PDA

View Full Version : IBM's 750VX Finalized?


MacRumors
Dec 11, 2003, 01:04 PM
AppleInsider claims (http://www.appleinsider.com/news.php?id=318) that IBM's 750VX ("Mojave") has been finalized but won't see introduction until Q3 2004.

The new chip will reportedly support a 400Mhz DDR front side bus and Altivec.

Older rumors (http://www.macrumors.com/pages/2003/05/20030509020831.shtml) about the chip offered a similar description ("Altivec-enabled G3"). No information of a 750VX chip has yet been released from IBM.

howard
Dec 11, 2003, 01:07 PM
Q3 2004?

thats when most people want G5 powerbooks.

anyone know what kinda speeds this is supposed to run at? maybe it'll be for the ibook or emac

Rocketman
Dec 11, 2003, 01:08 PM
Given the huge architecture improvenments over the Mot G4 this may be the rumored Powerbook G5 chip.

Rocketman

jbrown
Dec 11, 2003, 01:10 PM
but where will it be put???:rolleyes: :rolleyes:

johnnyjibbs
Dec 11, 2003, 01:10 PM
Well, 400MHz FBS is certainly a step-up from the Moto G4...

MattG
Dec 11, 2003, 01:12 PM
What's the clockrate of these things going to be? Could this be the final nail in Motorola's coffin?

ennerseed
Dec 11, 2003, 01:15 PM
Will this work in an Apple handheld?

sethypoo
Dec 11, 2003, 01:15 PM
Faster G3 chips! Altivec enhanced! Yay!

ebeitzuri
Dec 11, 2003, 01:20 PM
My guess is that this is the new consumer level chip. iMacs and iBooks will likely be getting this baby. We may even see it in Powerbooks if the G5 has issues.

rabatjoie
Dec 11, 2003, 01:20 PM
Originally posted by MattG
What's the clockrate of these things going to be? Could this be the final nail in Motorola's coffin?

The Appleinsider article says:

Clock-speeds should initially reach 1.8GHz mark before greatly surpassing the 2GHz milestone with its successor: the PowerPC 750VXe

that's very nice, haha. I might be going back to an ibook then.

CMillerERAU
Dec 11, 2003, 01:21 PM
this would be perfect for an ultra low cost consumer desktop. IMHO, a headless cube selling for $499 would really turn a lot of people on to Macs that would normally not want to give up their existing monitor, or are just really cheap minded. ;-)

gwuMACaddict
Dec 11, 2003, 01:51 PM
Originally posted by ennerseed
Will this work in an Apple handheld?

isn't it a little speedy for a handheld? i dunno, just asking...

greenstork
Dec 11, 2003, 01:53 PM
Man, it bums me out that Phil of Mac isn't around anymore, I need some vindication (http://forums.macrumors.com/showthread.php?s=&postid=500420#post500420)

All said, great news. The 750 vx will be a great mobile and low end chip. Bye bye Moto, good riddance.

willmg
Dec 11, 2003, 01:56 PM
So what are we hoping for then, Powerbooks will be due for an update in say 3-5 months, though most say G5 cant be in one till summer at the earliest. So maybe another G4 moto round on the PBs, then summer would bring either the G5 incarnation in PB and 750Vx in iBook or if there is some problem with G5 integration into PB the 750Vx could be swapped into them and the ibooks would have to suffer through another round without much change or use slower chips, memory, and smaller graphics? This doesn't sound to appetixing I dunno about you all I was hoping for a kick ass book to be announced by spring but seems like we could be in for serious length of time with moto still. :(

yoman
Dec 11, 2003, 01:57 PM
thats 3 times the speed of my current G3 ibook. :)

MacsRgr8
Dec 11, 2003, 02:01 PM
Originally posted by jbrown
but where will it be put???:rolleyes: :rolleyes:

Hopefully in the next gen. iBooks.
Just keep 'em as low cost as possible.

This is gr8 news!

greenstork
Dec 11, 2003, 02:05 PM
Without knowing that specs of these chips, I'll state for the record that I would MUCH rather have one of these 750VX chips in my PB than a G5, assuming power consumption is considerably lower on these chips.

The two most important priorities in a laptop (for me at least) are battery life (and somewhat related, heat) and speed. I think you have to strike a balance between the two. I own a PB because I like the extra power over an iBook and the difference in battery time between the two is negligible IMO.

That said, I am thinking (hoping) that this 750 VX will be more efficient than a G5. If it is considerably more efficient, it will be an excellent mobile processor that can scale up in speed along side the G5 while still maintaining long battery life and low heat. Ideal for a mobile processor.

edit: Let me be the first to say that the Powerbook will see one of these before a G5

CmdrLaForge
Dec 11, 2003, 02:08 PM
Q3 2004 ? What the hell ? If the chip is ready to go !

Why can't we see this chip earlier in a iBook at 2GHz ? Maybe because the G5 Powerbook is not ready earlier ! And this chip rocks !

pjkelnhofer
Dec 11, 2003, 02:09 PM
To me, it makes more sense to put a true G5 into the iMac and the Powerbook. Assume all heat and power issues can be worked out and save the 750VX for eMac and iBooks
The Powerbook and iBook are very close right now and this would seperate them again.
I don't understand why people think of the iMac as a some sort of Low-End consumer machine. $2200 is not a "consumer" PC even if it comes with at 20" LCD monitor built in. The iMac is do for a complete overhaul again soon.
The iMac could be updated early next year with a G5, and the eMac G4 speeded up. Then later in the year same thing with Powerbooks going G5 and iBooks getting the fastest G4's.
Then late next year (Christmas time anyone) low-end eMac's (which I think will go LCD when the iMac gets a new form factor) and iBooks all featuring IBM's *new* G5x (or whatever Apple calls it to drop the G4 moniker completely).

paulc
Dec 11, 2003, 02:15 PM
Uh, a "G4" IS a G3 with AltiVec. So this is essentially an IBM G4 chip.

As for G5s in laptops, let's not forget that as they shrink the die sizes/fab process, it will run cooler clock for clock. The 90nm chips shipping at some point will (should) run a lot cooler than the current 130nm ones.

I completely agree, battery life/run time is of paramount importance in a laptop rather than some numbers-speed issue.

BTW, "low cost" or even "ultra low cost" are phrases that one just can not mention in any discussion about our favorite fruit themed company.

XnavxeMiyyep
Dec 11, 2003, 02:17 PM
It will most likely be dubbed as a G4, as it wouldn't be as powerful as a G5, and Apple wouldn't want to reintroduce the name G3.

eric67
Dec 11, 2003, 02:19 PM
I am surprized that nobody notices what this release, in case it is true, means.....
if IBM is developping an altivec-enable G3, it implies :
- either Moto gave green light to IBM to do it because they want to stop G4 production
- either Apple really do not want to have anything to do with Moto anymore, and they might use this processor for replacing Moto G4, maybe before starting some legal action against Moto

(once they do not depend on Moto anymore)

But anyway it is cool to see IBM working on developping new processor for Apple

visor
Dec 11, 2003, 02:29 PM
i think we have the g4 already. doesn't use much energy, 32bit processing... why would we go back to the g3?
All it could be is the current g3 with altivec. I'd not expect to many structural redesigns of the g3 processor, since if it was a completely new proc, it woudn' feature the same serial number.

2ct

MacsRgr8
Dec 11, 2003, 02:41 PM
Originally posted by paulc
Uh, a "G4" IS a G3 with AltiVec. So this is essentially an IBM G4 chip.


Is this Mojave G3 multi processor compatible?
I mean, will I be able to order a Dual 2.0 GHz Mojave G3 (call it a G4)?
Then maybe Apple can use them in future low-cost Xserves aswell.

singletrack
Dec 11, 2003, 02:47 PM
IBM don't need anyone's permission to add Altivec to the 750. As part of the AIM alliance they have the right to implement it.

The other difference between G3 and G4 is the bus and multi-processor support, not just Altivec. The latter two may not be important though if they are only for single processor machines and aren't to be pin compatible with Moto G4 processors or use their bridge chips.

It's not quite a G4 but it sure isn't a G3.

pjkelnhofer
Dec 11, 2003, 02:56 PM
Originally posted by paulc
Uh, a "G4" IS a G3 with AltiVec. So this is essentially an IBM G4 chip.

That my understanding as well. As far as naming conventions go, how many different chips has Apple called G4? I remember when the G4 iBooks first came out no one was sure what model chip was really inside it.

Do people think they will call this chip a G4 as well? And, are we sure that these chips are even going to be in Mac's? IBM does sell chips to companies other than Apple.

greenstork
Dec 11, 2003, 03:01 PM
Originally posted by visor
i think we have the g4 already. doesn't use much energy, 32bit processing... why would we go back to the g3?
All it could be is the current g3 with altivec. I'd not expect to many structural redesigns of the g3 processor, since if it was a completely new proc, it woudn' feature the same serial number.

2ct

Well, the G4 has long been criticized for the fact that it is crippled by it's FSB, which stands to increase to 400 MHz with this 750 VX chip. This is also the reason that DDR memory is virtually useless on a G4. Unfortunately, the memory is far faster than the FSB which pretty much negates the fast memory.

In essence, this isn't reverting to a G3. It's IBM producing a G4, which kicks the ass of a current Moto G4. The ONLY beef on the G3 was the fact that it didn't have Altivec. Pound for pound, it is a far better chip than the Moto G4. It runs cooler, uses less power, and as evidenced by the projections for the 750 VX, runs far faster than any G4 (scales to 1.8 GHZ).

As for 32 bit processing, that's all the G5 is doing currently (in practice). The only 64 bitness of the current G5's is memory capacity since both Panther, and all applications, run in 32 bit. Now I know this will change, but it merits attention that the current G5 is only running 32 bit stuff.

dongmin
Dec 11, 2003, 03:19 PM
Originally posted by greenstork
Man, it bums me out that Phil of Mac isn't around anymore, I need some vindication (http://forums.macrumors.com/showthread.php?s=&postid=500420#post500420)

All said, great news. The 750 vx will be a great mobile and low end chip. Bye bye Moto, good riddance. well it's just a rumor. it's hardly anything solid. we'll see in 9 months time whether this 750vx is for real or not.

ThomasJefferson
Dec 11, 2003, 03:49 PM
I just decided to wait for the next ibook revision before buying a new one to replace my ibook600.

With a Q3 timeframe. I bet we see this in a Sept. 04 release. Making this 2 updates away.

I fear the next ibook revision will be a only a speed bump of the current chip. Not what I was really hoping for ....

k2k koos
Dec 11, 2003, 03:51 PM
Altivec or not, isn't that just a brand name of Moto?
IBM is mor ethan capable to produce avector processing unit that is compatible with"Altivec" instruction sets, and I don't think there are any legal problems either, afterall were they not all together in developing the PowerPC concepts in the first place?

may be someone else can shed some light on these "politics"?

Coem on with this procesor allready, these mobile intel chips are getting hot on the heals of our beloved powerbooks, we need to keep the edge here Apple!

CrackedButter
Dec 11, 2003, 03:51 PM
If this isn't being released till much later in 2004 you have to wonder what Apple is going to do with the current processors?

I havn't heard anything from moto as to what they are doing next with the G4 and it seems apple would be willing to put all processors from IBM into their machines.

If thats the case why the wait?

Get them in the machines NOW!

400mhz DDR bus might sound good now but what use is this in a years time? What will Intel put out in 12 months?

We all know for a fact laptops will be nearing 4Ghz on the windows side and PowerBooks will be on 2? Woopie doo!

<sarcasim>Wow! I cannot wait.</sarcasim>

Photorun
Dec 11, 2003, 03:52 PM
Originally posted by MattG
What's the clockrate of these things going to be? Could this be the final nail in Motorola's coffin?

One can only hope!

eric67
Dec 11, 2003, 03:57 PM
Originally posted by k2k koos
Altivec or not, isn't that just a brand name of Moto?
IBM is mor ethan capable to produce avector processing unit that is compatible with"Altivec" instruction sets, and I don't think there are any legal problems either, afterall were they not all together in developing the PowerPC concepts in the first place?

may be someone else can shed some light on these "politics"?

Coem on with this procesor allready, these mobile intel chips are getting hot on the heals of our beloved powerbooks, we need to keep the edge here Apple!

the G5 has an altivec-like vector-processing unit... let's call it "velocity engine" as Apple like to call it...

eric67
Dec 11, 2003, 04:02 PM
Originally posted by CrackedButter
If this isn't being released till much later in 2004 you have to wonder what Apple is going to do with the current processors?

I havn't heard anything from moto as to what they are doing next with the G4 and it seems apple would be willing to put all processors from IBM into their machines.

If thats the case why the wait?

Get them in the machines NOW!

400mhz DDR bus might sound good now but what use is this in a years time? What will Intel put out in 12 months?

We all know for a fact laptops will be nearing 4Ghz on the windows side and PowerBooks will be on 2? Woopie doo!

<sarcasim>Wow! I cannot wait.</sarcasim>

in low end PC labtop you will find 3-4GHz P4 by end of 2004, yes, but it is crap, even PC users recognize it. the real compettor for Apple is the Pentium M, which is at the moment at 1.6GHz....
a 4GHz P4 in a labtop does run at 4GHz...it is a fact, check on PC web sites, it is just a argument for marketing... for no brain users.
Just look at the price of a labtop with a 3GHz P4 and a labtop with a 1.4 or 1.6 GHz Pentium M, you will see the latter behind the most expensive...

Rincewind42
Dec 11, 2003, 04:03 PM
IBM & Motorola jointly worked on the instruction set known most popularly as Altivec. Altivec however is a Motorola trademark, which is why Apple doesn't uses it (and uses Velocity Engine instead). The original name was VMX. There are politics to IBM calling it's SIMD engine Altivec because that is a Motorola trademark. There are no politics to IBM having a VMX SIMD engine, because they have rights to such a thing. However until now they didn't see a big reason/market in making chips with VMX and competing with Motorola. With Motorola's chip foundry falling apart and being spun off, there is reason for IBM to produce such chips, if only because Apple promises to foot part of the research bill.

CrackedButter
Dec 11, 2003, 04:06 PM
Maybe, but people like i have just done will compare numbers...not everyone will have an understanding of how a processor is rated and how it runs.

Gymnut
Dec 11, 2003, 04:15 PM
Moto-who???

phreaker57x
Dec 11, 2003, 04:22 PM
Originally posted by XnavxeMiyyep
It will most likely be dubbed as a G4, as it wouldn't be as powerful as a G5, and Apple wouldn't want to reintroduce the name G3.

Maybe the new chip should be called the G4 Extreme, since Apple seems to like that word (Airport Extreme, Quartz Extreme, etc.). It would sound faster than the current G4s but won't steal the G5's speed connotations.

macphoria
Dec 11, 2003, 04:49 PM
This chip could definitely end up in iBook.

greenstork
Dec 11, 2003, 04:57 PM
Originally posted by CrackedButter

We all know for a fact laptops will be nearing 4Ghz on the windows side and PowerBooks will be on 2? Woopie doo!


The fastest Intel laptop chip is at 1.7 GHZ (at least it was afew weeks ago, haven't checked if they speed bumped lately).

The Intel Pentium M is a far superior mobile chip than the P4 you are referencing.

edit: Eric beat me to it.

ITR 81
Dec 11, 2003, 05:07 PM
If this ends up in the iBook then we could look at something like this:
lowend 1.6
mid range 1.8
topend 2.0

or

lowend 1.5
mid range 1.6
topend 1.8

This alone makes me think the PB's #'s would be abit higher as most folks don't look at the processor they only normally look at it's clock speed.

Is the 750 VX 64 bit? I would hope so, so Apple can introduce it's 64 bit OS sooner.

The 750 GX currently at 1.1 from IBM.

greenstork
Dec 11, 2003, 05:10 PM
Originally posted by macphoria
This chip could definitely end up in iBook.

Assuming Apple uses this chip, this statement is a bit of a no brainer, of course it would end up in an iBook eventually, if not to begin with.

ITR 81
Dec 11, 2003, 05:13 PM
Originally posted by CrackedButter
If this isn't being released till much ?

We all know for a fact laptops will be nearing 4Ghz on the windows side and PowerBooks will be on 2? Woopie doo!


And all 4GHz machine will weigh a ton compared to a iBook or PowerBook not to mention you will need to have it plugged in almost always because the P4 eats batteries for breakfast.

Most folks I know are getting Pentium M's which run alot cooler and run alot longer on battery power.

The only use for P4 now is in gaming laptops but once a laptop goes over 5lbs I don't consider it a laptop anymore.

CrackedButter
Dec 11, 2003, 05:14 PM
Originally posted by greenstork
The fastest Intel laptop chip is at 1.7 GHZ (at least it was afew weeks ago, haven't checked if they speed bumped lately).

The Intel Pentium M is a far superior mobile chip than the P4 you are referencing.

edit: Eric beat me to it.

Don't get me wrong, i don't disagree but people will look at the numbers... battery life and true performance will take a back seat on the otherhand.

macphoria
Dec 11, 2003, 05:16 PM
Assuming Apple uses this chip, this statement is a bit of a no brainer, of course it would end up in an iBook eventually, if not to begin with.
What are your thoughts on third party hardware developers using this chip to make processor upgrades? I'd like to see a 2Ghz G4 processor upgrade card for my old G4 PowerMac. This chip looks like a great candidate for such use also.

ffakr
Dec 11, 2003, 05:45 PM
Originally posted by paulc
Uh, a "G4" IS a G3 with AltiVec. So this is essentially an IBM G4 chip.
That's a miss-conception... just as the belief that the G3 is a next-gen 603 and the G4 is a next gen 604 is a miss-conception. I think the truth is a little more in between these two and yet not all there.

They are similar (G3s and G4s) because they are evolutions of the PPC development, but the G4 isn't just a G3 with altivec. They had different pipeline lengths, the G4 had better FP performance... Since then, the designs have each continued to evolve. There were changes to FP and altivec in the G4 when it was reved to the 745x series.

It isn't fair to say G3 + Altivec = G4 so this new G3 is just a G4.

How about we go with 'this new G3 is somewhat congruent to a G4!!'

:-)

swdrumcp
Dec 11, 2003, 05:51 PM
Well battery life is important. I have an IBM with a 1.6gh p4M. The computer is as powerful as the new pb, and as damn expensive. But the battery life sucks, its terrible. It doesnt last more than an hour and 15 minutes at the most. And a extra battery is over $200. (and I didnt buy this, i got a grant for it, lol). But this fact alone makes it almost useless for anything more than to sit on a desk.

Im still waiting on a full featured 12' pb. Backlite keyboard, 64mb graphics, same processor, etc. Apple make one and it would be my next computer!

Gyroscope
Dec 11, 2003, 05:51 PM
This could definitely be a new iPod CPU :D :D :D :D

ffakr
Dec 11, 2003, 05:53 PM
Originally posted by CrackedButter
We all know for a fact laptops will be nearing 4Ghz on the windows side and PowerBooks will be on 2? Woopie doo!

<sarcasim>Wow! I cannot wait.</sarcasim>

Don't wait... leave the Mac and buy one. You might get lucky and burn your bits off (http://theregister.com/content/archive/28245.html) in the process.

Intel has publicly stated that it will not try to push performance in the mobile space by pushing clock speed. The Prescott will produce over 100 watts at around 3GHz. You want a variation of that, only faster in a laptop? good luck.

The Pentium-M is the first post P4 attempt by Intel to make good on its promise to make more efficient processors.
It is entirely possible that you WON'T see PC mobile processors increase much in clock speed for a while.

<sarcasm>
Insiteful post though..
</sarcasm>

sorry, had to borrow from you a bit ;-)

stoid
Dec 11, 2003, 05:56 PM
Originally posted by ITR 81
The only use for P4 now is in gaming laptops but once a laptop goes over 5lbs I don't consider it a laptop anymore.

Dang, I sure thought my brand-new 15 inch AluBook was a laptop. Shucks, neither is the 17 incher nor the 14 inch iBook. That sucks.

I would consider a laptop that resembles a college textbook more than a notebook to be not a laptop. My friend with his 1.7 inch thick Windows POS registering 447MB of RAM? said that he likes that his computer is big and heavy and loud because that way he knows 'that there are parts in there'

Windows users sure seem to have a skewed view of the world...

ffakr
Dec 11, 2003, 05:57 PM
I've worked on Pentium-M machines and they do crank (at a relatively low clock speed) when you get them going. The problem is that it employs a clock slewing like the G5 but it takes too long to 'wake up' from periods of inactivity. It seemed to me (on a few systems) that the machine would choke for a while, figure out there was actually a heavy load, then speed up. When it got going (used Seti-at-home to check one out) it did get a lot of work done.

For typical work (idle, busy, idle cpu loads) it's kind of annoying.

mspock
Dec 11, 2003, 06:20 PM
I heard that the mojave chip support a 1000 k of L2 cache is that true ?

If yes, i will prefer a 1.8 ghz (1000k L2 cache) 400 FSB in a powerbook with great battery life, than a G5 powerbook with less battery life.

don't you ?

panphage
Dec 11, 2003, 07:38 PM
RE: Nail in Moto's Coffin:

Motorola no longer produces microprocessors. You can't put any more nails in Moto's micro coffin, Moto already did it itself. They are out of the business. Moto's micro division is now a totally seperate company. At least that's the last I heard. Never did get a name for this new company though.

greenstork
Dec 11, 2003, 07:44 PM
Originally posted by mspock
I heard that the mojave chip support a 1000 k of L2 cache is that true ?

If yes, i will prefer a 1.8 ghz (1000k L2 cache) 400 FSB in a powerbook with great battery life, than a G5 powerbook with less battery life.

don't you ?

It will support 4 MB of L3 cache, and as much L2 as the 750FX (which I regretfully don't know how much that is), according to AppleInsider.

bousozoku
Dec 11, 2003, 07:54 PM
Originally posted by ffakr
That's a miss-conception... just as the belief that the G3 is a next-gen 603 and the G4 is a next gen 604 is a miss-conception. I think the truth is a little more in between these two and yet not all there.
...

How about we go with 'this new G3 is somewhat congruent to a G4!!'

:-)

It's a terrible thing to even joke about the G4 being related to the 604. You certainly can't say that the G4 is a G3 with AltiVec. It's not. It's an AltiVec unit with bits and pieces of the G3 and 604e with too much glue keeping the speed down.

Whatever IBM does, it will be more than a match for any Motorola G4.

MrMacMan
Dec 11, 2003, 08:13 PM
Originally posted by ennerseed
Will this work in an Apple handheld?

Don't you mean its way, way to powerful for a handheld.

Originally posted by k2k koos
Altivec or not, isn't that just a brand name of Moto?
IBM is mor ethan capable to produce avector processing unit that is compatible with"Altivec" instruction sets, and I don't think there are any legal problems either, afterall were they not all together in developing the PowerPC concepts in the first place?

may be someone else can shed some light on these "politics"?


IBM can Legally use the works 'SIMD-Enabled' which means aka 'Altivec'...
Apple uses the works 'Velocity engine'...
Moto calls it 'Altivec'
IBM can call it 'SIMD'... all the same stuff

So basically they devolped it with them, they can use then name without fear of lawsuit.

Look, even if this chip is labled a 'G3' ut would still smoke my mac anyday.

Archmage
Dec 11, 2003, 08:55 PM
You all seem to think this is going into iBooks 2 or so revisions down the road.

Why not put the 750VX into the PowerBooks and the 7457's into the iBooks?

Then, after the G5 will fit in a Powerbook, put the 750VX into the iBook :)

wizard
Dec 11, 2003, 09:28 PM
Well I sure would!!

I would imagine that such chips would run 32 bit code about as fast as the G% anyways. They may be a bit slower for bandwidth hungery AltVec applications but who really cares, overall the performance increase will be rather significant. Mary a significant pefromance boost with industry leading power usage and you will have it made.

Thanks
Dave


Originally posted by mspock
I heard that the mojave chip support a 1000 k of L2 cache is that true ?

If yes, i will prefer a 1.8 ghz (1000k L2 cache) 400 FSB in a powerbook with great battery life, than a G5 powerbook with less battery life.

don't you ?

wizard
Dec 11, 2003, 09:38 PM
I don't think we will ever see a 970 in a PowerBook. I could be completely wrong here but I think Apple and IBM will look to a follow on to the 970 for inclusion in Laptops targetted at the 64 bit market.

Now what Apple and IBM may come up with is certainly in the air, but if they want to support 64 bits in the portable sector, it will have to be a low power solution. The other solution is to compete in the tranportable market with a LugaBook. A LugaBook could serve a different market that hte PowerBooks, the question is is that market large enough for Apple to attack?

I think the problem with the 7457 is pretty clear -- Motorola can not scale the chip to run at an acceptable speed!!!. The iBooks are already a lack a bit of competitveness as the curent G4's are to power hungery. From my pwerspective the market that the iBooks compete in is the ultra low power usage / long battery life domain. If they can not hold on to that niche there is little reason for them to exist.

Dave


Originally posted by Archmage
You all seem to think this is going into iBooks 2 or so revisions down the road.

Why not put the 750VX into the PowerBooks and the 7457's into the iBooks?

Then, after the G5 will fit in a Powerbook, put the 750VX into the iBook :)

joelypolly
Dec 11, 2003, 09:53 PM
The Pentium M is still the best mobile processor around. Period. Battery life can be over 5 hours of real use not like when Apple posts 5 hours... Still Mac portables are still playing catch up to PCs interms of perfomance.

TomSmithMacEd
Dec 11, 2003, 11:00 PM
Originally posted by pjkelnhofer
To me, it makes more sense to put a true G5 into the iMac and the Powerbook. Assume all heat and power issues can be worked out and save the 750VX for eMac and iBooks
The Powerbook and iBook are very close right now and this would seperate them again.
I don't understand why people think of the iMac as a some sort of Low-End consumer machine. $2200 is not a "consumer" PC even if it comes with at 20" LCD monitor built in. The iMac is do for a complete overhaul again soon.
The iMac could be updated early next year with a G5, and the eMac G4 speeded up. Then later in the year same thing with Powerbooks going G5 and iBooks getting the fastest G4's.
Then late next year (Christmas time anyone) low-end eMac's (which I think will go LCD when the iMac gets a new form factor) and iBooks all featuring IBM's *new* G5x (or whatever Apple calls it to drop the G4 moniker completely).

It would be sweet to see these. My comment is just on the iMac though... Who would want to buy one? They aren't fast, dont' do a whole lot... The only thing nice is the monitor. I don't know just my thoughts.

dguisinger
Dec 11, 2003, 11:05 PM
Originally posted by ffakr
That's a miss-conception... just as the belief that the G3 is a next-gen 603 and the G4 is a next gen 604 is a miss-conception. I think the truth is a little more in between these two and yet not all there.

They are similar (G3s and G4s) because they are evolutions of the PPC development, but the G4 isn't just a G3 with altivec. They had different pipeline lengths, the G4 had better FP performance... Since then, the designs have each continued to evolve. There were changes to FP and altivec in the G4 when it was reved to the 745x series.

It isn't fair to say G3 + Altivec = G4 so this new G3 is just a G4.

How about we go with 'this new G3 is somewhat congruent to a G4!!'

:-)

I disagree. There are several different G4s, each with different pipelines, optimizations, etc. But the instruction set stays the same, the core design stays the same. The G4 is a G3 with Altivec. You are trying to make the G4 more special than it really is. Sure, they may have increased the pipeline. Intel does that between P4 revisions....but its still the same processor at heart....just some operations are cut down into two for better clock scaling.... nothing special about that, its quite easy actually.

There is very little difference between the G3 and G4. Hell, even a bus change doesnt qualify as a CPU change.....because one CPU can support multiple BUS designs! It doesnt happen often, but it does occur from time to time.

tychay
Dec 12, 2003, 12:38 AM
Originally posted by swdrumcp
Well battery life is important. I have an IBM with a 1.6gh p4M. The computer is as powerful as the new pb, and as damn expensive. But the battery life sucks, its terrible.

A P4M is a crappy chip which is built for the ignorant masses who think clock speed is everything. BTW, it must be an old one because P4Ms are >2Ghz now. The one to benchmark against is the Pentium M, not the Pentium 4M, which is closer in design to a Pentium 3 than a Pentium 4. They run to about 1.7Ghz @ 130nm and are pretty efficient chips. In many benchmarks they outperform Mac G4 laptops (though it's hard for me to say that Apple is playing "catchup" here as others have claimed because as anyone who has used laptops for eight years will tell you, there's a lot more to a laptop than the speed/efficiency of the CPU).

All Intel "Centrino"s are spec'd with a Pentium M. In fact, much of the reason of the "Centrino" moniker is to divert attention of the consumer away from the Mhz of the chip. Intel depends on people looking at the Mhz for the desktop line, but they don't want them doing so for notebooks.

Rumor has it that the sequel to the Pentium M @ 90nm is delayed because the Pentium M is doing so well. That may be possible (they were at least due for a speed bump a few months ago and the PM is selling like hotcakes and commands a high price when put in "Centrino" notebooks), but it may be just as likely that Intel's production 12"/90nm fab (Fab11X) is having trouble producing anything at 90nm.

I remember I hinted as much a few months ago and I should note that the year is almost at a close and we have yet to see a single production 90nm chip coming from Intel--no Pentium 5, no new Pentium M.

BTW, for the record. Altivec is a Motorola term. IBM calls it "VMX" when they refer to it, though sometimes in their literature they use "Altivec-compatible" with a note that the term Altivec is a registered trademark of Motrola. Apple has called the same thing "Velocity Engine" since they launched the G4, which is really some impressive foresight when you think about it.

ffakr
Dec 12, 2003, 01:20 AM
Originally posted by dguisinger
I disagree. There are several different G4s, each with different pipelines, optimizations, etc. But the instruction set stays the same, the core design stays the same.
every PPC has the same instruction set (Altivec being the glaring exception). That's what being a PPC is all about.
That doesn't, in any way, mean that all PPCs are the same. A 603 and a 604e had the same instruction set, but they certainly weren't the same processors.

My un-scientific evidence that the G4 isn't a G3 w/ Altivec is pretty straightforward.. The G4 had a different floating point unit. We know this because the G4 had better floating point performance at the same clock speed.

Sure the G4 has changed a lot during it's life. Should these changes have caused a new series designation? I don't think so, maybe it could be argued though. They did come over several revisions though and they were built off the previous generations.
This strengthens the argument that the 750vx is NOT simply a G4 though. Even if we assume that the G3 and G4 shared the same core, they both evolved in different ways. The G4 has had changes to interger units, floating point units, load/store, big changes to altivec. The G3 has gone through changes of it's own. If anything, the two chips have diverged over the years. If they have moved apart over several reversions, you can't make one into the other by simply tacking on Altivec.

I'll see if I can find some documentation regarding whether or not the G4 and G3 shared core units at the time of the G4 launch.

ffakr.

ffakr
Dec 12, 2003, 01:29 AM
Originally posted by joelypolly
The Pentium M is still the best mobile processor around. Period. Battery life can be over 5 hours of real use not like when Apple posts 5 hours... Still Mac portables are still playing catch up to PCs interms of perfomance.
PentiumM is a great mobile processor, but those machines that boast really long battery life also have VERY large batteries. On vendor (IBM I think) has a P-M laptop that is rated at something like 9 hours, but to get that you need an oversized battery that extends the physical size of the laptop by an extra inch.

Apple's Laptops do extremely well with fairly average Li batteries. In fact, the 17" powerbook has a larger battery and I think it's rated at 6 hours (that might only be one speed revision though.. not sure).

My real complaint with the P-M are the tricks it need to do in order to provide such powersaving. It took me a while to figure out why the first P-M I worked on was only running at 500MHz. It scaled all the way back from 1.3GHz to 500MHz when not under load. Not a bad design (transmeta and the G5s do the same) but the system I worked on took too long to scale up when the load appeared. I found the lag to be noticable and annoying... like it was hanging everytime the machine got busy.

I'd like to see what the functional battery life of these P-M notebooks would be if they were forced to remain busy and clocked up the whole time.

I think Apple is still competing quite well in the laptop market. I'm perfectly happy with this 600MHz ibook (thanks Work) and the newer Powerbooks that I've worked on have been excellent computers. If I need the horsepower to play games, I put the ibook down and I boot up the Athlon tower. ;-)

ffakr.

jade
Dec 12, 2003, 01:31 AM
Originally posted by CrackedButter

We all know for a fact laptops will be nearing 4Ghz on the windows side and PowerBooks will be on 2? Woopie doo!

<sarcasim>Wow! I cannot wait.</sarcasim>


The only intel chip the powerbooks need to be concerned with is the Centrino chip. Many PC laptops are using desktop chips. If the next version of the powerbooks could have pentium 4-type (well g5) performance with centrino like battery life we will have a winner. Centrino is topping out at 1.7 now and is anticipated to reach 2.5 or so next year.

warcraftmaster
Dec 12, 2003, 01:36 AM
Originally posted by ITR 81
If this ends up in the iBook then we could look at something like this:
lowend 1.6
mid range 1.8
topend 2.0

or

lowend 1.5
mid range 1.6
topend 1.8

This alone makes me think the PB's #'s would be abit higher as most folks don't look at the processor they only normally look at it's clock speed.

Is the 750 VX 64 bit? I would hope so, so Apple can introduce it's 64 bit OS sooner.

The 750 GX currently at 1.1 from IBM.




ya they are at 1.1 ghz but it run's at 9w at 1.1 or top out (i think) well any way i some sad the there ibook g4 top out at 40 w (ya i said 40w) now you see the power diff he he. i dont think they will go 64 bits

Plutoniq
Dec 12, 2003, 02:45 AM
Wonder If the 750VX will be pin & voltage compatible with the 750GX & FX? Powerlogix could very easily continue there 750XX upgrade lines (Pismo, B&W etc.) with the VX if this the cease.

Reason why the 750GX (and hopefully 750VX) rule for upgrading older macs, the 1mb of full speed L2 Cache really helps with the slow frontside bus of these older Mac's. Lot less chance of having to read/write time critical instructions to RAM on a 66/100mhz bus if you've got 1mb of L2 cache running at 1.1gz+. For emulation, like VirtualPC or Sixtyforce (N64 Emulator), L2 cache size and speed is critical for smooth performance.

Viva la Pismo btw

p

IVIIVI4ck3y27
Dec 12, 2003, 06:28 AM
Originally posted by k2k koos
Altivec or not, isn't that just a brand name of Moto?
IBM is mor ethan capable to produce avector processing unit that is compatible with"Altivec" instruction sets, and I don't think there are any legal problems either, afterall were they not all together in developing the PowerPC concepts in the first place?

IBM actually developed the architecture and owns the rights to the name "PowerPC". It's an IBM registerred trademark that Apple and Motorola both have licensed.

The aforementioned "AIM" agreement though, has been disbanded. IBM and Motorola split their roadmap after the G3. The oddball thing is, I always thought it was going in the wrong direction. Of course... if Motorola had been able to keep their Mac clone division, it'd probably be a case where Motorola would be amped up about the Macintosh. Yet... the minute Jobs killed off cloning, I think that took the desire and will out of Motorola to focus on desktop development. In a lot of ways, if there was anything Apple could've done to be "choosy" with their clone vendors... they migth've been smart to retain Motorola and Motorola only. Although... that could've brought on some potential lawsuits...

The G3 name is a joint owned name, owned by Motorola and IBM and is applied to 3rd Generation of PowerPC's. The first generation were the 601/602(Bandai Pippin@world)/603/604/615 (64-bit processor, rumored to house an IBM BlueLightning Pentium-compatible core)/620 et al.

Second Generation was the 603e/604e series.

A merger of the two design roadmaps yielded a much faster processor (akin to the 604e) with the affordability of production and better efficiencies of the 603e. This was... the G3. It shared very little with it's previous designs, but it did evolve from both spectrums to get where the G3 was at. Fast... and very efficient.

The G4 is a revamped G3 design (meaning it did change beyond just slapping on AltiVec, don't read that as to mean it's anything but what I've said, the FPU, processor bus, and other technologies were alterred but it still was highly based on the G3 version of PowerPC architecture and design) employing AltiVec (a SIMD vector architecture, designed by Motorola), called Velocity Engine by Apple. The name G4 and it's associated rights thus far is owned by Motorola as IBM at the time had no interests in pursuing SIMD technology as IBM's application for PowerPC was not hinged around speciality processors, as they use it under very portable outlets; i.e. Linux and formerly A/IX. Neither of which are geared towards accessing specialty processors, with no major Linux distro focused on any SIMD technologies. Obviously... that has changed in IBM's case now...

Oh and for the record... IBM's PowerPC 970 *DOES NOT* use AltiVec. IBM's technology is a SIMD vector unit that is compatible, though not identical, with Motorola's AltiVec technology. Think of it as 3DNow vs. MMX on the PC side, although in this case... close enough that if you write for one, it works effectively for both. Sort of how an AMD vs. Cyrix vs. Intel processor was once compatible back in the 486sx/dx days.

So I expect the new processor to carry the "G4" name. I'm sure Motorola is at a stage where they'd rather focus on the embedded sector (more profitable and more in their key market anyhow), and with as often as the Somerset manufacturing plant (formerly owned by IBM and Motorola, now solely Moto) has been raided and plundered of engineers by it's competition, Motorola might be gearing very much towards dumping AltiVec-based processor production in favor of further evolutions of embedded chips; either PowerPC G3-evolutions... or more towards Digital DNA.

IBM does a lot of server-level hardware... to me the marriage between IBM and Apple always seemed a strongly beneficial one. Motorola seemed very much like their goal was to become a major player in the computer market riding on the coattails of someone. I think they had plans to make the Mac a veritable breeding ground of Motorola machines. The problem was... Apple socked them in the eye when they tried to beat them to market, and a lot of that was due to the slow-moving nature of Apple at that time. Morale was low as Apple was just recently taken over by Jobs... and it took him a bit to right the ship and set it's course.

The fact.

If Jobs didn't kill off the clones, there'd not be an Apple today... yet in doing so, he royally pissed some people off. Not that the clones were as good a machines reliability-wise. I firmly believe Radius/Umax were the best of the litter... and Umax was quirky at best, Radius machines used Apple mainboards and were geared towards video editing out of the box. PowerComputing were fast machines, but their reliability quotient and the required extra drivers to make them work (many machines require CD Toolkit for drivers for the CD-ROM's, a hassle) was a pain. The early StarMax Motorola machines weren't much better/worse than the PowerComputing machines. Daystar were obscure and high-end but not necessarily better in many ways than their lesser processored competition. The rest of the bunch was akin to getting an Acer or Packard Bell, except these were APSTech and MacTell machines.

Granted driver support would've likely changed if Amelio could've retained Apple, kept Apple afloat (suspect), boosted morale (laughable), and made Rhapsody a reality as it was going to be geared more and more to be like Windows and support a wider range of drivers (and probably become verrrrry buggy in trying to do so; it was after all planned to be designed for an open platform that was going to open Pandora's box... basically become the Mac version of the PC industry and spread itself verrrrry thin). Yet... with Jobs at the helm, cloning didn't stand much chance.


may be someone else can shed some light on these "politics"?


It's a good question, you can't really tell. My ex-gf's mom used to work for Motorola. Soon after Apple (under Jobs) jammed a broomhandle in the spokes of Motorola's CHRP/PPCP-based G3 Macintosh, the first G3 to debut on the market but never ship... Motorola has been "VERY" bitter about the whole deal. As far as fire-sale ditching of any machine with a multi-hued Apple logo on it. Hell she was allowed to bring many "old" Macs home without payment just to get rid of them since Motorola was moving their offices over to PC... In other words, Motorola dumped the Mac das pronto. It bordered almost on an irrational move as some of the machines that were dumped were fairly new at the time.

They also began actively attempting to pursue Microsoft for Windows CE/PocketPC compatibility at that time. I don't believe much ever came of that, even though some of the embedded PowerPC's are just as good as the very popular ARM/StrongARM and some of the smaller Intel-based PDA processors.

I don't particularly see Motorola sticking around, and I don't think they have the want to fight anymore. They see the writing on the wall, they killed their own G5 plans and roadmap long before the 970 was announced even, and I think they did so to tell Apple "Hey, find someone else". I don't think we'll see a fight out of them, and the faster they can avoid having to spend tons of $ on R&D for a processor lineup that probably barely breaks even for Motorola's intents and purposes... the G4 is probably something they long to axe at first opportunity. My belief... the faster Apple can move everything to IBM, the happier Motorola will be. Seems weird because you'd think they'd want that extra revenues... but desktop and laptop processors aren't particularly in the niche that Motorola I feel most wants to gear towards *now*. They probably can make 10x's as much $ selling to embedded applications... like ECU technology for GM cars, perhaps cell phone technologies for their own cell phone lineup, and maybe work harder on technologies for usage in car stereos, personal stereos, etc. etc. Maybe even accessory chips for computers rather than the main CPU. All of which are based on a similar architecture, and one where the R&D is recouped 10x's over via larger volume sales.

IBM on the other hand... found a way to turn a source of revenues for it's server-class Power4 based PowerPC chipset... and by simply catering to Apple's need for SIMD, they now have a processor they can use in blades running Linux, and sell x # of processors to Apple for use with OS X in their PowerMacs in higher volumes than their own machines, and therefore there's a % of every new Mac going into IBM's pocket that was rather miniscule with G3 only (last machine running G3 was iBook and it was destined to change). The R&D costs were already there, and it's a source of revenue. Win/Win for both Apple and IBM... plus embedded focus (Motorola) vs. desktop/server application focus (IBM, Apple)... which is more fruitful for the parties? It's only logical that IBM which isn't as geared on embedded as Motorola favors Apple better.

That isn't to say that there weren't some concessions to be made in the development process... but with servers brute speed in all manners isn't always the key focus (reliability being more prevalent, especially in blades where you can always toss in more processors to do the dirtywork anyhow if you need more speed). IBM could've likely more focused attack Itanium and Hammer head-on and surpass them, but to retain compatibility with the Mac... they probably made some revisions that in the short term slowed them down... but in the long term seek to grow both sides.

IVIIVI4ck3y27
Dec 12, 2003, 06:29 AM
Coem on with this procesor allready, these mobile intel chips are getting hot on the heals of our beloved powerbooks, we need to keep the edge here Apple! [/B]

I am a firm believer that this isn't "stalling" (to believe so would be to believe that companies actually like to watch the competition beat them), as I'm sure Apple and IBM would love nothing more than to be selling you this right now (for the fact it'd allow them to compete with and potentially blow away the competition). You have a usual span of a year from announcement 'til shipping, barring any potential ramp-up. It's scheduled for Q3 but I'd not doubt it if it's not out by early-mid Q2 as for stock sales reasons... it's best to be a bit behind on your guesstimate than be ahead and fall flat on your face. Granted... some might argue that they need to get moving faster... but processors in laptops are usually some of the slowest to evolve because it's not just straight "power" but "efficiency", and especially in the lower end of the scale where you're trying to sell competitively at lower pricepoints and have success with smaller margins. To me this processor is Apple's "Celeron" or Apple's "Duron" to what Intel and AMD sell on the PC-side for the future, and the fact it just evolved itself from the G3 platform which is spritely and cheap and verrrrry efficient, is a positive. Toss in a bit more horsepower, fatter pipes, and more grunt out of the addition of SIMD... and it's not something to sneeze at in the least.

I feel the shrunken PPC 970 and 980 are more critical mass than anything else as Apple and IBM need to play catch-up in this sector with the desktop PIV, the Athlon XP, and the latest Athlon 64-bit processors, much less anything further down the road. That's before you get into the scenario where IBM would love to have these trouncing Itanium and Hammer-based designs. They're still comparable or on-par in laptops with the G4, which isn't necessarily a "lead" but it's a case where IBM is having to do a lot at once and get everything online to get competitive and then try to trounce the competition. It takes time to get to that level when you've not been focusing your R&D in that capacity as much. Hell if Apple hadn't come back around, I'm not sure if we'd see any further evolutions of the PowerPC, and might only be seeing Power4 and Power5 based server processors without SIMD/AltiVec and geared towards IBM's sole needs. That is if they didn't concede the PowerPC (i.e. kill it) and embrace AMD or Intel for servers.

I'm sure Motorola's updates to the G4 will work stop-gap until IBM can get this processor online. That isn't to say that when this processor launches it'll be faster than the Intel/AMD competition, as it could very well be slightly slower. I feel we'll probably be seeing this in the low-end consumer machines anyhow (eMac and iBook), with the 970/980 ending up in the Powerbook, iMac (consumer machine but pretty much considered the "high-end" of the non-professional segment), Pro desktops (duh), and XServe/XServe RAID. Once the processor is out for a generation, we probably will see IBM pump the #'s up and go toe to toe or surpass. Just takes time, like I said earlier, to come out of the box running when you've been sitting debating what you're going to do. If anyone can make the PowerPC competitive though... don't doubt on IBM. They have the resources and engineering to go toe to toe with anyone. Most of the latest technological patents have come straight from IBM. Silicon on Insulator. Copper Wiring. High-tech fabs like the G5 are built on. You name it.

g4pismo
Dec 12, 2003, 09:03 AM
Originally posted by Gyroscope
This could definitely be a new iPod CPU :D :D :D :D


.. lets be the first to cluster them.. VA Tech ain't got nothing on a 200 node iPod cluster :-P

IVIIVI4ck3y27
Dec 12, 2003, 09:25 AM
Originally posted by g4pismo
.. lets be the first to cluster them.. VA Tech ain't got nothing on a 200 node iPod cluster :-P

::user screams at the thought of 200 iPod's synced playing Barry Manilow at the same time::

neilw
Dec 12, 2003, 10:14 AM
The 750VX seems like the future for the iBook, for sure, but at the speeds being mentioned it is a plausible interim step for other lines as well.

While I would like to see the G5 propagated through as much of the Apple line as possible, if by summer Apple could get the 750VX into a PowerBook running up to 1.8 GHz, that'd be an OK update.

That'd also free the iBook to jump to the 750VX. Even kept slower than the PowerBook, it would still allow for a 50% speed jump from the current G4 models.

The iMac, on the other hand, should jump to the G5, unless they are going to seriously cut the price and really target it to consumers. At its current prices, having a sub state-of-the-art processor doesn't make much sense to me. But it wouldn't shock me if the 750VX finds its way there first.

rdowns
Dec 12, 2003, 11:22 AM
Originally posted by neilw
The 750VX seems like the future for the iBook, for sure, but at the speeds being mentioned it is a plausible interim step for other lines as well.

While I would like to see the G5 propagated through as much of the Apple line as possible, if by summer Apple could get the 750VX into a PowerBook running up to 1.8 GHz, that'd be an OK update.

That'd also free the iBook to jump to the 750VX. Even kept slower than the PowerBook, it would still allow for a 50% speed jump from the current G4 models.

The iMac, on the other hand, should jump to the G5, unless they are going to seriously cut the price and really target it to consumers. At its current prices, having a sub state-of-the-art processor doesn't make much sense to me. But it wouldn't shock me if the 750VX finds its way there first.

If Apple forgoes a G5 for the 750VX in an iMac, I fear they will have lost a 17 year user. As much as it pains me to say, I have a Dell P4 with Win 2000 at work and that puppy screams and is very reliable. I have crashed exactly once in about a year. Check that, twice (iTunes for Windows was the second).

Same old Apple, no clue how to market their products (except the original iMacs). Was the case when I sold Apple products from the beginning to the first Power PCs. As much as I like Jobs, he has his head up his ass when it comes to marketing.

AidenShaw
Dec 12, 2003, 11:40 AM
Originally posted by ffakr
every PPC has the same instruction set (Altivec being the glaring exception). That's what being a PPC is all about.


Another glaring exception is the PPC970's lack of little-endian support....

Rincewind42
Dec 12, 2003, 11:44 AM
Originally posted by ITR 81
Is the 750 VX 64 bit? I would hope so, so Apple can introduce it's 64 bit OS sooner.

Apple can work towards a fully 64-bit OS anytime they want. And they won't have to leave 32-bit only CPUs behind - they just won't be able to run 64-bit only OS components. The OS Core is 64-bit clean, and some parts of it do run as if they are in a 64-bit memory space (Virtual Memory for one). If you installed it on a 32-bit CPU you'd just not get the 64-bit parts and would run as a you do now. The biggest hurdles right now are the fact that many of the frameworks are not fully 64-bit clean, and that you would need to ship twice as much code (a 32-bit and 64-bit version of every currently 32-bit library). So while they could introduce a 64-bit OS that still supported 32-bit CPUs, it will not likely be soon.

Daaavek
Dec 12, 2003, 11:46 AM
Originally posted by macphoria
What are your thoughts on third party hardware developers using this chip to make processor upgrades? I'd like to see a 2Ghz G4 processor upgrade card for my old G4 PowerMac. This chip looks like a great candidate for such use also.

If the 750VX 400MHz frontside bus capable, does that mean that it could take full advantage of the DDR RAM that you have to stuff in the MDD Macs? If so, that combined with an increase in clock speed would probably provide an awesome upgrade for those machines in particular. Thoughts???

IVIIVI4ck3y27
Dec 12, 2003, 03:11 PM
Originally posted by Daaavek
If the 750VX 400MHz frontside bus capable, does that mean that it could take full advantage of the DDR RAM that you have to stuff in the MDD Macs? If so, that combined with an increase in clock speed would probably provide an awesome upgrade for those machines in particular. Thoughts???

Ummmm... it'd make a good upgrade, but you'd not be able to realize the performance advantages from it in utmost. The reason? The big bottleneck with most G4 machines that limits the success of DDR RAM is the bus itself. While you're correct in presuming that the 400 Mhz. frontside bus would be a major boon, I don't particularly see this overcoming the on-board difficulties of the motherboard which are the real problems as always. The added horsepower would do wonders, but like shoe-horning a G4 into a 9600 (my poison of choice)... you might have a processor churning out 700 Mhz. but it's still going through a tighter bus than a Graphite G4, and therefore will not have anywhere near the speed that a purpose-designed machine with less bottlenecks would have. It's still quite livable if you ask me... but I'm obviously looking around for a good deal on a graphite G4 to upgrade to next, as even running at the same Mhz. the machine would be faster (plus it'd run OS X natively rather than doing the XPostFacto Hoola-hoop dance). ;) I just don't have the $ to buy new, and I kind of am fond to my 21" CRT monitor. The 750vx to me is a viable processor upgrade... just keep in mind that if Apple elects to put the 750vx into a low-cost desktop, buying *NEW* will be at a supreme advantage because the I/O of the G4 desktops would be antiquated, and the bus of the G4 machines is far slower than what the 750vx is vying for. In a nutshell... always buy new if you can afford it.

That said...

If Apple would give us a headless eMac... I'd be saving my pennies in a heartbeat. That'd the be ticket for the 750vx too, as it's cheap and has adequate performance for some of us that aren't ready to drop $1,500+ on a new computer sans monitor. All it'd need is Firewire 400/800, USB 1.1/2.0, a decent video card, and that's it. It wouldn't remotely begin to compete with the G5 and it'd have more people buying "NEW" machines from Apple than investing processor upgrades from Sonnet, Newer Tech, Powerlogix, and OWC. I don't even need a rash of PCI or PCI-X slots (maybe 1, if any). If I have good onboard video and sound (with potential for upgraded video over time) and USB/Firewire I have all of the expansion I need. It'd just need to come in around $400-500 and Apple would struggle keeping them on the shelves. eMac LC anyone? Video input... I could do with a DV through Firewire, or buy an ADS solution. Sound? I'm sure M-Audio would gladly devise a solution that'd work fine, as they have some USB-based inputs that would work smashingly well for a lot of people's needs. Drive bays should be standardized ATA/SATA. I'm sure I could perform a drive swap if there's necessary space and nowhere near the hurdles of the original cube. What I want isn't a fetish item (cube part 2)... I think the iMac covers that nicely for some. A nice, simple, compact desktop or tower would be nice. It could even be a larger cube if it makes Steve happy.

Then again if anyone wishes to donate to the "Help Marcus buy a G5 fund..." I'll gladly work on securing a P.O. Box for the foundation. ;) I'm sure I could work out a deal to send you to a secret website where the fruits of your donations would be visible as you look on with glee at me operating the machine monthly. ;) :D I'd personally respond to every email with loving care. Pretty-please?!?

Telomar
Dec 12, 2003, 03:58 PM
Originally posted by AidenShaw
Another glaring exception is the PPC970's lack of little-endian support.... This is not part of the PPC ISA and as such is not an exception. All PPCs follow the PPC ISA.

AidenShaw
Dec 12, 2003, 09:53 PM
Originally posted by Telomar
This is not part of the PPC ISA and as such is not an exception. All PPCs follow the PPC ISA.


What about the PPC instructions: Load Half Word Byte-Reverse Indexed (lhbrx), Load Word Byte-Reverse Indexed (lwbrx), Store Half Word Byte-Reverse Indexed (sthbrx), and Store Word Byte-Reverse Indexed (stwbrx)?

Do some research....

Telomar
Dec 12, 2003, 11:10 PM
Originally posted by AidenShaw
What about the PPC instructions: Load Half Word Byte-Reverse Indexed (lhbrx), Load Word Byte-Reverse Indexed (lwbrx), Store Half Word Byte-Reverse Indexed (sthbrx), and Store Word Byte-Reverse Indexed (stwbrx)?

Do some research.... All are still present and supported with the PPC970. What broke VPC was not the lack of a PPC instruction. You're free to validate this on your own time. I might suggest first ensuring you are familiar with the PPC 970 by referring to the PPC 970's programming documentation located here. (http://www-306.ibm.com/chips/techlib/techlib.nsf/techdocs/F6153E213FDD912E87256D49006C6541)

blue&whiteman
Dec 14, 2003, 10:08 AM
why are so many of you down on motorola? they have been making apple cpus since the first mac in 84 and even before that. they have been there with apple 100% even when it wasn't much of a profit any longer.

motorola has had a big part in making macs what they are and always have been so give them some slack. just because they can't live up to the expectations of all you speed freaks doesn't mean they should be just up and forgotten. they have helped make apple what they are today.

IVIIVI4ck3y27
Dec 14, 2003, 11:30 AM
Originally posted by blue&whiteman
why are so many of you down on motorola? they have been making apple cpus since the first mac in 84 and even before that. they have been there with apple 100% even when it wasn't much of a profit any longer.

motorola has had a big part in making macs what they are and always have been so give them some slack. just because they can't live up to the expectations of all you speed freaks doesn't mean they should be just up and forgotten. they have helped make apple what they are today.

I'm not so much down on Motorola as I do understand that companies are only appreciated when they produce and meet demand. In Apple's case, it was a scenario where Motorola was not up to task, and like any "worker" in a business, they're being collectively put out to pasture in favor of those that do perform.

Let's face it... I've always thought Motorola was a cool company. I always thought it'd be neat to have a Mac with a pair of the "Batwings" on it because I've always thought Motorola to be pretty cool (although their consumer electronics have always been more intriguing than their desktop attempts, with exception to the PPCP/CHRP machine, and none of us have seen it in person). Very true, they did create the 680x0 lineup of processors that was used in all early Macintosh computers from the original to the final Quadra and Performa lineup. Hell, goes back to before the Macintosh if you really want to get clinical.

So yes, Motorola is a very cool company... fact is though, they didn't perform after being kicked to the curb by Apple when Apple nixed the clones. Can I particularly blame them? Not entirely... if I was Motorola I'd probably be pi55ed off too. Yet, when you sign a contract with a company and don't produce what you say you're going to produce, when you're going to produce it... there comes a time for loyalty, and a time for reality. Motorola's time of loyalty has long since been over. That's the bottomline.

Yet I also feel their reason to support the Macintosh platform and Apple was in limbo the day Apple put a fork in their release of their original G3 machines. If that day hadn't come about, and Apple had reworked an agreement where Motorola and *ONLY* Motorola was allowed to build and *ONLY* build clones that didn't affect Apple's current marketshare and bottomline, I feel Motorola would've produced on the G4, we'd not even be concerned about the PowerPC 970, and things would be kosher. So it's a bit of give and take here. I know there are many Apple zealots trolling on here, much as there are PC zealots trolling on here and elsewhere bashing Apple... but the fact is, Apple and Motorola both have committed their own acts. It's probably best to let bygones be bygones and let Motorola and Apple both move on in their separate ways.

Apple and IBM is the future of the Mac...

wizard
Dec 15, 2003, 02:05 AM
If what you say is true about virtual memory being 64 bit on todays G4 Macs, how come we are limited to 4Gb of addressable virtual memory range on G4 Macs? It seems to me if what you say is true, a process should have a much larger virtual address space available to it. Maybe I got some info crossed but this is my current understanding of the G4 environment.

I'm not disagreeing with you about 32 bit support, there is a lot of potential life left in 32 bit hardware. Well there would be if Aple had a decent supplier for 32 bit hardware. Of course that is what this thread is all about.

Thanks
Dave

Originally posted by Rincewind42
Apple can work towards a fully 64-bit OS anytime they want. And they won't have to leave 32-bit only CPUs behind - they just won't be able to run 64-bit only OS components. The OS Core is 64-bit clean, and some parts of it do run as if they are in a 64-bit memory space (Virtual Memory for one). If you installed it on a 32-bit CPU you'd just not get the 64-bit parts and would run as a you do now. The biggest hurdles right now are the fact that many of the frameworks are not fully 64-bit clean, and that you would need to ship twice as much code (a 32-bit and 64-bit version of every currently 32-bit library). So while they could introduce a 64-bit OS that still supported 32-bit CPUs, it will not likely be soon.