Jump to content

Apple Wants New Intel Chips Early


Swad
 Share

16 posts in this topic

Recommended Posts

Check out this interesting story on our Main Page.

 

Should Apple get the newest Intel chips early?

 

:) Oye... This just keeps getting worse... Jobs states in the last article that OS X will only run on Intel Based Macs, now we find out he's gunning for top of the line chips. From the sounds of it, this mactel system will be expensive. I've said it before, and I'll say it again -- Apple needs to lure people from Windows, and to do that they have to keep costs down. A proprietary Intel based system isn't any better than a proprietary PPC mac. I don't mind if Apple makes expensive Intel based hardware, they should still make Tiger run on as many computers as possible. Runs on anything, but if you really want to run Tiger well with full Quartz support, run it in a Mac / Intel system.

 

To get people to try the OS, you first have to target a copy that will run using VESA 2.0 - 3.0, on as many PC's as possible. A lot of this work has already been done by the OpenSource Darwin Project. Once you have the so that it will run dual-boot with Windows, and priced about $75 bucks, people should give it a try. Then once they've switched, and now they're using Tiger as their primary OS, when its time for an upgrade I could see people buying a Mactel system. I don't understand why they won't open up the OS to allow people to try their product. I would even do what Knoppix or SuSE are doing -- put together a bootable runs-only-from-the-DVD version and give it away. Point is, if you get people to try OS X, I think there are a ton of people who will like it, and as a result will go buy a mactel system because of it.

 

If CodeWeavers is successful in getting a copy of Crossover Office for mactel systems, this makes the point even better. I've been using Crossover since it first came out for Linux, and it's really an awesome program that allows you to run a large percent of Windows Apps. This is key for the person who has a large inventory of software they've purchased for their windows system. Being able to run them on their Mactel system would push this user into the OS X world. Again the only way this is going to happen is if Apple makes OS X available to run on as many PC systems as possible, makes it inexpensive for people to try, and just for kicks bundle a special copy of Crossover Office by Codeweavers the same way Xandros Linux does. Think of it, one system that will run PPC / Intel OS X Binaries, OpenSource X windows apps (KDE and GNOME), along with Windows apps, What you'll have is a PC that can instantly has thousands of titles available. But Apple will never do it... :(

Link to comment
Share on other sites

I doubt Intel will do anything...

 

As I mentioned in another thread, I suspect that this is actually a "done deal", whereas Apple secured the privilidge of getting new Intel chips before agreeing to the Intel switch. I mean, I would not expect anything less from Steve Job's marketing and negotiating brain. The alternative being presented, that Apple is just now trying to cut in line seems unlikely. Otherwise, I think that Intel has a lot good reasons to play along with this game. It adds the Apple/Jobs hype factor to their chips, even when Dell gets them second.

 

AppleInsider also picked up this story with a corresponding forum discussion: http://www.appleinsider.com/article.php?id=1285

Link to comment
Share on other sites

Why can't Apple use Yonah for the first wave of computers, then eventually upgrade computers to using Woodcrest and Merom processors later on down the road?

 

Well, I think this is basically what most informed people expect at this point and sooner rather than later. However, I am not sure the "uprade" word will apply here as Yonah is 32-bit while the others are 64.

Link to comment
Share on other sites

Why can't Apple use Yonah for the first wave of computers, then eventually upgrade computers to using Woodcrest and Merom processors later on down the road?
Yep. In view of the roadmap, Yonah is an excellent choice for early entry-level appliances. Actually I don't really see why Apple would want to put Merom CPUs also in their initial offerings. IMO, that would make their middle- and upper-level systems less attractive. The first x86-based Mac Minis and iBooks featuring a Yonah processor would already be very innovative designs and appeal to many "early adopters" who want to get a glimpse of what OS X on Intel looks like. With upcoming mid- and pro-level machines like the iMac, Powerbooks and -Macs featuring 64-bit support and improved overall performance with Merom, Conroe and Woodcrest CPUs, Apple will then have the opportunity to market the improvements as another leap in development, enabling them to cash in on the afficionados again. Looking at the lively iPod market (and watching the mass of people here who purchased parts and built their own new, but potentially quite shortlived Pre-"Mactels"), I have absolutely no doubt that -- provided "the price is right" -- the die-hard Apple fan crowd will buy twice: first the small and then the bigger systems when they're finally out, no matter if there's just a six-month timespan between.

 

Anyway, there are so many things you can do with a Yonah-based Mac Mini or iBook you have in spare -- using it for yourself, giving it to friends or relatives, re-selling it again with little loss -- if you decide to get another, updated Macintosh. Even with new Merom machines on the shelves, old Yonah will still be quite attractive.

 

The last time I had a look, small form-factor Pentium-M boards were only available for industrial appliances and costed almost a fortune (usually about 300 USD, IIRC, just for the plain board). If the Apple folks manage to market a complete Mac Mini with a Pentium-M derived CPU for under 600 USD (regardless of which clock speed, memory configuration, hard drive size, type of the optical drive, graphics chip, number of cores this machine will have, and whether it features 64-bit support or not as an additional gizmo will be even more irrelevant), it will be a bargain and catch the attention of the average Joe as well as the geeky folks who are looking for a neat, small, cool and quiet but nevertheless fast small form-factor PC to replace their aging VIA EPIAs. Generic boxes like Dells might offer better performance and features for a lower price, but even if we ignore OS X as a unique selling point, they still simply play in another league and are no direct competitors to the Mac Mini. The same is true for the iMac, but I guess Apple will face direct competition in the notebook sector.

 

I am also convinced that if they succeed to put out their entry- and mid-level x86 stuff relatively early next year (whether it's all Yonah- or also partly Merom-based won't matter so much, methinks), there will be less pressure on them to keep up the pace with the pro stuff and to anticipate workarounds in case Intel may venture difficulties and delays with their roadmap on server- and workstation-class processors. Apple don't really have a choice with notebook and small desktop processors yet, but AMD may serve as an alternative supplier in the workstation and server business on a long-term scale and that might be a viable backup plan (although very unlikely) if Intel fails to deliver in time.

 

 

First they literally flame Intel for running too hot, then they eventually do something smart and suddenly become best friends with Intel out of nowhere.
Intel deserved the "flaming" for their hot and inefficient processors as much as they deserve the praise for their excellent revised designs. By dumping the overheating G5, skipping the ill-fated P4 and opting in on Intel's new promising P-M designs, Apple's decision makers stay true to themselves, as odd as it may sound.
Link to comment
Share on other sites

Yes, I agree with the last statement - the switch is really more inline with Apple's traditional design ethos than anything. It's not a radical shift from the standards they've always had - the only shift is that of identity and the maker of the processor.

 

After thinking about it though, I'm pretty sure this will have to be the last time that Apple will ever change architectures.

 

And, for the record, I think they don't deserve the chips early, although I assumed that this was one of the aspects of the deal they made to switch to Intel. If not early processors, what other hidden benefits might they be getting?

Link to comment
Share on other sites

Yes, I agree with the last statement - the switch is really more inline with Apple's traditional design ethos than anything. It's not a radical shift from the standards they've always had - the only shift is that of identity and the maker of the processor.

Well, although I basically agree with the terry's post, I do not agree with this part at all. First of all, I do not think there is anything particularly wrong the G5 chip and neither does the vast majority of the Mac community. Furthermore, the lastest round of IBM announcements, the low voltage G5 (970FX) and dual core G5 (970MP), show that IBM had the goods Apple wanted. Finally, as CPU design evolves from a dual-core to an n-core mode, IBM is clearly way ahead with the Cell architecture.

 

From my point of view, the switch to Intel was simply done for the very fact that Apple is implementing a radicial shift in it's Macintosh product strategy. If Apple wants to regain marketshare, as it must to survive over the long term, it must push Macintosh/OS X into the mainstream, which is, of course, x86. Apple has clearly been planing this more for years while this latest business with dishing IBM and hyping Intel is just social engineering. Someone like Steve Jobs certainly knowns that Intel's roadmap is equally useless as IBM's. I mean, certainly he has not forgotten how he promised a 3GHz G5 by July 2004.

 

As opposed to whether or not this is Apple's last switch, I think a more interesting question is if Apple will actually drop PPC entirely or not. The issue really comes down AltiVec versus SSE3. For professional digitial content users AltiVec is vastily superior to SSE3. While IBM has already laid out the Cell archetecture as the future for AltiVec, Intel has no plans to even fix what is presently wrong with SSE3. The point is that Apple risks losing its "Pro App" market and recent success in supercomputing to Linux on Cell because Intel in not interested in vector processing. I am not sure if Apple is going to let all that go for the minimal cost of continuing to support a PPC version of OS X.

 

Ultimately, I suspect the sucess of the Cell chip will determine whether PPC OS X has any future, but so far the Cell is proving to be much less than what people expected for game consoles. On the flip side, I suspect the success of x86 OS X will lie in whether or not Apple decides to license it out to the likes of Dell and HP, which of course would clearly be a radicaly departure from the "whole widget" business model.

Link to comment
Share on other sites

Well, although I basically agree with the terry's post, I do not agree with this part at all. First of all, I do not think there is anything particularly wrong the G5 chip and neither does the vast majority of the Mac community. Furthermore, the lastest round of IBM announcements, the low voltage G5 (970FX) and dual core G5 (970MP), show that IBM had the goods Apple wanted. Finally, as CPU design evolves from a dual-core to an n-core mode, IBM is clearly way ahead with the Cell architecture.

 

From my point of view, the switch to Intel was simply done for the very fact that Apple is implementing a radicial shift in it's Macintosh product strategy. If Apple wants to regain marketshare, as it must to survive over the long term, it must push Macintosh/OS X into the mainstream, which is, of course, x86. Apple has clearly been planing this more for years while this latest business with dishing IBM and hyping Intel is just social engineering. Someone like Steve Jobs certainly knowns that Intel's roadmap is equally useless as IBM's. I mean, certainly he has not forgotten how he promised a 3GHz G5 by July 2004.

 

As opposed to whether or not this is Apple's last switch, I think a more interesting question is if Apple will actually drop PPC entirely or not. The issue really comes down AltiVec versus SSE3. For professional digitial content users AltiVec is vastily superior to SSE3. While IBM has already laid out the Cell archetecture as the future for AltiVec, Intel has no plans to even fix what is presently wrong with SSE3. The point is that Apple risks losing its "Pro App" market and recent success in supercomputing to Linux on Cell because Intel in not interested in vector processing. I am not sure if Apple is going to let all that go for the minimal cost of continuing to support a PPC version of OS X.

 

Ultimately, I suspect the sucess of the Cell chip will determine whether PPC OS X has any future, but so far the Cell is proving to be much less than what people expected for game consoles. On the flip side, I suspect the success of x86 OS X will lie in whether or not Apple decides to license it out to the likes of Dell and HP, which of course would clearly be a radicaly departure from the "whole widget" business model.

 

Apple is part of the PowerPC "AIM" partnership. That means they have rights to Alti-Vec. Why could they not license it to Intel for integration into all their Intel chips? Would that not be possible?

 

Benefits all around.

Link to comment
Share on other sites

Apple is part of the PowerPC "AIM" partnership. That means they have rights to Alti-Vec. Why could they not license it to Intel for integration into all their Intel chips? Would that not be possible?

While I do not know for sure, I would suspect that whatever rights Apple has to AltiVec preclude giving it to Intel for general use. (Again, I believe that Apple does not want to be on "custom" hardware, but rather that the switch to x86 in all about compatibility and moving into the mainstream) And even if Apple could just give AltiVec away it is not like Intel could just tack it on to the side of Dothan or something, there would be a lot of technical problems that would probably require a major overhaul of any host CPU. Finally, as I said above, the real problem seems to be that Intel has no interest in doing to improve the vector performance of its processors. Basically, the way I see it is that the Windows world has never really embraced MMX/SSE#, partly because it has been done poorly from the beginning and is unstable, so Intel considers it to a waste of time and die space, while the Mac community had to leverage AltiVec, which was done right the first time, to survive while losing the megahertz race in the G4 era.

Link to comment
Share on other sites

Intel deserved the "flaming" for their hot and inefficient processors as much as they deserve the praise for their excellent revised designs. By dumping the overheating G5, skipping the ill-fated P4 and opting in on Intel's new promising P-M designs, Apple's decision makers stay true to themselves, as odd as it may sound.

 

One more thing, Apple certainly could have had Intel's Itanium chips at a super bargain price. Intel has spent billions developing the Itanium and almost nobody will touch them, and it is not because they suck as they lastest verisions are quite good. As an instruction set per se, x86 does suck compared to PPC or Itanium's EPIC: it is over twenty years old and even Intel hates it. Unfortuantely, it is just another piece of legacy {censored} that Microsoft has stuck us with. There is one main reason that Apple want to be on x86, again, my point here is that Apple is switching to x86 for platform compatibility.

Link to comment
Share on other sites

another piece of legacy {censored}
You might laugh at me, but I'm quite happy that in the PC world legacy support remains so important. I do not want to trade in my good, robust 13-year old HP LJ4 for any new printer, and I also still do use quite a good amount of old 16-bit software for which I do not have an equivalent, modern replacement yet, and which I like to run seamlessly with my other apps and not on a separate desktop in an emulation environment. Of course this also means that I am facing a hard time with the upcoming 64-bit Windows, which will break that 16-bit compatibility. I know, all this whining probably makes me look like a dinosaur, but I swear, I'm just in my late twenties!

 

One piece of old software I favor very much is for instance Micrografx ABC SnapGraphics 2.0 from 1994. This is so easy to use and produces such beautiful outputs that I'm constantly impressed everytime I try to do something new with it. I do not know any contemporary software package that is so much fun to use, including Visio and SmartDraw. If you don't know it, you can still get Version 1.0 (which isn't quite as nice as 2.0, however) for free from this source:

 

http://www.winsite.com/bin/Info?500000027108

 

Further info on 2.0:

 

http://www2.chass.ncsu.edu/garson/pa540/snapgrap.htm

http://www.findarticles.com/p/articles/mi_...v13/ai_17101356

 

Or take Lotus Agenda from the late 1980ies, that's also a great program which in many aspects is yet unrivalled:

 

http://www2.support.lotus.com/ftp/pub/desk...a/dos/2.0/misc/

http://cheerleader.yoz.com/archives/000141.html

 

But these are of course not the only 16-bit legacy applications I use... There are also some very expensive old literature and encyclopedic databases that I do not want to miss.

 

So the bottom line I want to draw is, to some legacy support might appear like {censored}, for others it is more like a blessing.

Link to comment
Share on other sites

But these are of course not the only 16-bit legacy applications I use... There are also some very expensive old literature and encyclopedic databases that I do not want to miss.

 

So the bottom line I want to draw is, to some legacy support might appear like {censored}, for others it is more like a blessing.

 

I am really not arguing against this per se, but shouldn't have parallel and PS/2 ports, and 3.5" floppy drives disappeared from new PC hardware by now?

 

My main point was that Microsoft could have and should have taken advantage of the numerous opportunities, the lastest being of course Longhorn/Vista, to declare x86 dead and work with semiconductor industry to set a new and better designed standard. Instead, exactly the opposite has taken place, AMD has extended x86 and Microsoft has killed Windows for Intel's fledgling Itanium. This is yet another example of Microsoft systematically failing to take leadership role, when IBM owned the industry they were never ever this bad.

Link to comment
Share on other sites

I am really not arguing against this per se, but shouldn't have parallel and PS/2 ports, and 3.5" floppy drives disappeared from new PC hardware by now?
Funny you didn't also include serial ports... What's wrong about still having the opportunity to use such legacy stuff if one wants to?

 

My main point was that Microsoft could have and should have taken advantage of the numerous opportunities, the lastest being of course Longhorn/Vista, to declare x86 dead and work with semiconductor industry to set a new and better designed standard. Instead, exactly the opposite has taken place, AMD has extended x86 and Microsoft has killed Windows for Intel's fledgling Itanium.
Have I missed something? I thought Intel's Itanium and its instruction set were never planned to go mainstream, but to be reserved to high-end use. IIRC, there's still an Itanium version for Windows coming, it's just not for the desktop, yet for some very specific high-end use.

 

This is yet another example of Microsoft systematically failing to take leadership role, when IBM owned the industry they were never ever this bad.
What are you talking about? Applications for the average user? When IBM was a synonym for PCs, the market was very different from today, still in its beginning and much smaller than now. This opened up much more opportunities to try something new than it is possible today. And Microsoft must have become big for a reason, they certainly haven't done everything wrong. I know, Windows 9x was all horrible, but the market demanded for it as a connection piece in the long transition from DOS to Windows NT. They can't hit their large user base in the face by expecting them to dump their existing hard- and software from one day to another. This is totally irrealistic.

 

Besides that, Microsoft aren't "systematically failing to take leadership role", they're rather having their conservative users in mind. I am under the impression that you don't really have an idea how much old soft- and hardware is still in use all around the globe, even in corporate environments. There must be a reason why Microsoft started to develop Eiger (now officially named "Microsoft Windows Fundamentals for Legacy PCs", "WinFLP"), a cut-down Windows XP SP2 version for Pentium-class PCs with 64MB RAM and 500MB of free hard drive space.

 

Windows Fundamentals for Legacy PCs for Software Assurance Available March 2006!

http://tinyurl.com/bj7ds

 

Microsoft Windows Fundamentals will convert a legacy PC to a thin client in March 2006

http://www.brianmadden.com/content/content.asp?id=498

 

WinFLP, first install experience

http://tinyurl.com/94wgh

Link to comment
Share on other sites

 Share

×
×
  • Create New...