Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Roller said:
In his keynote last week, Jobs confirmed longstanding rumors that Apple had been developing an Intel-compatible version of OS X for years (a secret double-life, as he put it).

In order to support PPC-based Macs, Apple's going to have to contunue to develop both versions for several years, well after the entire product line goes Intel and Leopard ships. At least Apple has a model for that in the 68k-PPC transition.

The question is will Apple continue to hedge its bets beyond that? For that matter, will they also create in-house versions for other processors like the Cell?
wouldnt that be a lot of trouble for something steve jobs said wasnt going to be in desktop machines
 
Wyrm said:
You are correct.

Out of order execution helps with a compiler that can't perform the out of order at compile time. I'm sure there are some special cases that the compiler couldn't do this, but I can't think of any right now, and I'm sure they are as rare as a Coelacanth in your bathtub.

I think most compilers now address this, so all that silicon is effectively redundant now. Well, except for ancient compiled code that is, and where does that exist in the Mac world?

-Wyrm

Out of order comes into effect when you have previous instructions waiting for data or execution units, while latter ones are ready.
 
MacTruck said:
No 4ghz chips ever to be produced. In reality 2.8ghz is the fastest anybody can make a chip. P4 is expanded to increase clock speed not performance. Look at AMD, Intel Pentium M ( a real cpu unlike P4) and IBM. All have maxed out at these speeds roughly. 65nm chips coming out? How did they tackle the problems inherit in the 90nm process? The only way computers will get faster are less bloated OSes and software. If I were to run windows 95 on my 2ghz Pentium M thinkpad it would be the fastest computer ever. XP slows it down just like OSx slows down a new mac compared to OS9.

"Intel's leading strained silicon technology, first implemented in our 90-nm process technology, is further enhanced in the 65-nm technology. The second generation of Intel strained silicon increases transistor performance by 10 to 15 percent without increasing leakage. Conversely, these transistors can cut leakage by four times at constant performance compared to 90-nm transistors. As a result, the transistors on Intel's 65-nm process have improved performance without significant increases in leakage (greater electrical current leakage results in greater heat generation)."

http://www.intel.com/technology/silicon/si08042.htm


I am sure once intel reaches 2.7ghz with Pentium M (if they ever do) they will be using liquid cooling and heatsinks the size of shoe boxes for their cooling needs as well.

Pentium-M's have been overclocked to 2.6 Ghz with no problem. http://forums.legitreviews.com/about1525.html

Lets just get something straight, intel has nothing revolutionary in the pipeline.

Depends on how you defined revolutionary. I think Jobs is onto something with the low-power as revolutionary. Taiwan Inc. has recently begun releasing desktop mobos for Pentium-M processors. When I travel I like to use my XP notebook (w/ Pentium-M) because it has a longer battery life and runs cooler (though, admittedly, part of the issue with the battery life is the ***-up battery in my 12").

Voice activation, dual core. Whatever. It is all hype. Let me see it in the store then I will believe it.

Can't argue with this.
 
what a pretty Coelacanth

Wyrm said:
Out of order execution helps with a compiler that can't perform the out of order at compile time. I'm sure there are some special cases that the compiler couldn't do this, but I can't think of any right now, and I'm sure they are as rare as a Coelacanth in your bathtub.
For starters, cache misses are dynamic and cannot be predicted by the compiler - especially if the program contains conditional statements and loops. (The compiler can make some guesses about cache behaviour - but in a multi-tasking, multi-processing there's a lot of randomness.)

When the compiler does it, OOOE is called "scheduling". It's fine if you're willing to schedule your application for a particular processor, but you have to choose whether it should run better on a PPC970 or a PPC970fx (even among closely related chips, there can be differences in scheduling, cache behaviour, ....).

When the chip does OOOE, however, the chip itself can "reschedule" dynamically - accounting for both randomness and architectural differences.
 
Why Jobs unveil so soon?

poundsmack said:
intel had to do some good enticing to apple too....i mean intel must have made this deal worth wild for apple


Yes. But the thing I still don't understand is why Jobs would unveil this information so early. People have been saying all along that this is really going to drive down the sales of the current line up. How could Jobs overlook that fact? Does he not want to sell computers for another what few years before the Intel line-up comes to existence?

Apple is making the switch to Intel and ditching the old stuff. Who would want to buy the old stuff now? I mean does anyone else see how this could really hurt Apple? I just don't see the logic.
 
Jmitch said:
Yes. But the thing I still don't understand is why Jobs would unveil this information so early. People have been saying all along that this is really going to drive down the sales of the current line up. How could Jobs overlook that fact? Does he not want to sell computers for another what few years before the Intel line-up comes to existence?

Apple is making the switch to Intel and ditching the old stuff. Who would want to buy the old stuff now? I mean does anyone else see how this could really hurt Apple? I just don't see the logic.

As has been mentioned countless times before in this and other threads, Jobs has done everything to ensure sales will not decline. He has indicated this will be a gradual transition, he introduced XCode 2.1 with universal binaries so existing and new PPC machines will not become useless once the Intel systems are released, and said there are still many PPC products in the pipeline. He has gone to great lengths to make the transition a smooth one.

Also, there is absolutely nothing wrong with his timing - it is perfect, in fact. Do you think the developer community would prefer to find out about this switch only a few months before it happens? That would royally screw them over! How would they possibly have enough time to prepare and port their apps? The whole thing would be a mess and there would be a lot of ill feelings towards Apple as a result. With the announcement coming now, Jobs has given all the developers out there fair warning and had allowed for them to prepare and begin coding with universal binaries so that when the migration actually begins, everyone will be prepared and everything will run smoothly. Jobs announced this when he did because he had the developers in mind, the timing was perfect.
 
Wyrm said:
I'm still for making OSX support both architectures, forever! That way you give IBM a little incentive to improve the PPC970. A little more cache perhaps? I mean at least Motorola was active in improvement... not getting very far, but they were trying.

I agree. With Steve's penchant for pissing people off or throwing one of his kiddy tantrums, we could find ourselves going through a similar transition 5 or 6 years from now when he dumps Intel because he's not happy about something. A frightening thought, yet not out of the question.
 
kaos said:
Just from my own sampling of Apple fanatics - friends who buy a lot of Apple gear, from G5s to PBs and iPods - I think there may be a far more significant Osborne effect than Jobs and Apple realize. Personally, I used to update my PB every 9mos or so. I'm going to try and hang on to it now until Intel's real notebook procs hit in '07.

Fortunately for Apple, I think you are the minority with your 9 month update cycles on $2000 - $3000 laptops...
 
AidenShaw said:
For starters, cache misses are dynamic and cannot be predicted by the compiler - especially if the program contains conditional statements and loops. (The compiler can make some guesses about cache behaviour - but in a multi-tasking, multi-processing there's a lot of randomness.)

When the compiler does it, OOOE is called "scheduling". It's fine if you're willing to schedule your application for a particular processor, but you have to choose whether it should run better on a PPC970 or a PPC970fx (even among closely related chips, there can be differences in scheduling, cache behaviour, ....).

When the chip does OOOE, however, the chip itself can "reschedule" dynamically - accounting for both randomness and architectural differences.

This hits it. The Cell is not designed for a modern multi-tasking operating system like MacOS or XP. The Cell is designed to run an OS that is running just one Media/Render-intensive program at a time.
 
Dont Hurt Me said:
IBM doesnt even advertise their own 970s, they advertise their Intel Powered machine.

IBM does advertise their blade servers on TV/cable. those are available with IBM or Intel chips.
 
beatle888 said:
im not so sure about that whole media/computer systems now. i just ordered a new cable box from my cable company for $9 a month. it records up to 70 hours of tv. its like a tivo. it seems that the cable companies already got a head start in this area. they already have movies you can preview>watch>record...and its happening NOW.

cable companies have the content>delivery method>users and now an inexpensive box.


if you ever used TiVo you would see how much better their software/interface is.

cable companies also already have you paying for the box, and that cost is subsidized with your cable bill. you cable company is also alrady feeding your box a schedule gris (TiVo has to make a national schedule your Tivo can sync with). my understanding is that cable companies added PVR features because they were losing customers to DSS (which has has PVR for years now). supposedly if they had a way to keep out theft, but not have to provide boxes the cable companies would back it. they see boxes as needed technology now, but a big expense.

that's why the cable companies, in general, strongly support the future of cable cards (1). the thing is they hate generation 1 cards (out now) because they are one way cards. they will not allow interaction (you to buy pay per view movies or video on demand stuff). once the 2nd generation of cable cards rolls out i wonder how many cable companies will still promote PVR boxes. you could buy an off the shelf TiVo and pop in the card from your provider.

(1) if you do not know what a cable card is, it is literally just a little card with the proprietary info from your cable company. it allows them to know you have the device on the cable network and it will allow the channels you subscribe to. instead of the $9 a box cost to rent a month they are more like $1 or $2. you use the TV (or other device, like tivo) tuner to pick a channel and not a set top box. if you look on the back of a new larger TV you will see a cable card slot. there is a government mandated rollout date that is in progress now... starting with larger TVs and working its way down.
 
~Shard~ said:
As has been mentioned countless times before in this and other threads, Jobs has done everything to ensure sales will not decline. He has indicated this will be a gradual transition, he introduced XCode 2.1 with universal binaries so existing and new PPC machines will not become useless once the Intel systems are released, and said there are still many PPC products in the pipeline. He has gone to great lengths to make the transition a smooth one.

Also, there is absolutely nothing wrong with his timing - it is perfect, in fact. Do you think the developer community would prefer to find out about this switch only a few months before it happens? That would royally screw them over! How would they possibly have enough time to prepare and port their apps? The whole thing would be a mess and there would be a lot of ill feelings towards Apple as a result. With the announcement coming now, Jobs has given all the developers out there fair warning and had allowed for them to prepare and begin coding with universal binaries so that when the migration actually begins, everyone will be prepared and everything will run smoothly. Jobs announced this when he did because he had the developers in mind, the timing was perfect.

Exactly. Also, think about Tiger's release. A large number of apps were ported to work on Tiger by its release date. When the first Intel box hits the shelf I bet that most apps will be ready to go. The only reason sales will drop will be ignorance and misinformation. PPC Macs will be supported for many years to come and you can bet that developers will support PPC and Intel Macs for years to come as well.
 
kaos said:
Just from my own sampling of Apple fanatics - friends who buy a lot of Apple gear, from G5s to PBs and iPods - I think there may be a far more significant Osborne effect than Jobs and Apple realize. Personally, I used to update my PB every 9mos or so. I'm going to try and hang on to it now until Intel's real notebook procs hit in '07.

And, given the transition timeframe, I'm going to do what was unthinkable before this transition was announced - take a real look at Longhorn and evaluate moving back to Windows. I think it's likely I'll stick with OS X, but I would probably not have even given Windows serious consideration without this move by Apple. Hopefully (for Apple) I'm more the exception than the rule, but I think this was the wrong time to make such a transition.

obviously nobody knows for sure, but the notebooks may be first to jump in mid 2006. the assumption is that Apple will take all the G4 based machines and roll them to Intel starting about June 2006. the actual quote from Jobs said that by the time WWDC 2006 happens the first of the Intel based systems will be out. since Intel has has experience getting their chips into laptops etc it is possible that Apple will want to move the laptops before the Mac Mini. the Mini will probably be using a laptop processor anyway (so the rumors say). who knows what will happen with the eMac in the next year or so anyway. if an iMac can be a G5 i would think an eMac could be too. the fate of the eMac is probably dictated by schools and how many still want a CRT Mac. that leaves Powerbooks and iBooks with G4 chips, and doubtful G5 upgrades coming.

to be honest i think i will be buying at least one more PPC Mac.. maybe two. both my desktop machine and my ibook are getting old. i have been pondering an upgrade for a little while now (i just keep spending the money on something else). i also have no fear of owning a PPC Mac when the Intel ones come out. i trust it will work for a while to come. at some point it may not run the newest software, but it will still work. any computer you buy will get old. that's just how it is.

while the laptops may be on the sooner side of the upgrades... i do not think my desktop tower can hold out for 2+ years. if you watch the keynote he said that most all Macs *should* be over to Intel by June 2007, and if not then by the end of 2007. 2.5 years without upgrading an already older desktop machine is not sounding like something i want to do. i am happy enough with the PPC chips. i really believe a lot of the decision was Apple looking at the future. the problem is not just that IBM is a little behind speeds today, but what they PLAN on having out in 5 years versus what Intel plans to have out in 5 years. maybe IBM is not optimistic about turning the next POWER chip into a PPC? maybe their ideas do not sit well with Apple. who knows. IBM will still be selling the chips to game consoles, and they still sell them in servers.


as for switching to M$ Windows......
that's silly. Apple will be shipping some Intel based Macs before Longhorn is on the shelf. 10.5 will ship late 2006 or early 2007. that's before all Macs are even Intel, so no matter how paranoid your friends are we know that OS will totally support PPC chips. i think it's a safe bet that 10.6 will support PPC chips too, and Apple is slowing down on OS X releases. say 10.5 comes out in January 2007, then it is not unrealistic to expect 10.6 around January 2009? remember not everyone upgrades the day the new OS ships. even if PPC support is not in 10.6, your computer you buy this week can be totally supported till at least sometime in 2009? that seems adequate to me. that's longer life than M$ windows machines will last. they say a lot of PCs being sold today do not meet the expected requirements for Longhorn.
 
jcgerm said:
The only reason sales will drop will be ignorance and misinformation.

That pretty much sums it up. I'm curious however who will actually be ignorant to all of this - the tech savvy Mac users like us, who follow the goings-on of Apple closely, watched the Keynote, etc., understand the situation and realize this is not something to be concerned about with respect to compatibility, etc. On the other end of the spectrum, there are Mac users who couldn't even tell you what processor is in their Mac right now, and may not even be aware when Macs make the move to Intel, and won't care, because everything will "just work". I guess it will be the people in the middle whom we'll have to properly educate and ensure there is no misinformation out there. But hey, that's what these Forums are for! Communication and education is a good thing. 🙂 😎
 
MacVault said:
yea, but... who the heck would be interested in buying a PowerPC(Mac) from now on? So, maybe there's no need for IBM to develop anymore. Just make a few of the current designs to sell til the switch to Intel.

a lot of people do not care. they buy a Mac for OS X, not for PPC chips. if they need a new computer now, they are not going to wait till 2007. i have talked to Mac people this weekend that did not even hear about the Intel switch. they were mildly curious, but did not really care. it will have no impact on their plans to upgrade. some were planning on buying new Macs within the next month or so. one was going to buy and iBook in the next few days. you may call them dumb, but that's not how they see it. not everyone freaks out about being one step behind when a new Mac comes out. they buy their computer to help them do their work, not for bragging rights. these are people that do design work or other things that will benefit from having a newer machine.

how many people have a relatively high end mac and use it for nothing more than internet access, iTunes, text editing and iPhoto? i know a few.
 
m-dogg said:
Fortunately for Apple, I think you are the minority with your 9 month update cycles on $2000 - $3000 laptops...

fortunately? i would think the shareholders would like a lot more customers like that!
 
Jmitch said:
Yes. But the thing I still don't understand is why Jobs would unveil this information so early. People have been saying all along that this is really going to drive down the sales of the current line up. How could Jobs overlook that fact? Does he not want to sell computers for another what few years before the Intel line-up comes to existence?

Apple is making the switch to Intel and ditching the old stuff. Who would want to buy the old stuff now? I mean does anyone else see how this could really hurt Apple? I just don't see the logic.

it's not old. there will be new PPC hardware coming before the intel switch. they are not ditching, they are planning for the future. unless you are a developer, this change will not mean much of anything to you. it should mean faster machines, but to the average user it will be no different than if IBM suddenly got their chip fab in high gear.

like i said in another post, i know people that are unfazed and still planning to upgrade. people buy a new machine when they need it. if you can wait 2.5 years just to have the first Intel machine, then you must not really do anything important on your computer, or anything processor intensive (making me wonder why you NEED the first of the new Intel chips anyway).

let me tell you something..... cars will eventually get much better fuel efficiency and have safer designs. all current gasoline cars will one day be obsolete. people are still buying these old designs. internal combustion design is ridiculously old technology. i have no idea why any people buy them. do you? they must not read newspapers or watch TV.
 
Osborne effect

It's interesting that everyone is so concerned about declining sales in the interim period before the first Intel Macs are introduced. It makes sense. I think power users and notebook buyers may hold out. As a bit of a mix of both types of users, I've been holding out for the first major PowerBook update, whether it's a G5, dual core Freescale G4, or now, a Pentium-M.

What's great though is there have been countless posts on these forums in the past complaning about how Apple was so tight-lipped, and didn't offer any sort of roadmap.

Well now we have one. Granted it's drawn in the sand with a pointy stick, but it's something. I'll be very curious to see 1) how it affects sales and 2) if this is a new trend for Apple. If we start seeing consistent, frequent updates, maybe Apple will be willing to be a little more forthcoming. Apple's past schedule has basically consisted of infrequent, significant upgrades, so everyone made a game of figuring out the best times to purchase.
 
kaos said:
Just from my own sampling of Apple fanatics - friends who buy a lot of Apple gear, from G5s to PBs and iPods - I think there may be a far more significant Osborne effect than Jobs and Apple realize. Personally, I used to update my PB every 9mos or so. I'm going to try and hang on to it now until Intel's real notebook procs hit in '07.

And, given the transition timeframe, I'm going to do what was unthinkable before this transition was announced - take a real look at Longhorn and evaluate moving back to Windows. I think it's likely I'll stick with OS X, but I would probably not have even given Windows serious consideration without this move by Apple. Hopefully (for Apple) I'm more the exception than the rule, but I think this was the wrong time to make such a transition.

For Powerbooks I agree to certain extent. G5 products might not suffer too much since it will take at least one year and probably two years until most apps are updated (and optimized) for Intel procs. During this transition period, any G5 will run circles around any Intel-Macs when using Rosetta based apps.
 
Peripheral & Device Drivers

As it stands, peripheral support from some companies is lackluster/shakey at best when it comes to the Mac. I can't help but wonder how many manufacturers will decide this move to Intel is the final straw and decide to drop Mac support for their peripherals and/or devices? Anyone have an idea how much work will be required for drivers to work with OS-X on Intel?
 
I will buy a PowerMac 2.7 in August but???

I won't wait for the intel's definitely, I need a high end mac now but should I wait a little more? Maybe new IBM chips will come? Apple will update its towers definitely before the intels, probably twice. But still, when will the update come?
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.