Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Apple in Talks with Intel?
- Thread starter MacRumors
- Start date
- Sort by reaction score
iGary said:
No it doesn't. I just wish my PowerBook was faster though. Why? So I could run Virtual PC faster, for one. All my work is done on a PC. I have one program that necessates use of an ISA card since the company refuses to update the card to PCI. I can't even run that program off of Virtual PC.
It sucks being in a heavily WinDoze-centric field. If I was in the life sciences, physics, or chemistry at least I would see some Macs...
csubear said:
Me too. I was just getting to the end of the posts and you guys beat me to it.
Intel & Power PC. Makes some sense, doesn't it.
Considering how all next-gen game consoles are now based around PowerPC technology (and the fact that microsoft is now creating content for a PowerPC chip), maybe Intel has realized that they need to manufacture alternates to x86 otherwise they are going to start losing out on a huge market. Apple is unhappy with IBM, and everyone else seems to be heading to IBM, seems to leave Intel and Apple on some sort of common ground.
Is there anyone here with real industry knowledge that could give us some input as to whether this may/may not happen???
Oops, They Did It Again!
Actually this is the very same path that Be, Inc. (BeOS) and NeXT, Inc. took and we know how that worked out.
Pray that this does not go down as Apple going software/iPod only.
slu said:
Actually this is the very same path that Be, Inc. (BeOS) and NeXT, Inc. took and we know how that worked out.
Pray that this does not go down as Apple going software/iPod only.
I am very surprised...
with the anti Intel bias here. Why does it matter what CPU it runs on? My Apple experience is a combination of hard- and software. Love OSX, love my multiple Macs. But for one, I don't care if it runs on 68000 series, PowerPC or Intel CPU. As long as OS/X is stable, as fast (or faster) as it is. If Apple could make it run on tomatoes or oranges inside, why would I care. Because Intel is also working with M$, doesn't necessarily mean it's has a bad product.
with the anti Intel bias here. Why does it matter what CPU it runs on? My Apple experience is a combination of hard- and software. Love OSX, love my multiple Macs. But for one, I don't care if it runs on 68000 series, PowerPC or Intel CPU. As long as OS/X is stable, as fast (or faster) as it is. If Apple could make it run on tomatoes or oranges inside, why would I care. Because Intel is also working with M$, doesn't necessarily mean it's has a bad product.
but switching to x86 is a stupid move, intel sucks their cpu's are way too hot and too slow, the only thing going for them is the pentium M and thats just because it can scale back when needed.
there is anti intel bias for a reason it's a stupid move to there architecture, it's like rumors ford reverting the model T back to steam power because it's more popular at the moment and complaining about people being steam biased.
there is anti intel bias for a reason it's a stupid move to there architecture, it's like rumors ford reverting the model T back to steam power because it's more popular at the moment and complaining about people being steam biased.
Hector said:
Seems a very intelligent point of view. It just does suck.... hmmmmm. Benchmarks for now prove that they are faster, so what is really so bad about them? By the way, just learn to spell...... dose, there, 2 in 3... makes a very good statement. Grow up!
Oh, by the way, I am not stating Intel's CPU's are superior to the PowerPC, but currently they are making much more progress in multi-core and just plain speed than IBM at this time.
Oh my god, you really can't be serious? self-respecting computer scientist? So, just because I take usabilty over technical design, I am not self-respecting? Or a computer scientist? Please explain your academic degree... (if you have one)cube said:
MarcelV said:
benchmarks do not prove intels cpu's are faster the dual 2.5GHz G5 beats the dual 3.4GHz xeon in 4/6 tests (http://www.barefeats.com/macvpc.html), if you dont like those benchmarks find me some better ones, I dare you.
multi core progress? dont make me laugh, the dual core 3.2GHz p4 gets hammered into the floor by the athlon X2 and neither of them will ship until august/september except for in one dell pc, intel cpu's are no faster they run hot and have very little future you'd be an idiot to think it would be a good idea for apple to switch to them, the only suitable cpu for apple with future would be the itanium and thats very very expensive (~$10k).
you clearly know nothing of cpu's and your on such thin ice you go so low as to attack my spelling/grammer, always a sign of weakness in opinion, some of us have better things to do than be an anal retentive over their spelling/grammer.
jayscheuerle
macrumors 68020
Whatever..
As long as it runs all my current software without modification.
As long as there's a performance boost.
As long as it's equally secure.
- then I couldn't care less about what chip is in my machine.
As long as it runs all my current software without modification.
As long as there's a performance boost.
As long as it's equally secure.
- then I couldn't care less about what chip is in my machine.
jayscheuerle said:
Maybe you do, but a lot of Mac users are more like cult users... and I'd guess that the majority are geeks... and geeks don't wan't another x86 bodge job computer... How about a modern architecture with huge parallel memory bandwidth and (to quote Steve Jobs) 'masive branch prediction logic... I don't know what that does... predicts branches...'
Nice modern clean architecture should the future, not bolt-ons 'r' us x86... which I guess still have the 640KB RAM limit... and a patch to get around it.
We already have modern clean architecture in the G5... once the technical glitches are worked out, we can move up the speed ladder. Remember... it's the whole industry that has hit a wall... Intel, AMD and IBM all have the same problem... yet IBM have increased their CPU speed more than the others in the past 18 months.
Pentium 4 speed increases have pretty much stalled also... or is everyone forgetting that?
Apple to Intel...
I have not taken the time to read all 22 or so pages posted here, but I did come across an interesting article which I'm sure has already been mentioned here, but check this out.
Is it possible that Apple would "leak" for their own benefit...say it isn't so!
I have not taken the time to read all 22 or so pages posted here, but I did come across an interesting article which I'm sure has already been mentioned here, but check this out.
Is it possible that Apple would "leak" for their own benefit...say it isn't so!
CELL Processor
Well, we already ARE using the CELL processor - it is a variant of the PowerPC family. This one is a special case of the PowerPC and the partnership producing it is IBM, Sony and Toshiba. Also notice that Microsoft is using a PowerPC base for the next XBox. For Apple to go Intel now seems backwards. There are however heat problems with the G5 series especially for portable devices. The new MiniMac could double as a foot warmer.
Well, we already ARE using the CELL processor - it is a variant of the PowerPC family. This one is a special case of the PowerPC and the partnership producing it is IBM, Sony and Toshiba. Also notice that Microsoft is using a PowerPC base for the next XBox. For Apple to go Intel now seems backwards. There are however heat problems with the G5 series especially for portable devices. The new MiniMac could double as a foot warmer.
JonMaker said:Hey, I wonder if this thread will get 1000 negatives
Why not? It would show that we have 1000 stupid Mac users.
csubear said:
Thanks! I'll look into fink! I've just installed X11 for Tiger. Of course, a graphical interface would be nice to do circuit simulation. I can always run MicroSim DesignLab in a pinch...
Hector said:
Apple has a broader product line than just the dual processer powermac. How does the whole line compare price/performance wise? Cutting $100 out of the cost of the mac mini while at the same time allowing it run their windows software in a window at full speeds makes it a lot easier for windows people to buy a macintosh and would mean a lot more for Apple's market share than getting the G5 desktop machine to 3 GHz.
Doing the same with the iBook gets them a lot more of these school contracts.
Also, it seems to me that while the PowerPC has always had architectural advantages, what continues to cause Apple problems is that building CPUs for personal computers is really a side business for IBM while it's Intel's entire reason for existing. PowerPC is supposed to be a competitive advantage for Apple, but it hasn't worked out that way - Motorolla failed to deliver time after time and now IBM is a year behind the 3GHz promise and doesn't yet have a notebook solution.
alandail said:
pretty well when you look at it matching every spec, only the single 1.8GHz PM and the 14" ibook are lagging.
the mini should be compared to micro/nano-ITX pc's and it beats the crap out of all VIA ones in cpu speed and graphical power, and is a hell of allot cheaper than the P-M ones. the ibook is plent fast for a low end laptop is small compact and dose not have a shared memory gpu which all sub £2k pc laptops do, the same go's for the 12" powerbook, also when you comapre the powerbooks to comparable pc laptops of size shape battery life cpu/gpu speed and weight nothing really compares, as for the emac, it's a solid AiO there is not really a comparible pc apart from a dell which when you configure one with near equal specs they cost about the same.
useing an x86 cpu would not magically let macs execute VB or use directX, those need API's API's which apple cant run on there OS, also to put an intel cpu in the mini would be plain silly, a pentium M would push the price up allot, a P4 would make it burst into flames and a celery is slow as crap.
All this talk of making OS X able to run Windows software isn't looking at a key fact.
Windows software GUIs have no consistency and work in a different way to OS X GUI applications. Can you imagine trying to use Expose with an MDI application?
Also, if OS X did have some Windows compatible API... would that not affect OS X software development... wouldn't developers just write Win32 software instead of Cocoa software if OS X will run Win32 applications?
IMO, making OS X able to run Windows applications natively has far too many bad points which outwiegh the good points.
Windows software GUIs have no consistency and work in a different way to OS X GUI applications. Can you imagine trying to use Expose with an MDI application?
Also, if OS X did have some Windows compatible API... would that not affect OS X software development... wouldn't developers just write Win32 software instead of Cocoa software if OS X will run Win32 applications?
IMO, making OS X able to run Windows applications natively has far too many bad points which outwiegh the good points.
A few summary points
Just a few thoughts on the earlier discussion (haven't read all 10,000 responses yet):
Just a few thoughts on the earlier discussion (haven't read all 10,000 responses yet):
It has been widely rumoured that Apple has had an Intel version of MacOS X that has been kept synced up with the latest PowerPC build for years. Given OpenStep, the core OS upon which X was built, was ported to PPC from Intel, making incremental changes to an existing x86 code base is a lot easier than starting from scratch. Jobs has been down this road before and keeping a parallel build of X for Intel would be a prudent move. The biggest changes Apple would need to do it: Optimize the heck out of it (the Intel version is unlikely to be as performance tuned as the actual commercial PPC product), provide a PPC-binary emulator like Apple did so successfully with 68K binaries, and use proprietary features that Wintel PCs don't use, like OpenFirmware and even going back to the Apple ROM. Migrating the developers to create x86 will be the hard part, but Apple has mastery in this area as well when it went from 68K to PPC. Can anyone say FAT binaries? And guess what? MacOS X supports them already.
Technically, this is feasible. Desirable? Not in my book. The jump from PPC to x86 may actually result in an overall slower system given that a slower PPC chip often has more computational power than an x86 at a faster clockspeed. Plus the G5's have amazing bandwidth, though the latest Pentiums are nothing to sneeze at in this department. Altivec is much better than MMX and its successors, too, though both are underutilized on both platforms. Proper use of FAT binaries involves context switches which are performance hogs, and a transition from optimized PPC software to optimized x86 can take years for consumer software, meaning end users won't see much of a benefit until they are running all new software. We have been done this path before, only the jump from 68K to PPC was much greater than going to x86, and that was the saving grace.
Yes, I think this is leverage against IBM which like Moto before it has hugely dropped the ball in delivering faster PPC chips. IBM hates losing money to it's competitors and this will light a fire under the semiconductor execs. Also, I think the Fab investment IBM will make to release new PPC processors for Sony and Microsoft will directly benefit ALL PPC development efforts, even for other PPC chiplines such as the G5 and the G6. Apple under Jobs is a shrewd bargainer (witness how Jobs has been bargaining with Disney under his Pixar helm). Jobs plays hard ball...and he has very hard balls. Jobs might quietly even show IBM a version of Tiger running deftly on a x86 box, just to set their balls on fire. (Throw in an alpha copy of Mac X Photoshop running natively on x86, and you'd have a powerful blow torch for those blue balls.)
Parting comment: I don't see Apple releasing an x86 box as an open system--unless it was essentially a server solution only running x86 Darwin. Apple has shown it couldn't compete with its carefully selected and licensed clones; it wouldn't have a snowball's chance in the bigger x86 pond where every computer is a commodity box with miniscule profit margins and higher risk, akin to commodity speculation on pork bellies. And NeXT, which provides the real leadership inside of Apple, has already gone down this path to a T---and with basically the same OS.
NeXT jumped from 68K to the openplatform x86 world with OpenStep and then...got bought by Apple in the bargain bin after it was clear they had failed. While MacOS X is arguably superior to its parent, OpenStep PPC, it is still basically BSD UNIX with a Mach kernel with a very similar API (Cocoa) and running a similar graphics framework (Quartz uses PDF, which is derived from PostScript; OpenStep used Display PostScript). Carbon and Classic are nothing more than transitional technologies and Apple has made it clear that it wants to wean its user base from enjoying the same level of backwards compatibility that XP users currently enjoy (I can even run 1983 DOS binaries on modern XP systems...and it is a good thing, boys and girls) by not even including MacOS 9 on new systems or even on a 5-cent CD. In a few years, those of who like software compatibility will be running Mac-on-Linux-on-Mac to fill the gap once Apple yanks Classic completely. Without the transitional stuff, you've got an OS that still closely resembles OpenStep, with a few improvements like the IOKit and a new look and feel (which many NeXTers still don't like as much as what OpenStep provided.)
But AppleNeXT has been down this path before and won't repeat the exact same, doomed to fail strategy. That doesn't mean they might not try OpenSource Darwin x86 servers or even totally locked-box x86 Mac systems, however, but as to the latter, I already gave a reason above why I find this unlikely.
Idiocy Abounds
The x86 never had a 640KB limit - it was the bonehead PC architechture that imposed that. The x86 (8086) had 20-bit address lines and 2^20 is 1MB - the memory limit was 1 Megabyte. But the PC reserved the upper 384KB for use by adapter cards.
The current x86 (586+) chips no longer use that idiotic segment+offset addressing, now having a flat memory space so no "patch" is needed. There is no 640KB limit.
That aside, x86 does still have loads of backwards-compatibility slag on it, and the PowerPC is just a better architechture. All this nonsense about it being "behind" because the x86 has faster clock speeds is just that ... utter nonesense. Intel also finds itself in the position that it must stop promoting clock speed as the measure of CPU speed.
The architecture of the machine outside of the CPU chip itself is also important to overall performance (which is partly why MHz/GHz as the sole measure is idiotic) and the Macintosh is, again, far superior to the "Industry Standard Architecture" that PCs must conform to in order to all be interoperable.
What technical glitches? There are technical reasons that need to be considered when increasing clock speeds, but it's not like there are G5 bugs that need to be fixed before working on higher clock speeds.
Intel can't take over making PPC chips unless licensed by IBM to do it, and what do you think the odds of that happening are?
Each version of Winblows grinds to a crawl and needs faster hardware and more memory etc. while each version of OS X has performed better than the last given the same hardware.
This whole thing is entirely absurd. There are plenty of reasons why Apple would be meeting with Intel that have nothing to do with CPUs for their computers, and it just goes to show how totally and incredibly stupid these allegedly knowledgeable analysts really are. Wall street and the mainstream press consistently gets tech stories all wrong; even "IT" magazines aimed at high-level CIO types (who are not tech-savvy) are constantly screwing up.
Forget it. It's not happening. Long Live PPC.
-bb
nomore said:
The x86 never had a 640KB limit - it was the bonehead PC architechture that imposed that. The x86 (8086) had 20-bit address lines and 2^20 is 1MB - the memory limit was 1 Megabyte. But the PC reserved the upper 384KB for use by adapter cards.
The current x86 (586+) chips no longer use that idiotic segment+offset addressing, now having a flat memory space so no "patch" is needed. There is no 640KB limit.
That aside, x86 does still have loads of backwards-compatibility slag on it, and the PowerPC is just a better architechture. All this nonsense about it being "behind" because the x86 has faster clock speeds is just that ... utter nonesense. Intel also finds itself in the position that it must stop promoting clock speed as the measure of CPU speed.
The architecture of the machine outside of the CPU chip itself is also important to overall performance (which is partly why MHz/GHz as the sole measure is idiotic) and the Macintosh is, again, far superior to the "Industry Standard Architecture" that PCs must conform to in order to all be interoperable.
What technical glitches? There are technical reasons that need to be considered when increasing clock speeds, but it's not like there are G5 bugs that need to be fixed before working on higher clock speeds.
Intel can't take over making PPC chips unless licensed by IBM to do it, and what do you think the odds of that happening are?
Each version of Winblows grinds to a crawl and needs faster hardware and more memory etc. while each version of OS X has performed better than the last given the same hardware.
This whole thing is entirely absurd. There are plenty of reasons why Apple would be meeting with Intel that have nothing to do with CPUs for their computers, and it just goes to show how totally and incredibly stupid these allegedly knowledgeable analysts really are. Wall street and the mainstream press consistently gets tech stories all wrong; even "IT" magazines aimed at high-level CIO types (who are not tech-savvy) are constantly screwing up.
Forget it. It's not happening. Long Live PPC.
-bb