolleyes:

This must mean Apple just doesn't care about performance. They are just looking to milk us suckers out of more cash. I mean, look how little progress they made with OS X 10.0 to 10.2 in 18 months. BASTARDS!!!!!

GMAFB. Apple is actively seeking a resolution to the performance issue. It will be resolved in due time. This mamby pamby chicken little BS is getting really fscking tired.

[TyleRomeo]

Originally posted by e-coli



Funny. I've been doing the Audi spots for years, and they're all done on NT based Avids.

When we bought our uncompressed stations, a decent Mac version wasn't available yet.

Different strokes for different houses, I guess. But you can't make the assumption that ALL high-end editing and finishing is done on Macs, because it isn't. That was the point of my original post. You missed it.

BTW, Unity isn't an editing system. It's a media management tool. :rolleyes:

...and just FYI, MTV is famous for not using uncompressed high-end editing. As a matter of fact, they frequently use the DV or DVC codec. Hardly high-end. ;)

Different strokes for different folks. I have never come across a place that chooses to uise PCs over Macs is what I'm saying. OHH AND THE MTV GUYS USE DIGIBETA, is that good for you?

tyler

[ibjoshua]

Originally posted by edenwaith
Here is a line from the book "Building Cocoa Applications" on page 279:

"Very little of Mach and Mac OS X are written in the PowerPC's native assembly language. As a result, Mac OS X can easily be ported to other microprocessors simply by recompiling the system and rewriting a small number of device drivers."
that is very interesting but it doesn't address the issue of all the carbon code out there.

If this holds true, then, yes, Mac OS X can appear on an Intel-based PC. Also, I've heard that there have been early builds of Darwin on PCs. Even if Mac OS X is ported successfully over to PCs, ...
okay let's get this one straight. if you believe this eWeek (http://www.eweek.com/article2/0,3959,496270,00.asp) article then it has not only been done but they (Apple) have the latest cut of 10.2 running on x86 processors. In fact the OSX developers frequently get asked to make changes to keep the OS compatible with x86.

... Also, there would be the problem of successfully porting their OS to PCs, with the multitude of hardware out there. Look at Linux. One challenge is to get as much hardware, new and old, to run with the OS. If Apple decided to license OS X only to particular computer manufacturers who would build systems that were guaranteed to work with OS X, then that might be their start into breaching the PC world. But would that be enough, or should they then spend more development time in building drivers (or catering to 3rd party hardware manufacturers) to create drivers for OS X.
i don't believe any 'reliable' sources said anything about porting OS X to PCs just using x86 chips in macs.


I suppose, what Apple is ideally going for is more PCs users to drop their Dells and their Windows, and buy Mac hardware and Mac software, bringing money into Apple from two sides.
bingo!. think you might have hit on something there. :)

But as we've seen in the case of Microsoft, huge profits can be garnered from software sales since the physical costs of software are quite low, especially in comparison to hardware. A few manuals and CDs probably cost a few bucks at most, and considering the high cost of MS's software, the company from Redmond is making a nice little profit with each sale.
valid point but i don't thank Apple would ever get out of hardware design. who in their right mind would go head on with MS??

[ibjoshua]

post script:

i don't think Apple will ever go down the clone road again. it just didn't work. who'd want to get undercut by a bunch of inferior products on the hardware side of the business only to then have the software guys have to support all the 'non-standard' configurations?

[Kethoticus]

I said this Power4 stuff was BS, and if this guy's testimony is valid, I was right.

Maybe this was referring to the Power4 and not a modified PPC based on the Power4. I don't know. Maybe Apple declined the offer because they do have something up their sleeve that's unbelievable. Of course I have nothing authoritative to offer on that front.

But what I can say is this: every few months I hear some new, incredible rumor. It promises something near-unimaginable on the horizon, something that every Mac fanatic raves will bury the Wintel duopoly once and for all.

How long have I been hearing this stuff? For years. The original Power PC chip was supposed to accomplish this overthrow 9 years ago. Then OS X was. Then the G4 chip, with its altivec instruction set, was. There have been other false promises in between, if memory serves. They've all consistently led to the following:

a) Disappointment on the parts of some Mac fans, who usually threaten to leave the Mac platform once and for all and / or preach its imminent doom;

b) Mac apologists who do just the opposite and make excuses for the disappointment, then latch on to the next too-good-to-be-true false promise / rumor and begin waving the Mac flag in their drooling, Apple's-going-to-conquer-the-world-someday patriotism once again.

I feel like I'm the only one who actually perceives this cycle and is sick of it. It's a joke. People, save yourselves the emotional energy and mental "blue balls" and wait to see what actually transpires. And do not base purchasing decisions on rumors, particularly those that seem too good to be true. I bought a Mac in 1999 because I heard that OS X was coming out in Jan 2000. How long did I have to wait for it to be finally released?!? Just a case in point.

Folks, if you need super speed, don't hope for the Mac to come out with a 2GHz G5 next January. It ain't gonna happen. Get a fast PC because that is what is available now. Trash all hopes you have in vaporware. And if you don't need speed, then the Mac is a wonderful box that can do many consumer-oriented things quite elegantly (PC World's recent home DV editing articles attesting to that).

Okay, I stand ready to weather your flames.

[ibjoshua]

I feel like I'm the only one who actually perceives this cycle and is sick of it. It's a joke. People, save yourselves the emotional energy and mental "blue balls" and wait to see what actually transpires. And do not base purchasing decisions on rumors, particularly those that seem too good to be true. I bought a Mac in 1999 because I heard that OS X was coming out in Jan 2000. How long did I have to wait for it to be finally released?!? Just a case in point.

no you're not the only one and you're not the only one who feels it's necessary to go into long diatribes about everyone else on these forums either.

the site you are currently reading is called macrumors. that's what it's all about fella, rumors.

[iwantanewmac]

I know what they have in store for the future.......
Incredibly meager updates 2 times a year and they will hit 2 GHZ in 2004.
Sounds like an apple strategy to me :)

[pc_convert?]

It all sounds like BS to me.

Quoting from MacRumors in July (http://www.macrumors.com/pages/2002/07/20020719012400.shtml)
Steve Jobs: "The roadmap on the PowerPC actually looks pretty good and there are some advantages to it. As an example, the PowerPC has something in it called AltiVec, we call the Velocity Engine -- it's a vector engine -- it dramatically accelerates media, much better than, as an example, the Intel processors or the AMD processors... so we actually eke out a fair amount of performance from these things when all is said and done. And the roadmap looks pretty good. Now, as you point out, once our transition to Mac OS 10 is complete, which I expect will be around the end of this year or sometime early next year and we get the top 20% of our installed base running 10, and I think the next 20 will come very rapidly after that. Then we'll have options, then we'll have options and we like to have options. But right now, between Motorola and IBM, the roadmap looks pretty decent.Personally I put more trust in what Steve says rather than a public posting by Joe Shmo...

Steve like options...so they dismiss a Power4 derivative...sounds really plausible.

Catch that bit at the end?
But right now, between Motorola and IBM, the roadmap looks pretty decent.I think that sums it up.

[g21]

Thanks Bretm, you're absolutely right - though I'd say mac loyalty still exists among editors, they just haven't had much of a choice of late!

I work in ads and videos in London. Here you find a lot of Avid Symphony on-lines running uncompressed video on NTs, and a lot of off-line Avids running on unbelievably ancient macs. Avids running on G4s are extremely rare, and trying to find an Avid that will talk to a firewire drive is hell on earth. FCP is starting to attract some interest, but while it's very cute, and MUCH cheaper, it's not yet quite as refined and productive to use as Avid Media Composer - but that's another story.

If Apple wants to compete in the professional post market, what it really needs to do is run Shake - or even After Effects - on something that's as least as fast as Smoke, Flame and Inferno, which run on SGIs and are the industry standard for a lot of this work. I don't have the comparative benchmarks (does anyone know the answer?), but ideally I'd like to see a mac running 2-3 times faster to do this kind of work. Realistically though, you have to note that these Flame systems are hugely more expensive. In practice, running AE on a new dual 1ghz is a surprisingly pleasant experience. Apple should do well in the editing market with current hardware, and if they could come up with something a couple of times faster at this kind of price level, they will sell A LOT of hardware in the professional compositing and effects market. Even some kind of external render farm or PCI card would do the trick.

[QuiteSure]

Someone from IBM says that Apple has declined to use the Power4 chip for its products. Possible interpretations:

1. The statement is correct and Apple has another plan. This is good news. Let's hope Apple reveals itself soon.

2. The statement is correct and Apple is run by idiots. I don't have any reason to believe that this is true. It doesn't appear to me from recent history that Apple is run by idiots.

3. The statement is incorrect and Apple is actually going to use the Power4 chip. This is good news, Apple simply employing more misdirection.

4. The statement is incorrect and the discussion with Apple never existed. This is neutral. I can go back to sleep.

L' shana tova!

[dekator]

What horse will Apple bet on ?
Apple has done a lot of things right lately... and yes, it's not about MHz. But Apple *needs* more powerful machines. Their loyal users (and others) know it isn't just about MHz but still, more power is needed.
It's ironic: In order to keep up in terms of performance Apple jumped on the AltiVec train. It has become dependent on it (optimizing code for it) so that Apple apps can keep up in performance with Wintel ones. Now, already the AltiVec train is derailing... Apple, I know it's a pain to re-optimize and retreat from the AltiVec hype... but you should jump off that train *now* before it takes you with it down into the abyss.
Apple, don't shoot yourself in the foot again... one (hardware) is already getting limp...

[luiss]

Many comments seem to imply that since IBM's 'PowerPC' vector unit is not compatible it would be easier for Apple to just go to x86.
Mac OS X currently runs on G3s, with no velocity engine. If Apple were to use a PowerPC processor from IBM, it would simply use the non-Altivec code path for executing code, until Apple and 3rd parties recompile. Seems like a much better migration story than needing new binary versions of everything for running on x86.

[DaveGee]

Originally posted by QuiteSure
Someone from IBM says that Apple has declined to use the Power4 chip for its products. Possible interpretations:

1. The statement is correct and Apple has another plan. This is good news. Let's hope Apple reveals itself soon.

2. The statement is correct and Apple is run by idiots. I don't have any reason to believe that this is true. It doesn't appear to me from recent history that Apple is run by idiots.

3. The statement is incorrect and Apple is actually going to use the Power4 chip. This is good news, Apple simply employing more misdirection.

4. The statement is incorrect and the discussion with Apple never existed. This is neutral. I can go back to sleep.

L' shana tova!


QuiteSure,

There is atleast one more and one that wouldn't make the comments from the IBM rep a total lie. One that I think is pretty close to the truth...

Someone from IBM says that Apple has declined to use the Power4 chip for its products. Possible interpretations:

5. The statement WAS true at a given point in time....

Just something to think about...

Dave

[Cappy]

I just find it fascinating that people are so hung up on this that they're parsing what someone stated they heard from someone else who stated what they heard from someone else who stated what someone else heard. How much farther might that go beyond what's being posted?

How in the freaking world can you "factually" derive anything from that?!?!?!

C'mon people...parsing posts and breaking them down can look cool and impressive but just because it's posted on the internet doesn't mean it's true.

[Cappy]

Originally posted by Kid Red


uh, wrong. Apple has had X on Intel for quite sometime now. There is an article at MacNN that discusses this very issue. Also, Apple may be ready to switch to an AMD processor by this time next year if IBM or Moto don't follow thru.

There is no confusion with "X" running on Intel.

Again...just because it's posted on the internet does not mean it's true.

[GigaWire]

wow...

[Cappy]

Originally posted by alex_ant

I also don't think the Office.X-x86-Mac OS X combination would be incredibly appealing to businesses. Sure, they can run OS X on the machines they already have, but why should they?

1) They've already paid for Windows
2) The conception still exists (mostly rightfully so) that Windows is what software that businesses use runs on
3) "Nobody ever got fired for buying Wintel"
4) Lots of people have gotten fired (or at least laughed at) for even mentioning anything having to do with Macs
5) Windows is what people know; OS X would require massive retraining of personnel
6) OS X is unproven and its future is uncertain, whereas everybody knows with absolute certainty that Microsoft and Windows will be around for a long long long time
7) Everybody else uses Windows, so why shouldn't we
8) Windows is faster
9) Win2k is more professional looking; OS X's interface looks like a toy to business managers

All these points have been constantly reinforced for the past decade plus, and they aren't going to go away anytime soon. Basically, I do think OS X on x86 could suit plenty of businesses, but the same thing was said about OS/2, and I don't envision OS X making a big dent in the business world anytime in the near future, x86 or not.

Alex

No offense but I'm going to assume you're not in the business of dealing with licensing and support of Windows units. Licensing 6.0 from MS is absolutely frustrating the hell out of companies. There is a very large number of companies being polled regularly that are stating they want to look into alternatives. Large numbers like 20-45% or so. It's pretty staggering actually.

Linux gets mentioned typically but I can safely say that Linux is not even close to ready for such a primetime move. A few servers, yes, but not at the level needed. I don't see this licensing move changing for awhile as MS has so much control over the market that this is their only way to attempt keeping stockholders happy with constant revenue coming in.

Frankly Apple needs help to break into business markets. A big name like IBM would make all of the difference in the world. The fact that they are pushing the Unix'ness of the OS does help it in arguing the whole Mac OS X is unproven.

Keep in mind that I'm looking at Macs moving into a 10-15% marketshare number. There is nothing absolute about this in that it would be so good to be true that everyone wanted to do it.

They have to change their business model though. I think they can compete with MS better than they can Dell, HP, and others with x86 boxes.

[eddively]

did like 1/3 of the people say this is a positive thing. How could Apple not using: the best chip as far as we know, a good thing?

Regardless, they better have something else up their sleeves, or the article is wrong, which it could be. Motorola better be working that .09nm..and it better actually be a PPC chip for desktops etc, instead of an embedded chip only--which I am pretty sure it will be, since it is Moto and all.

[aasmund]

Remember when apple switched to PowerPC? Back then everyone thought that RISC processors was the future and that x86 would die, furthermore Mirosoft had a PPC version of windows and IBM was shipping PPC based PC's. However then Intel sat down in its lab and cooked up some mighty powerfull x86 based chips and PPC based PCs were history.

So Apple wanted to be running on the same platform as PC's. And I think they still want to. They don't want to be stuck in PPC hinterland. obviously if Motorola cooks up a G5 before christhmas they will play along for a while, however if Moto still doesnt get it's act together I feel pretty certain that Apple will switch to some kind of PC like architecture.

Offcourse they have benn known to make stupid decisions - like not embracing USB 2.0 etc. But I think they have been pondering too long about this to make a bad decision.

[yankeedoodle]

Originally posted by Choppaface
well i hope apple gets its act together somehow....jobs said that he envisioned catching up with intel in late 2003, so he's got a year to do it....

Where do you have that statement from? Did I miss something?

[eddively]

Originally posted by Cappy
I just find it fascinating that people are so hung up on this that they're parsing what someone stated they heard from someone else who stated what they heard from someone else who stated what someone else heard. How much farther might that go beyond what's being posted?

How in the freaking world can you "factually" derive anything from that?!?!?!

C'mon people...parsing posts and breaking them down can look cool and impressive but just because it's posted on the internet doesn't mean it's true.

Just cuz you say something, or see something or use something, doesn't mean it is true either. You could be fooled. I don't think that a mac foru is the proper place for a "theoretical reality" debate.
Regardless...I'm gonna go for actually existing.

[eddively]

Originally posted by yankeedoodle


Where do you have that statement from? Did I miss something?

Yeah, I can't place where they said it, but I'm positive I've heard that before from .. I think it was Steve-o.

[dekator]

Originally posted by aasmund


So Apple wanted to be running on the same platform as PC's. And I think they still want to. They don't want to be stuck in PPC hinterland.
<Sigh> How often has it to be said ?
IBM is forcing its RISC architecture, Intel has almost completed its next generation ***RISC*** processor. RISC will be the future and it is not "hinterland" at all. It's the front. It's possible that Apple will move to Itanium but they will certainly never switch to x86 (CISC) which is basically dying... No more circles to be squeezed out of that.

L'shana tova, Quitesure !

[LethalWolfe]

...we're crossing lines here. Avid DS and Symphony are compositing/FX and finishing proggies. Not editing proggies. Film/Media composer are examples of Avid's editing proggies. And I've never seen an editing workstation at a post house that wasn't a Mac. I'm sure some people do use PCs, but the editing side of things is very Mac heavy. Now the finishing/compositing area is a different beast, and I even conceded that in my first post. Even though I've seen more Discreet proggies (Smoke/flame etc.,.) than I have Symphony or DS (granted DS is the new kid on the block).


And I don't know where you (e-coli) got the idea that you can't have an uncompressed Avid based editor on a Mac 'cause you can buy uncompressed flavors of Media/Film Composer. But I think that is a moot point 'cause many places (at least that I've been to here in the States) might have 5 or 6 off-line editors and only one on-line SGI machine running something from Discreet (or a similar ratio).

And I don't see Apple losing the post industry (read "editing") to PCs 'cause FCP is maturing very fast, has a huge buzz surrounding it, and houses in LA are already starting to switch to FCP 3.0 in favor of upgrading to new Avid MC systems.

In summary, in regards to editing Apple is still sitting pretty (and is gaining ground in the FCP vs. M/FC battle). In regards to FX, compositing, and finishing Apple can't really compete.


Lethal

[beatle888]

Originally posted by Kethoticus
I said this Power4 stuff was BS, and if this guy's testimony is valid, I was right.

Maybe this was referring to the Power4 and not a modified PPC based on the Power4. I don't know. Maybe Apple declined the offer because they do have something up their sleeve that's unbelievable. Of course I have nothing authoritative to offer on that front.

But what I can say is this: every few months I hear some new, incredible rumor. It promises something near-unimaginable on the horizon, something that every Mac fanatic raves will bury the Wintel duopoly once and for all.

How long have I been hearing this stuff? For years. The original Power PC chip was supposed to accomplish this overthrow 9 years ago. Then OS X was. Then the G4 chip, with its altivec instruction set, was. There have been other false promises in between, if memory serves. They've all consistently led to the following:

a) Disappointment on the parts of some Mac fans, who usually threaten to leave the Mac platform once and for all and / or preach its imminent doom;

b) Mac apologists who do just the opposite and make excuses for the disappointment, then latch on to the next too-good-to-be-true false promise / rumor and begin waving the Mac flag in their drooling, Apple's-going-to-conquer-the-world-someday patriotism once again.

I feel like I'm the only one who actually perceives this cycle and is sick of it. It's a joke. People, save yourselves the emotional energy and mental "blue balls" and wait to see what actually transpires. And do not base purchasing decisions on rumors, particularly those that seem too good to be true. I bought a Mac in 1999 because I heard that OS X was coming out in Jan 2000. How long did I have to wait for it to be finally released?!? Just a case in point.

Folks, if you need super speed, don't hope for the Mac to come out with a 2GHz G5 next January. It ain't gonna happen. Get a fast PC because that is what is available now. Trash all hopes you have in vaporware. And if you don't need speed, then the Mac is a wonderful box that can do many consumer-oriented things quite elegantly (PC World's recent home DV editing articles attesting to that).

Okay, I stand ready to weather your flames.


here! here! hehe, i agree with you 100%
this is fact, yet this is a rumor site, anywhoo
mac will get back on track, but i wouldnt wait
around for them to, thats silly, get what you
need now, where ever you can.

[Kethoticus]

the site you are currently reading is called macrumors. that's what it's all about fella, rumors.

You seem to be one of the few people who understands this. For many, these rumors turn into anticipated fact. Then they crash and produce disappointment and denial.

When I tried to express my belief that this chip was never going to see the light of day in Macs, many came forward and told me why they felt I was wrong. And they presented good arguments, too. For many, these rumors are equivalent to official announcements. For a long time, that's how I perceived them as well.

[reyesmac]

What I get mad about is the fact that I have to wait 4 years to go from 350mhz to 867mhz and I have to pay more now than I did in the past. Technology should get cheaper and faster. If you bought the high end iMac 2 years ago, you had a 500mhz machine and it cost $1,500, now it would cost you $2,000 and you would only go up to 800mhz with only altivec giving it the noticeable speed boost. In a bad economy, why would they raise prices? They thought things would be doing better by now and so they saw nothing wrong in making more expensive computers. They are wrong, if they decided to keep the iMac prices the same, they would still be selling like hotcakes reguardless of speed.
I would never switch to the dark side, but they at least know their processors will almost double every year. If we had some assurance of that on the Mac side, I think rumor sites wouldnt have become as popular. Instead what we used to get was 50mhz upgrades and now 100mhz upgrades. As far as I am concerned, the G4 is just like the 604chip. The 604 was faster than other chips, but not much faster than the 603 and was soon eclipesd by the truly faster G3. If moto would have been on the ball, we would have had the G5 by now. Now when we do get the G5, all those iMacs with G3's in them will look old and week, the stopgap G4's will be seen for what they really were.
Apple needs to stop planning for the future and just use a chip that is fast that exists NOW, not one that is on the drawing boards. And a chip that will not be in short supply like every other G4 they had.

[MacCoaster]

Originally posted by dekator

<Sigh> How often has it to be said ?
IBM is forcing its RISC architecture, Intel has almost completed its next generation ***RISC*** processor. RISC will be the future and it is not "hinterland" at all. It's the front. It's possible that Apple will move to Itanium but they will certainly never switch to x86 (CISC) which is basically dying... No more circles to be squeezed out of that.

L'shana tova, Quitesure !
Itanium is *NOT* RISC. It's EPIC.
RISC isn't exactly the future. Today's processors have gotten so complex that RISC/CISC/EPIC doesn't really matter anymore.

[MisterMe]

Originally posted by MacCoaster

Itanium is *NOT* RISC. It's EPIC.
RISC isn't exactly the future. Today's processors have gotten so complex that RISC/CISC/EPIC doesn't really matter anymore.

EPIC is RISC by another acronym that Intel coined because it doesn't want to admit that the RISC developers were correct. Remember that the Itanium at it core is a HP-PA chip set implemented on a single chip. HP-PA is RISC. If HP-PA is RISC, then the Itanium is RISC no matter what Intel and its apologists claim.

[alex_ant]

Originally posted by MisterMe
EPIC is RISC by another acronym that Intel coined because it doesn't want to admit that the RISC developers were correct. Remember that the Itanium at it core is a HP-PA chip set implemented on a single chip. HP-PA is RISC. If HP-PA is RISC, then the Itanium is RISC no matter what Intel and its apologists claim.
RISC and CISC used to be two different processor design philosophies - they were never anything more than philosophies, though. Calling a processor RISC nowadays means nothing, because the terms RISC and CISC are obsolete. There is no such thing as a modern pure RISC or pure CISC CPU; the two concepts have merged into whatever you want to call what they are today. Although Itanium descends from the HPPA, and is able to run HPPA code, it is not an HP-PA. (The HPPA was never pure RISC, anyway.) EPIC is a more accurate description for the Itanium because that term entails the technologies Itanium uses that other current CPUs don't (and the technologies Itanium doesn't use that other CPUs do).

Alex

[Scottgfx]

Originally posted by LethalWolfe
...we're crossing lines here. Avid DS and Symphony are compositing/FX and finishing proggies. Not editing proggies. Film/Media composer are examples of Avid's editing proggies. And I've never seen an editing workstation at a post house that wasn't a Mac. I'm sure some people do use PCs, but the editing side of things is very Mac heavy. Now the finishing/compositing area is a different beast, and I even conceded that in my first post. Even though I've seen more Discreet proggies (Smoke/flame etc.,.) than I have Symphony or DS (granted DS is the new kid on the block).


And I don't know where you (e-coli) got the idea that you can't have an uncompressed Avid based editor on a Mac 'cause you can buy uncompressed flavors of Media/Film Composer. But I think that is a moot point 'cause many places (at least that I've been to here in the States) might have 5 or 6 off-line editors and only one on-line SGI machine running something from Discreet (or a similar ratio).

And I don't see Apple losing the post industry (read "editing") to PCs 'cause FCP is maturing very fast, has a huge buzz surrounding it, and houses in LA are already starting to switch to FCP 3.0 in favor of upgrading to new Avid MC systems.

In summary, in regards to editing Apple is still sitting pretty (and is gaining ground in the FCP vs. M/FC battle). In regards to FX, compositing, and finishing Apple can't really compete.


Lethal

From my friends and relatives in the Atlanta post scene, most of the Avids are PC based these days.

When a reseller came to my TV station a few months ago to demo Media Composer, they brought a Compaq PC based system. It was the worst Avid demo I have ever seen!

At a very large media company in Atlanta, (Think really, really big) they have an entire floor of a building outfitted with Avids. All are PC based.

Apple's situation with Avid pre-dates the release of FCP. I heard stories way back in 1996 that Avid wasn't going to develop Media Composer for Copeland. I never really heard why. Are there any discussions as to why Copeland was a failure?

[AmigaMac]

http://www.osnews.com/story.php?news_id=1694

go and make your vote count!

[MacBandit]

Originally posted by Scottgfx


From my friends and relatives in the Atlanta post scene, most of the Avids are PC based these days.

When a reseller came to my TV station a few months ago to demo Media Composer, they brought a Compaq PC based system. It was the worst Avid demo I have ever seen!

At a very large media company in Atlanta, (Think really, really big) they have an entire floor of a building outfitted with Avids. All are PC based.

Apple's situation with Avid pre-dates the release of FCP. I heard stories way back in 1996 that Avid wasn't going to develop Media Composer for Copeland. I never really heard why. Are there any discussions as to why Copeland was a failure?


Think bloatware. They really didn't have a good base to build on like OSX does. So it went in to development hell where the time and money build it was indefinite. So what do you do at that point? Scrap it, fire a few people, hope you have enough money to start over, and ask an old friend for some help.

[ibjoshua]

Originally posted by Kethoticus
For many, these rumors are equivalent to official announcements. For a long time, that's how I perceived them as well.

point taken

just try not to get wound up about it. ;)

i_b_joshua

[Pauls]

I seem to remember on related threads on the new IBM chip, that some other sources close to IBM and others to Motorola that the switch to the new IBM processor was going to happen.

Considering that there is not substatiation in any of these rumours, it's merely one rumour against another. Which do you believe. I think they cancel each other out and we're back to square one.

Apple has options with PowerPC from both vendors, just like Microsoft has with Intel and AMD etc. Processor speed ultimately won't be the thing that will make Apple survive. Systems are getting fast enough now that soon benchmarks won't matter. What will make Apple survive the market share issue, is that it won't matter if you have a Mac in a Windows world, it will be effortlessly compatible. What will sell a Mac in the near future is 'Wow, I want one of those...' or 'Wow, I want to do that...'

[gopher]

Apple is not behind as some people think. With these applications able to take advantage of Altivec:
http://www.apple.com/powermac/processor.html
Absoft Pro Fortran
Adobe After Effects
Adobe Illustrator
Adobe GoLive
Adobe Photoshop
Adobe Premier
Alias|Wavefront Maya
Apple DVD Studio Pro
Apple Final Cut Pro
Apple iDVD, iMovie, iTunes
Apple Mac OS X
Apple QuickTime
Apple WebObjects
Avid Media Composer
Connectix Virtual PC
Deneba Canvas
Digital Origins EditDV
Discreet Logic Combustion
Discreet Logic Cleaner
Emagic Logic Audio
FileMaker Pro
Heuris MPEG Power Probe
Id Software Quake
Macromedia Dreamweaver
Macromedia FreeHand
Macromedia Fireworks
Macromedia Flash
Maxon Cinema 4D
Media 100 CineStream
Metrowerks CodeWarrior
MOTU Digital Performer
Netscape Navigator
NewTek LightWave 3D
Propellerhead Reason
Sorenson Video Codec
Toon Boom Studio
Wolfram Mathematica

And speeds up to 90% faster than Pentium IV:
http://www.apple.com/powermac/specs.html

thanks to Altivec, it really all depends how well designed the program you are using is at using Altivec. If for some reason your program isn't fast enough for you, complain to the developer, not Apple. Just because your program isn't fast, doesn't mean it all should be assigned blame to hardware or the OS. Before we grouch about whether or not the next Power4 makes it into Macs, let's try to at least get 3rd party software to catch up with hardware. If programs like Genentech's Blast can run 5 times faster on a Mac than a Pentium, it goes to prove software developers for the most part don't understand or don't care how to code their software to take the fullest advantage of the G4. Remarkably, Altivec is optimizing Adobe Premier to some extent, but not enough to make it worth it on a Mac as some people have posted benchmarks showing it isn't well coded enough for the Mac. Funny that Photoshop on the other hand is. I think Premier users need to contact Adobe and ask them why they aren't taking the fullest advantage of the G4 and optimizing their software for the G4.

It doesn't matter how fast the hardware is, if the software isn't there to keep itself optimized with it.

[LethalWolfe]

Originally posted by gopher

And speeds up to 90% faster than Pentium IV:
http://www.apple.com/powermac/specs.html



You made a good point (concerning 3rd party software), but linking to an "offical" Apple benchmark kinda blows yer credibility, IMHO.

Lethal

[alex_ant]

Originally posted by gopher
Apple is not behind as some people think. With these applications able to take advantage of Altivec:
http://www.apple.com/powermac/processor.html
Absoft Pro Fortran
Adobe After Effects
Adobe Illustrator
Adobe GoLive
Adobe Photoshop
Adobe Premier
Alias|Wavefront Maya
Apple DVD Studio Pro
Apple Final Cut Pro
Apple iDVD, iMovie, iTunes
Apple Mac OS X
Apple QuickTime
Apple WebObjects
Avid Media Composer
Connectix Virtual PC
Deneba Canvas
Digital Origins EditDV
Discreet Logic Combustion
Discreet Logic Cleaner
Emagic Logic Audio
FileMaker Pro
Heuris MPEG Power Probe
Id Software Quake
Macromedia Dreamweaver
Macromedia FreeHand
Macromedia Fireworks
Macromedia Flash
Maxon Cinema 4D
Media 100 CineStream
Metrowerks CodeWarrior
MOTU Digital Performer
Netscape Navigator
NewTek LightWave 3D
Propellerhead Reason
Sorenson Video Codec
Toon Boom Studio
Wolfram Mathematica
And it's important to note that, even with various parts of these programs hand-optimized for AltiVec, they still in many cases do worse in task-based benchmarks than their equivalent PC versions. GCC 3.1 in Jaguar has been improved to do some automatic optimization for AltiVec, but the degree to which it does this is nowhere near enough to boost performance to a great degree even in applications that lend themselves to matrix and vector operations. Meaning AltiVec optimization still requires much developer effort.

Not to mention the PCs that do beat the Macs in these benchmarks often cost much less than the Macs they beat. If a dual 1.25GHz G4 can beat a dual 2.2GHz Athlon at whatever task, that's great, but when you consider the Athlon costs half as much, well, that's not so great.
And speeds up to 90% faster than Pentium IV:
http://www.apple.com/powermac/specs.html

Depending on Apple for a source of objective benchmark information, I'm surprised that percentage isn't higher, like 300 or 400%. :rolleyes:
thanks to Altivec, it really all depends how well designed the program you are using is at using Altivec. If for some reason your program isn't fast enough for you, complain to the developer, not Apple. Just because your program isn't fast, doesn't mean it all should be assigned blame to hardware or the OS.
Why not complain to Apple? They're the ones who are selling these CPUs that use non-industry-standard math units that require tons of effort to exploit fully.
Before we grouch about whether or not the next Power4 makes it into Macs, let's try to at least get 3rd party software to catch up with hardware. If programs like Genentech's Blast can run 5 times faster on a Mac than a Pentium, it goes to prove software developers for the most part don't understand or don't care how to code their software to take the fullest advantage of the G4.
It also goes to show that the particular algorithm that Blast employs suits itself well to the AltiVec unit, and that because few other programs are so lucky, they will never achieve anywhere near the performance of Blast, hand-optimized or not.
Remarkably, Altivec is optimizing Adobe Premier to some extent, but not enough to make it worth it on a Mac as some people have posted benchmarks showing it isn't well coded enough for the Mac. Funny that Photoshop on the other hand is. I think Premier users need to contact Adobe and ask them why they aren't taking the fullest advantage of the G4 and optimizing their software for the G4.
The reason is obvious: Optimizing for AltiVec costs time and money. Let's assume that Product X contains two million lines of code. And let's assume that a competent programmer who earns $50,000/year is faced with the task of optimizing 50% (a million lines) of this program for AltiVec. Working at 1,000 lines per 8-hour day (a liberal estimate), it will take him over three years. That's obviously not acceptable, so Company X is forced to hire a few more programmers to make this job possible in a reasonable timeframe - let's say two more programmers, who earn the same amount and together can get the job finished in under a year. The total cost is $150,000 in labor for that year of development. This isn't even taking into account expenses for hardware/software, testing, marketing, etc.

This is $150,000 that Company X never would have had to spend (and doesn't have to spend on anyone ELSE's processor) if Apple's CPU was not so esoteric. It sounds to me like this is much more Apple's problem than third-party developers' problem. "Hey, port your software on over to the Mac. Nevermind that to achieve top performance, you'll need to employ labor-intensive programming tricks which will only run on our processors, and nevermind that the future of our processors isn't looking that great anyway, and who knows, we might jump platforms entirely in the next year without any advance notice, so when and if that happens, tough beans! But port to the Mac anyway! As we said, our CPU roadmap is a trade secret, but if you port to our platform, we'd be happy to provide you with a AltiVec Technical Summary PDF for free download, as thanks for being a valued Macintosh developer."

I just don't understand how the blaming of developers can be justified. If Macs actually had strong general-purpose FPUs, or if Apple would put more effort into its compiler, we wouldn't b ehaving this problem.

Alex

[GPTurismo]

well, not to come in late, but IBM has two chips coming out.

One is based on the Power4 arch while the other is a PowerPC arch with some power4 based enhancements.

The Power4 is strictly for the high end market and PowerPC is for a wide variety of markets.

GPT

[KingRocky]

Apple needs to light a fire under that cell-phone company's arse and tell 'em to shape up, or ship out!!

There are plenty of companies out there who'd love to build processors/chipsets for Apple, but Apple has chosen MOTO as their sole vendor. This keeps prices artificially high and innovation stagnant. The PC world has 4 different companies manufacturing CPUs/chipsets. Prices are low and innovation comes fast and furious.

I really wish that they'd threaten to pick another vendor just so MOTO would get the lead out and make some world-class CPUs at reasonable prices.

The MHz gap isn't the problem; it's the price disparity with PCs that is the problem. Until Apple can stop marketing to Apple customers, they'll never increase their marketshare. And yes, they market to their own customers. When the Joe at CompUSA sees a Compaq for $499, and then an iMac for $799 and a G4 for $1699, which do you think he will buy?

[nixd2001]

Originally posted by KingRocky
The MHz gap isn't the problem; it's the price disparity with PCs that is the problem. Until Apple can stop marketing to Apple customers, they'll never increase their marketshare.

Part of Apple's price disparity is entirely of their own making - they just charge very high prices. For a smidgen over what Apple would charge me for taking a new PMG4 from 256MB to 512MB (i.e for roughly 256MB at Apple prices) I've just bought 1GB. That's nothing to do with Moto prices (although they may well be above the odds due to lack of competition).

[A total of 1.25GB may be overkill at present though - Dervish, as my machine is called, hasn't managed to swallow more than 750MB yet - so I've nearly 512MB sitting there not helping much at present. But I'm working on it!]

[MacBandit]

Originally posted by nixd2001


Part of Apple's price disparity is entirely of their own making - they just charge very high prices. For a smidgen over what Apple would charge me for taking a new PMG4 from 256MB to 512MB (i.e for roughly 256MB at Apple prices) I've just bought 1GB. That's nothing to do with Moto prices (although they may well be above the odds due to lack of competition).

[A total of 1.25GB may be overkill at present though - Dervish, as my machine is called, hasn't managed to swallow more than 750MB yet - so I've nearly 512MB sitting there not helping much at present. But I'm working on it!]

All PC manufacturors charge outrageous prices for upgrading ram etc.. Try outfitting a Dell or Compaq or any other major name brand similarly to a Mac and you will see that the price difference is simply because you get much more on a basic configuration Mac.

[nixd2001]

Originally posted by MacBandit


All PC manufacturors charge outrageous prices for upgrading ram etc.. Try outfitting a Dell or Compaq or any other major name brand similarly to a Mac and you will see that the price difference is simply because you get much more on a basic configuration Mac.

ok - never tried it either. Never bought a big name PC, other than the Vaio laptop, but I've never upgraded the memory in that.

[KingRocky]

Originally posted by MacBandit


All PC manufacturors charge outrageous prices for upgrading ram etc.. Try outfitting a Dell or Compaq or any other major name brand similarly to a Mac and you will see that the price difference is simply because you get much more on a basic configuration Mac.


ummm. . .

I don't know where you get your prices from, but I can get an AMD AthlonXP 2200+ for a whopping $145 and 512MB of PC2100 DDR SDRAM for $95.

An upgrade to Mac's fastest processor costs OVER 4 TIMES AS MUCH!!! Not to mention that most Macs still use slow PC100 SDRAM.

A processor upgrade for my G3 or G4 COSTS AS MUCH AS AN ENTIRE PC.

And there's no difference in hardware levels between Mac and a typical mid-range PC. Both have DVD-RWs, lots of RAM, big HDs, ethernet, USB, etc. Apple simply beats the PC in execution, that's all. Nice, pretty acrylic cases, mice, keyboards and speakers. All of which add $ to the price, as well as that overpriced CPU.

If Apple wants to increase marketshare, they need to be offering a BETTER DEAL than what the PC manufacturers are offering, not just the same deal.

[Telomar]

Originally posted by alex_ant

The reason is obvious: Optimizing for AltiVec costs time and money. Let's assume that Product X contains two million lines of code. And let's assume that a competent programmer who earns $50,000/year is faced with the task of optimizing 50% (a million lines) of this program for AltiVec. Working at 1,000 lines per 8-hour day (a liberal estimate), it will take him over three years. ...

This is $150,000 that Company X never would have had to spend (and doesn't have to spend on anyone ELSE's processor) if Apple's CPU was not so esoteric. It sounds to me like this is much more Apple's problem than third-party developers' problem.
...
As we said, our CPU roadmap is a trade secret, but if you port to our platform, we'd be happy to provide you with a AltiVec Technical Summary PDF for free download, as thanks for being a valued Macintosh developer."

I just don't understand how the blaming of developers can be justified. If Macs actually had strong general-purpose FPUs, or if Apple would put more effort into its compiler, we wouldn't b ehaving this problem.

Alex
For a start I stuggle to think of any program of reasonable complexicty where you would optimise anywhere near 50% of the code for AltiVec. It is a very specialised unit. I mean do you seriously believe 50% of office could be done with AltiVec?

Secondly the troubles faced with programming for Apple's vector unit are no differnent than those you face when programming for Intel's.
At the end of the day for the best performance on the PC side you still have to get in there and do a lot of the coding by hand, which is just as slow as for the Velocity Engine. The main reason for greater adoption on the PC side is the increased competition, which largely doesn't exist on the Mac side. It's an easy way to differentiate performance and that is the first thing people look at.

The only other major difference is the compilers on the PC side are better but I'd expect Apple to keep working on their own. Apple does have a research group that does nothing but focus on higher level methods of programming for altivec or reprograms libraries to take advantage of altivec. For instance with Jaguar came a whole bunch of math libraries optimised for altivec. All you need to do is integrate them and the code is optimised for certain functions already.

Apple is more secretive with their roadmap than most but they do release some details to certain people. I can say their future roadmaps have Altivec compatibility built in though. They aren't abandoning it. Of course you always need to look at the fine print.

I agree Apple needs to improve their FPU and add a second one and certainly compilers can be improved but that's coming.

[alex_ant]

Originally posted by Telomar
For a start I stuggle to think of any program of reasonable complexicty where you would optimise anywhere near 50% of the code for AltiVec. It is a very specialised unit. I mean do you seriously believe 50% of office could be done with AltiVec?
That is a good point. Because AltiVec is so specialized, it does not lend itself well to many applications, which means that many programs are doomed to running on the G4's slow integer unit and very slow FPU. Which means that any objective performance advantages the G4 has over whatever else deserve an asterisk next to them.
Secondly the troubles faced with programming for Apple's vector unit are no differnent than those you face when programming for Intel's.
At the end of the day for the best performance on the PC side you still have to get in there and do a lot of the coding by hand, which is just as slow as for the Velocity Engine.
My point was that extensive hand-optimization is a requirement to achieve competitive performance on the Mac, whereas great performance is practically guaranteed out-of-the-box on Intel because Intel's compiler is so good - it's able to optimize somewhat well for SSE2 and MMX automatically (better than GCC for PPC optimizes for AltiVec), and it continues to improve. Not to mention the much faster integer and FP units on P4s and Athlons.
The main reason for greater adoption on the PC side is the increased competition, which largely doesn't exist on the Mac side. It's an easy way to differentiate performance and that is the first thing people look at.

The only other major difference is the compilers on the PC side are better but I'd expect Apple to keep working on their own. Apple does have a research group that does nothing but focus on higher level methods of programming for altivec or reprograms libraries to take advantage of altivec. For instance with Jaguar came a whole bunch of math libraries optimised for altivec. All you need to do is integrate them and the code is optimised for certain functions already.
This is good, and it's what Apple needs if it wants to stay on the AltiVec path, but they need to do a better job of making those research dollars translate into benchmark points. Sure Apple is improving - it's processors are getting faster, and its math libraries are getting more efficient, etc. - but the problem is that x86 is improving faster.

Alex

[ibjoshua]

Originally posted by KingRocky
There are plenty of companies out there who'd love to build processors/chipsets for Apple...
i'm curious. who wants to make chips for Apple? just asking.

.. but Apple has chosen MOTO as their sole vendor.
Motorola is not a sole vendor. IBM also make chips for Apple.

[uberman42]

What is the deal about Apples using sun Sparc chips. A friend of mine @ sun mentioned that there are mules in cupertino running sparc boxes. can someone substantiate this

[senjaz]

Grrrr.

I'm normally passive and enjoy reading most of the discussions here. I wish that Alex would follow my lead and stay quiet unless he actually has something useful to say. Always trying to sound knowledgable when he obviously knows next to nothing on what he chooses to sound off about.

Here we go with AltiVec...

Very little of an average application can be vectorised and the parts that are, are often not difficult to. You can spend time finding an optimal algorithm for a particular problem to get the best gain from the vector unit but getting a general performance increase is not difficult.

Apple has some examples covering common optimisations here: http://developer.apple.com/samplecode/Sample_Code/Devices_and_Hardware/Velocity_Engine.htm

Coding for AltiVec is as simple as including a header file <AltiVec.h> and lining up data for it's vec_ C instructions. I don't profess to doing any MMX coding but that always used to be hand coded in assembler, with support for autovectorising code in some recent compilers.

The G4 processor has a weak floating point unit but doesn't really need anything more. Most math intensive code can be vectorised. When it can't be the general floating point unit is sufficient. (I believe that the G4 actually has 2 floating point units but that each only deals with certain types of operations)

The main weakness of AltiVec is it's lack of support for double precision floats. However for higher precision than this where processors don't normally have native support, better performance can be achieved with AltiVec again: http://developer.apple.com/hardware/ve/pdf/oct3a.pdf

On to benchmarks. They mean next to nothing as they are abused so often, especially the spec bench marks that the Intel world is so fond of quoting. First, the Spec bench marks do not stress the overall system which is important if you want to use benchmarks to base expectations for real world results. Second the spec benchmarks are run as-is. The only optimisations that can take place on the code must be done by the compiler, so unless you have an auto-vectorising compiler you will not get any benefit for that benchmark. Thus a chip with an inferior vector unit (P4) with an auto vectorising compiler can get a better spec result than a superior vector unit (G4) without auto vectorisation, because the vector unit is not being used.

If you have any doubt on which is superior. Arstechnica is a fantastic site for info on such things: http://www.arstechnica.com/cpu/1q00/simd/simd-1.html
http://arstechnica.com/cpu/01q2/p4andg4e/p4andg4e-1.html
http://www.arstechnica.com/cpu/1q00/g4vsk7/g4vsk7-1.html

And especially relevant to this:

http://www.arstechnica.com/cpu/2q99/benchmarking-1.html

Although modern compilers are good they can't know the context of the problem you are trying to solve. Optimisation by hand will always have the potential for better results.

Apple needs to add auto vectorising support to GCC, I do not believe it does so at all right now, even GCC 3. Apple knows it needs to improve the compiler and has in fact been hard at work doing that. GCC 3 has many improvements rolled into it that are from Apple many to do with vector support other general optimisations.

Apple should continue what they are doing now, improving the compiler in a way that benefits us developers and not aiming for high bench mark scores that only the uneducated seem to deem as relavent.

For reference, I am a Java/C programmer working for Marconi plc. I get to play with Objective-C and Macs in my spare time.

[KingRocky]

Originally posted by i_b_joshua

i'm curious. who wants to make chips for Apple? just asking.


Motorola is not a sole vendor. IBM also make chips for Apple.

You are correct about IBM. However, IBM doesn't seem interested at the moment in sharing their shiny new Power4, which I'm sure spanks some G4 booty.


And what vendor wouldn't want to expand their business by manufacturing CPUs for the Macintosh?

Although I'm sure that the costs of licensing the PowerPC/AltiVec architecture would be cost-prohibitive, any of the Taiwanese manufacturers could do it--VIA, SiS, UMC, etc. There's no special magic to processor manufacturing that only MOTO/IBM know.

I don't think that AMD would be interested--they've already invested heavily in the X86-64 architecture for their next-generation chip.

[thegrayrace]

Originally posted by bertinman

I still feel that intel based osx is not the best idea none-the-less, because it would ruin Apple's control of hardware which is why Apple's systems are so great--They work together.

-- bert :cool:

Wow. There are really a LOT of misconceptions about the possibilities of Apple going with an AMD or Intel processors.

It would not ruin Apple's control over the integration of their hardware at all. Apple would *never* enable Mac OS X to be installed on any AMD/Intel box, that would be a huge disaster. There is no question that if Apple went x86, Mac OS X for x86 would only run on Apple hardware. This is not a difficult task to accomplish, at all.

[KingRocky]

Originally posted by thegrayrace


It would not ruin Apple's control over the integration of their hardware at all. Apple would *never* enable Mac OS X to be installed on any AMD/Intel box, that would be a huge disaster. There is no question that if Apple went x86, Mac OS X for x86 would only run on Apple hardware. This is not a difficult task to accomplish, at all.

Where there's a will, there's a crack. . .

M$ said that you had to register XP--cracked
M$ said that XP SP1 would fix the problem--cracked

You underestimate the time and resources of the hacker community. No, I'm not one myself.

MacOS on X86 hardware would be a step backwards unless it's on a 64-bit platform. But that's beside the point. There is NOTHING WRONG with the current PPC/AltiVec architecture that a little competition won't fix. With Apple being it's sole customer for G4 chips, MOTO isn't interested in spending the time and $ in turning the G-series into a real world-class processor.

[thegrayrace]

Originally posted by Kethoticus
But what I can say is this: every few months I hear some new, incredible rumor. It promises something near-unimaginable on the horizon, something that every Mac fanatic raves will bury the Wintel duopoly once and for all.

How long have I been hearing this stuff? For years. The original Power PC chip was supposed to accomplish this overthrow 9 years ago. Then OS X was. Then the G4 chip, with its altivec instruction set, was. There have been other false promises in between, if memory serves. They've all consistently led to the following:

While I mostly agree with you, you have to admit that when it comes to raw power, the Macintosh platform has lost the lead it once had. The 680x0 chips blew away the x86 chips available at the time. The PPC 60x chips blew away the x86 chips available at the time. The G3, when it was first released, blew away the x86 chips available at the time. But in the last 4 years, Apple has fallen far behind in the raw power of their hardware (unless you believe 5 or 6 out of a few hundred Photoshop benchmarks equals raw power).

However, the POWER4-based desktop chip is not a pipe dream. It exists, and it could very well put Apple back on top in the hardware arena. Or, at the very least, allow them to catch up on the significant ground they have lost.

People who proclaim Apple is doing everything fine and hasn't fallen behind don't see what I see. I work at an Apple-only reseller/service center, and the graphics, video, and audio professionals aren't falling for Apple's benchmarks. We're constantly losing the high-end customers to the x86 world due to the speed issue. iBook, PowerBook, and iMac sales have remained pretty constant the 2.5 years I've been here, but Power Macintosh sales have noticibly decreased. Those Power Macintosh towers we do sell are to Mac-faithful, not new customers. Apple is behind in this area and they need to catch up if they want to remain strong in a market where speed is a MAJOR issue.

[thegrayrace]

Originally posted by KingRocky
Where there's a will, there's a crack. . .

M$ said that you had to register XP--cracked
M$ said that XP SP1 would fix the problem--cracked

You underestimate the time and resources of the hacker community. No, I'm not one myself.


It isn't about simply cracking a method of software protection, although there would surely be software-hardware verification. The logic board (and ROM on the logic board) is just as significant as the processor.

There are other systems out there using the PPC chip.

IBM had servers which used the G3 chip. Mac OS X hasn't been hacked to run on those machines to the best of my knowledge. They won't run OS 8.x or 9.x.

The BeBox used 603e chips, it can't run OS 8.x or 9.x.

Correct me if I'm wrong.

[KingRocky]

I was speaking strictly from a PC user's perspective. I honestly don't know of ANY hacked or cracked Macintosh software/OS.

However. . .

Since OS X is UNIX based, I can forsee a time when somebody out there gets it to run on their PC. Granted, it would take a BIG effort, but I have no doubt that it could be done.

[thegrayrace]

Originally posted by KingRocky


Where there's a will, there's a crack. . .

M$ said that you had to register XP--cracked
M$ said that XP SP1 would fix the problem--cracked

You underestimate the time and resources of the hacker community. No, I'm not one myself.

Also, if Apple was to go x86, they'd most likely have AMD and/or Intel manufacturing them slightly modified chips. Again, not difficult to accomplish.

It would NOT be as easy as building your own x86 box and tossing in your hacked version of Mac OS X. There likely would be all sorts of hardware modifications that would need to be done. Drivers for all third-party hardware would need to be written or compiled for OS X (kind of hard to do when the system won't even boot off of your motherboard, use your video card, recognize your keyboard). Nothing is impossible, but it is unlikely, and even if someone pulled it off, what's it hurt Apple if a few dozen people are running an unstable hacked version of Mac OS X on unsupported hardware?

They'd only be worried about the average Joe being able to install Mac OS X on his Dell or Compaq or the machine he built with parts at Fry's Electronics, which would not be happening. The hardcore hackers and hardware engineers aren't going to be a major concern anyway.

[kenohki]

Originally posted by thegrayrace


It isn't about simply cracking a method of software protection, although there would surely be software-hardware verification. The logic board (and ROM on the logic board) is just as significant as the processor.

There are other systems out there using the PPC chip.

IBM had servers which used the G3 chip. Mac OS X hasn't been hacked to run on those machines to the best of my knowledge. They won't run OS 8.x or 9.x.

The BeBox used 603e chips, it can't run OS 8.x or 9.x.

Correct me if I'm wrong.

However, if you ran Mac OS X on Intel hardware you suddenly attract the attention of all the Windows and Linux hackers/crackers who will do it just to say they can.

IBM's PPC machines would never have been a target for a Mac OS mule because they're so damn expensive. No one wants to pay more than even Apple prices to run Mac OS. Pretty much the same story with BeBox. However, if you can buy a cheap, build-it-yourself PC and get a hacked version of Mac OS X to run, that might be interesting for the price (especially to that crowd).

If you need an example, just look at how the game hackers are beating Microsoft's hardware/software copy protection duo on the XBox.

[thegrayrace]

Originally posted by kenohki


However, if you ran Mac OS X on Intel hardware you suddenly attract the attention of all the Windows and Linux hackers/crackers who will do it just to say they can.

IBM's PPC machines would never have been a target for a Mac OS mule because they're so damn expensive. No one wants to pay more than even Apple prices to run Mac OS. Pretty much the same story with BeBox. However, if you can buy a cheap, build-it-yourself PC and get a hacked version of Mac OS X to run, that might be interesting for the price (especially to that crowd).

If you need an example, just look at how the game hackers are beating Microsoft's hardware/software copy protection duo on the XBox.

True, it would attract more attention, and sure, it would be possible to hack, no question about that.

However, Apple, with relative ease, could put all sorts of hardware-software verification schemes throughout the operating system, not only in the installation. It would be tough work to break. Every "Security Update" one installs via Software Update could patch the system to perform hardware-software verification again. =)

Drivers for all the third-party hardware would also be a major issue. Someone could perhaps create a hacked build of Mac OS X to run on some specific hardware, but we'd never see a hacked version of Mac OS X that included even 5% of the drivers that are included in Windows XP, for example. All third-party hardware would need drivers which would have to be written and compiled specifically for Mac OS X, and these would have to be on the hacked install CD. To do that, one would have to hack a copy of Mac OS X already installed on Apple hardware to remove any hardware-software verification. Once that is done, they'd have to compile an install CD that includes OS X drivers for all the hardware they intend to use. Not simple tasks.

Again, it's all possible, it's just not anything I think Apple needs to worry about severely hurting their hardware sales if they went the x86 route. I certainly would prefer they went with the POWER4-based chip. But a lot of people have the misconception that if Apple went with x86, anybody would be able to grab a retail copy of Mac OS X for x86 and install it on their Dell.

[chubakka]

Apple isn't moving to x86. And talking about it aint gonna happen.

[alex_ant]

Originally posted by senjaz
Very little of an average application can be vectorised and the parts that are, are often not difficult to. You can spend time finding an optimal algorithm for a particular problem to get the best gain from the vector unit but getting a general performance increase is not difficult.
...

The G4 processor has a weak floating point unit but doesn't really need anything more. Most math intensive code can be vectorised. When it can't be the general floating point unit is sufficient. (I believe that the G4 actually has 2 floating point units but that each only deals with certain types of operations)

The main weakness of AltiVec is it's lack of support for double precision floats. However for higher precision than this where processors don't normally have native support, better performance can be achieved with AltiVec again: http://developer.apple.com/hardware/ve/pdf/oct3a.pdf

The point still remains that Apple is asking developers to re-write however much of their code for this oddball AltiVec thing that nobody else uses. If it were so easy, all Mac developers would have done it by now. Until all desktop processors have similarly functioning vector processors onboard, or until whichever OS X compiler is capable of useful auto-vectorization, developers - if they want to involve themselves with the Mac - are forced into doing something they shouldn't have to do.
On to benchmarks. They mean next to nothing as they are abused so often, especially the spec bench marks that the Intel world is so fond of quoting. First, the Spec bench marks do not stress the overall system which is important if you want to use benchmarks to base expectations for real world results. Second the spec benchmarks are run as-is. The only optimisations that can take place on the code must be done by the compiler, so unless you have an auto-vectorising compiler you will not get any benefit for that benchmark. Thus a chip with an inferior vector unit (P4) with an auto vectorising compiler can get a better spec result than a superior vector unit (G4) without auto vectorisation, because the vector unit is not being used.
That's the whole point of SPEC, and it does serve a useful purpose. It's designed to approximate general performance of identical code across different platforms. If the test suite were allowed to be modified, then it wouldn't be a cross-platform benchmark anymore. There are two ways of looking at CPU-bound benchmarks like SPEC:

1) They're good, because they are decent approximators of general real-world CPU performance, and other factors of overall system performance (disk I/O, video, etc.) are likely similar enough across the Mac-PC divide that they won't account for too much disparity from this angle.

2) They're bad, because they do not have any bearing on the maximum potential of a processor. Well, Apple can go on and on about how the dual 1GHz is capable of 15 gigaflops and so on, but until the dual GHz is actually capable of regularly achieving 15 gigaflops on general floating-point tasks, then that number is theoretical and meaningless to the end user. Maybe it applies to a couple Photoshop filters.
Although modern compilers are good they can't know the context of the problem you are trying to solve. Optimisation by hand will always have the potential for better results.
But that means less when the x86 world is much faster by default than the Mac world. A developer today has a choice: 1) writing standard cross-platform code that will run pretty well by default on the platform that 95% of the desktop market uses, and 2) writing the same code and then combing through however much of it by hand to optimize it for what less than 5% of the desktop market uses, ultimately achieving performance anywhere that still might not be much better than x86. Of course it all boils down to what sells, but people have a tendency of being partial to what performs best.
Apple should continue what they are doing now, improving the compiler in a way that benefits us developers and not aiming for high bench mark scores that only the uneducated seem to deem as relavent.
If Apple continues what they're doing now, they will only fall further and further behind the rest of the world. Although improving the compiler is great, and it may be their only short-term option, they need faster chips in order to stay competitive, and those chips have to be able to achieve good performance without significant developer effort.

Alex

[alex_ant]

Originally posted by thegrayrace
However, Apple, with relative ease, could put all sorts of hardware-software verification schemes throughout the operating system, not only in the installation. It would be tough work to break. Every "Security Update" one installs via Software Update could patch the system to perform hardware-software verification again. =)
Wasn't it Steve Jobs who said (to paraphrase) "Every security scheme based on secrets eventually fails"? I believe this would be the case with an x86 Mac. You can put all the secret encryption chips or whatever onto the motherboard that you want, but it would only be a matter of time before the proprietary x86 Mac would be cracked wide open and low-level emulators became available for any PC under the sun. The Linux fanboys would be on this thing like flies on ****.

Not to mention the questionable probability of success of an x86 Mac...

All this talk of hacking just makes me uneasy. It's against what the Mac is all about: Simplicity. People shouldn't have to understand that the OS is separate from the hardware and that it needs special drivers and the manufacturers haven't released drivers for OS X-x86 yet and software publishers haven't recompiled their code yet and blah blah blah. They want a computer that works, and that's all. A hardware branch would upset this.

Alex

[nixd2001]

Originally posted by alex_ant

Wasn't it Steve Jobs who said (to paraphrase) "Every security scheme based on secrets eventually fails"?

He may have said it, but I doubt he originated the idea.

[nixd2001]

A lot of the underlying ideas I agree with here. But not some of the actual arguments used. So, at the danger of being labeled a troll, here's some comments (and don't take these personally - they're comments thrown to all the readers):

Originally posted by alex_ant

The point still remains that Apple is asking developers to re-write however much of their code

Much is the wrong word. The rewrite - ie away from C (or whatever) and to something to get benefit from AltiVec - is for the purposes of performance. It is extremely rare of performance bottlenecks to be anything other hotspots - the old 90% of the time in 10% of the code adage. There's not much to do for a word processor, other than leaning heavily on class libraries that Apple has spent lots of time working on overall (one hopes). But anything that could be considered to be performing data processing (think anything with audio, video, graphics, FP data sets) is likely to be extremely amenable to optimisations in small, key locations.

So I agree with the sentiment that it's a shame that you have to recode some things, but thing the balance isn't quite right here.


for this oddball AltiVec thing that nobody else uses.

It's probably as oddball as MMX, etc. In fact it's not at all oddball if you consider some of the processors outside the "general processor" category. Large register, split carry, repeated operation.... TI had a graphics co-proc doing this a good decade ago (340 series I think). And it's not that odd if you venture in to the world of DSPs, differing word sizes, true Harvard architecture, etc. I used an Analog Devices SHARC DSP a few years ago - data items in registers manipulated as 16, 32, or 48 bits, with data memory banked into units where an incremement of 1 moved up by one of these units (ie there was simply no concept of byte addressing). Oh, and that was VLIW too.

If you've just poured months into a large program, and have some respect for what you're doing, then

1) you should know where the bottlenecks are.
2) you should already have tried to isolate the key routines to make later optimisations easy.
3) the chance to achieve a 10x speed gain is almost irresistable
4) unless you're a VB coder who doesn't know what an interrupt or scalable algorithm is, you'll find ways to optimise if they exist.


If it were so easy, all Mac developers would have done it by now.

Here we start getting close to the real nub - is it really a case of being easy versus difficult, or actually that AltiVec won't make much difference? For comparitively simple operations over large data sets, there are benefits from AltiVec (and the like). But more general computing (let's say a Java VM just to choose a random example) is probably more just data bandwidth bound than anything else. In such situations, there will be no AltiVec optimisations, but because they're not going to contribute much versus being too difficult.

Until all desktop processors have similarly functioning vector processors onboard, or until whichever OS X compiler is capable of useful auto-vectorization, developers - if they want to involve themselves with the Mac - are forced into doing something they shouldn't have to do.


So I don't quite agree with this.

...
But that means less when the x86 world is much faster by default than the Mac world. A developer today has a choice: 1) writing standard cross-platform code that will run pretty well by default on the platform that 95% of the desktop market uses,

This 95% figure is interesting. We've been reading it quite happily for a while now, but maybe it's worth considering what it means.

Does Intel have 95% market share - no AMD ensures that it so.

Does Microsoft have 95% market share - yes (let's assume this at least for the sake of the argument).

Okay - what has Microsoft got 95% market share with? Their operating systems (note the plural). Even this 95% doesn't mean any one of 95, 98, 2000, XP, etc, has 95% market share.

Applications - Let's guess IE is the most used piece of code MS produce. This still probably has less than 95% share - not all MS users are necessarily using the 'net and some (not many!) are using altnerative browsers.

Office is probably MS's biggest selling application. This sure as hell hasn't got 100% of this 95% to itself.

In other words, this is classic lies, damned lies and statistics. The potential market share for a PC developer may be 95%, but this isn't a realistic figure for even MS, let alone more "lowly" developers.

It all comes down to market segmentation. Adobe may say that PC's represent about 20 times the market as Macs for Photoshop, but I imagine that the Mac community is more focused and would expect a smaller sales ratio than 20:1 in reality. (Can any one be bothered to trawl Adobe's SEC filings and pull out whatever figures they give?)

If we start talking "esoteric" applications, such as professional design (print, video, etc), then this 95% is probably prime MS FUD. Sales to Mac users might actually be higher!

[/b]

[quote] and 2) writing the same code and then combing through however much of it by hand to optimize it

the "by hand" bit is probably wrong. If you don't already know where its slow (eg writing a video application - not difficult to guess!) there are tools that will make this a fairly trivial process.


If Apple continues what they're doing now, they will only fall further and further behind the rest of the world. Although improving the compiler is great, and it may be their only short-term option, they need faster chips in order to stay competitive, and those chips have to be able to achieve good performance without significant developer effort.

Alex

This is where I really get closest to agreeing. But, as that wouldn't be providing troll fodder (not my intention, honest), consider this.

Let's assume we've factored out where AltiVec is relevant (possibly including compilers, but let's not go there).

A lot of the messages here recently seem to have focused very much on clock rates. Some have cited Power4 and derivates, but almost always in an off-hand fashion. But ask the question, what actuallty needs to be faster?

Much as I find it strange to say this, I think the BareFeats numbers might be relevant here. Combine the apparent oddities of these figures with the white paper about G4 upgrades and cache performance, and I think there's something lurking here (not everyone is going to be surprised by this). I think G4 at present is not MHz bound but MB/S bound - ie improving memory bandwidth might have a lot more effect that increasing clock rates or achieving higher IPC/superscalar performance.

Keeping everything about a G4 the same apart from doubling the speed of the memory bus and performing some ripple through on timings might well be a considerably cheaper way of getting better performance than things like Power4Lite.

But if this hypothesis is true, why hasn't it happened? I've no answer to that, other than it's far more in Motorola's (and maybe IBM's) hands than Apple's.

Phew - rant over.

[alex_ant]

Originally posted by nixd2001
Much is the wrong word. The rewrite - ie away from C (or whatever) and to something to get benefit from AltiVec - is for the purposes of performance. It is extremely rare of performance bottlenecks to be anything other hotspots - the old 90% of the time in 10% of the code adage. There's not much to do for a word processor, other than leaning heavily on class libraries that Apple has spent lots of time working on overall (one hopes). But anything that could be considered to be performing data processing (think anything with audio, video, graphics, FP data sets) is likely to be extremely amenable to optimisations in small, key locations.

So I agree with the sentiment that it's a shame that you have to recode some things, but thing the balance isn't quite right here.
These are good points. I didn't say "much" though, I said "however much" - in an ideal world, developers wouldn't have to do anything special to achieve maximum performance on the Mac. But I think we mostly agree here, and I admit my rewriting-50%-of-the-code figure or whatever it was was way high.
It's probably as oddball as MMX, etc.
Well, Intel's compiler has made enough strides that optimizing for MMX and SSE2 to some extent happens automatically. The auto-vectorization isn't perfect, but it's better than none at all (I would guess :)). So that kind of separates the oddballs from the not-quite-as-oddballs.
In fact it's not at all oddball if you consider some of the processors outside the "general processor" category. Large register, split carry, repeated operation.... TI had a graphics co-proc doing this a good decade ago (340 series I think). And it's not that odd if you venture in to the world of DSPs, differing word sizes, true Harvard architecture, etc. I used an Analog Devices SHARC DSP a few years ago - data items in registers manipulated as 16, 32, or 48 bits, with data memory banked into units where an incremement of 1 moved up by one of these units (ie there was simply no concept of byte addressing). Oh, and that was VLIW too.
There are indeed processors much weirder than the G4 out there, but out of the current crop of desktop & low-end server processors, the G4 with its AltiVec is probably the most oddball of the bunch.
If you've just poured months into a large program, and have some respect for what you're doing, then

1) you should know where the bottlenecks are.
2) you should already have tried to isolate the key routines to make later optimisations easy.
3) the chance to achieve a 10x speed gain is almost irresistable
4) unless you're a VB coder who doesn't know what an interrupt or scalable algorithm is, you'll find ways to optimise if they exist.

...
(moved)
the "by hand" bit is probably wrong. If you don't already know where its slow (eg writing a video application - not difficult to guess!) there are tools that will make this a fairly trivial process.
All good points. Perhaps I overestimated how difficult AltiVec accelleration really is. But time and economics do play into this. Every week a programming team spends optimizing is a week the Mac version of Software X has to be delayed, and every dollar that team is paid is a dollar that will be appearing in red on Company X's balance sheets. Probably just about all Mac developers could go hog-wild with AltiVec if they really had to, and still stay in business, but no company likes to spend money it shouldn't have to spend.
Here we start getting close to the real nub - is it really a case of being easy versus difficult, or actually that AltiVec won't make much difference? For comparitively simple operations over large data sets, there are benefits from AltiVec (and the like). But more general computing (let's say a Java VM just to choose a random example) is probably more just data bandwidth bound than anything else. In such situations, there will be no AltiVec optimisations, but because they're not going to contribute much versus being too difficult.
Yup - there are a lot of tasks AltiVec isn't suited for. And this reinforces what I said about Apple needing a CPU with better general-purpose performance.
This 95% figure is interesting. We've been reading it quite happily for a while now, but maybe it's worth considering what it means.

Does Intel have 95% market share - no AMD ensures that it so.

Does Microsoft have 95% market share - yes (let's assume this at least for the sake of the argument).

Okay - what has Microsoft got 95% market share with? Their operating systems (note the plural). Even this 95% doesn't mean any one of 95, 98, 2000, XP, etc, has 95% market share.

Applications - Let's guess IE is the most used piece of code MS produce. This still probably has less than 95% share - not all MS users are necessarily using the 'net and some (not many!) are using altnerative browsers.

Office is probably MS's biggest selling application. This sure as hell hasn't got 100% of this 95% to itself.

In other words, this is classic lies, damned lies and statistics. The potential market share for a PC developer may be 95%, but this isn't a realistic figure for even MS, let alone more "lowly" developers.

It all comes down to market segmentation. Adobe may say that PC's represent about 20 times the market as Macs for Photoshop, but I imagine that the Mac community is more focused and would expect a smaller sales ratio than 20:1 in reality. (Can any one be bothered to trawl Adobe's SEC filings and pull out whatever figures they give?)

If we start talking "esoteric" applications, such as professional design (print, video, etc), then this 95% is probably prime MS FUD. Sales to Mac users might actually be higher!
I think this is my fault... instead of saying "the platform 95% of the market uses," I should have said "the platform comprising 95% of the market in current sales" or something like that. Sure there are tons of legacy Windows PCs, but there are tons of legacy Macs just as well.
This is where I really get closest to agreeing. But, as that wouldn't be providing troll fodder (not my intention, honest), consider this.

Let's assume we've factored out where AltiVec is relevant (possibly including compilers, but let's not go there).

A lot of the messages here recently seem to have focused very much on clock rates. Some have cited Power4 and derivates, but almost always in an off-hand fashion. But ask the question, what actuallty needs to be faster?

Much as I find it strange to say this, I think the BareFeats numbers might be relevant here. Combine the apparent oddities of these figures with the white paper about G4 upgrades and cache performance, and I think there's something lurking here (not everyone is going to be surprised by this). I think G4 at present is not MHz bound but MB/S bound - ie improving memory bandwidth might have a lot more effect that increasing clock rates or achieving higher IPC/superscalar performance.

Keeping everything about a G4 the same apart from doubling the speed of the memory bus and performing some ripple through on timings might well be a considerably cheaper way of getting better performance than things like Power4Lite.

But if this hypothesis is true, why hasn't it happened? I've no answer to that, other than it's far more in Motorola's (and maybe IBM's) hands than Apple's.

Phew - rant over.
If the G4's memory bus were quadrupled tomorrow, that would be fantastic, but I would be happier if I knew that Apple had a strong and capable architecture with a bright, Moore's-Law-compliant future ahead. Increasing the bus speed on the G4 sounds like a nice short-term solution to the x86-PPC performance disparity, but Intel and AMD certainly aren't standing still. They're moving a lot faster than Motorola is, and that's the whole problem. :)

Alex

[beatle888]

Originally posted by nixd2001
Borris
[Would the current claimant to the Borris ID care to to relinquish it, perchance....]

liked your post....hope you get your user id you requested Borris:p