Apple - Intel talks just a herring?
Summary: Earlier this week, the Wall Street Journal reported that Apple was in talks with Intel regarding usage of the latter's microprocessors. Today, Apple relies strictly on PowerPC chips from IBM but has had difficulty keeping pace with Intel-based competitors, particularly on the notebook front where the company has complained that IBM can't deliver a decent mobile offering.
Earlier this week, the Wall Street Journal reported that Apple was in talks with Intel regarding usage of the latter's microprocessors. Today, Apple relies strictly on PowerPC chips from IBM but has had difficulty keeping pace with Intel-based competitors, particularly on the notebook front where the company has complained that IBM can't deliver a decent mobile offering. Most believe that if Apple moved to Intel, the company would be one baby step away from the birth of a Mac clone industry whether Apple endorses it or not. Shortly after the news surfaced, News.com's Michael Kanellos offered his analysis of the risks that would be involved in such a move (including the cloning issue/problem). But I liked the way this report by Forbes really got a gander at Apple's soft-white underbelly. Both analyses quote Instat/MDR chip analyst Kevin Krewell. Both also talk about how Apple is like IBM's neglected stepchild and that it could be using discussions with Intel as leverage to get more attention from IBM. Both are good middle-of-the-road takes on this on-again off-again rumor (it's certainly not a romance yet). But for something totally left wing, read why our own Paul Murphy thinks Intel is the wrong move if Apple really wants to switch. Says The Murph, "For the record, my belief is that Apple’s best route out of its strategic CPU dilemma is UltraSPARC."
Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.
Talkback
Ultrasparc?
The best explanation for these rumors is that Apple just wants a little attention from IBM. However, Apple would benefit from Intel mobile technology with the Pentium M for use in their own laptops.
Sorry George, you're dead wrong
<P>
All of Sun's real research muscle -and there's a lot of it- is behind throughout computing. Even the engineers seconded to AMD in 2001/2 (at least those still with Sun) are now working on Rock and Niagara. Stay tuned, these are world beaters - orders of magnitude more powerful than Xeon.
I don't have much confidence in Sun
As for your history lesson on CPUs, I tend to disagree. AMD never had a good product until the arrival of the Athlon. Before the Athlon, their CPUs were horrible for floating point performance. Back then, AMD lost a lot of credibility with their processor rating scheme where they exaggerated their performance rating numbers and recently they tried to do that again with their Athlon 3200+ speed ratings.
In the current single core race, the Intel CPUs have a slight edge on AMD, but the AMD dual-cores are killing the dual-core Intel chips in performance while utilizing 1/2 the power. Since the coming of the Athlon, it?s pretty much been a seesaw battle between AMD and Intel for the performance crown. While you're right about AMD's 64 bit strategy forcing Intel to embrace 64 bit extensions, I think you give too much credit to Sun.
Yes and no
Can you say VIKING? Sun has had many missteps in the CPU race (Viking was a replacement for some other previous chip that didn't work). UltraSparc IV is the latest (Fujitsu to the rescue!).
OTOH, Sun has had breakthroughs also. The original SPARC put them in the workstation lead for good (bye, bye Apollo), and their "open" licensing for SPARC DID net them Fujitsu (the terms of the license are darn good). I think SPARC has helped BOTH Fujitsu and Sun.
Sun has a good plan and great vision, the only thing in their way is . . . Murphy's Law.
Suffice to say that the Itanium is the tortoise in the processor race. EPIC is just TOO smart of an idea to just die out. As I said before, as the end of Moore's Law draws near (>40nm processes will not work), and Silicon real estate becomes static in size (finite), the EPIC architecture offers MORE than CISC/RISC (just get rid of the x86 "partial core").
As for APPLE, my own suggestion is to partner with MIPS. Maybe get MIPS to get in on that SPARC license and produce cheap UltraSparcs. Or maybe have MIPS partner with ARM and build their next generation chips. Hard to say, what kind of fabs does MIPS have?
the Apple/Intel rumor is just that.
1) convert the Mach kernel to x86 (could be done)
2) convert all the applications to x86 (could be done but would be very expensive) This would require support from every software developement house that writes for MacOS (a little software company in Redmond, WA comes to mind).
The chip architectures are different, PowerPC is bi-endian while x86 is littleendian (iirc, could be bigendian...).
Just wont happen
Apple COULD use the Transmeta chip - and thus no recompiling - but Transmeta is getting out of the chip business. The ONLY other chip I know that has "downloadable personalities", and could support the PPC op codes is . . . Alpha. Gee, WHO can make THAT?
Hogwash
difference.
NeXT Step - Father of OS X ran on Intel HW
But also, it's not about clones if parts of the computer become Intel chips instead of motorola. Microsoft xBOX360 and IBM are 8000 pound gorillas. Apple would be smart to leverage other providers for technology chips inside like video, wimax wireless and so forth.