AMD wants to worm into Apple

AMD wants to worm into Apple

Summary: AMD believes that Apple will eventually buy its CPUs in order to offer an alternative to Intel.

SHARE:
TOPICS: Processors
15

Until AMD comes out with something that can come close to the Core 2 Duo, why should Apple bother? AMD logoSome bold words from AMD yesterday:

Hector Ruiz, chief executive officer of Advanced Micro Devices Inc., said Apple Computer Inc. will eventually buy its semiconductors to provide an alternative to Intel Corp. chips.

Apple started selling computers based on Intel microprocessors in February, and its exclusive use of them will drive it toward using Advanced Micro parts, Ruiz said in a speech yesterday at the Commonwealth Club in San Francisco.

"Everybody wants choice,'' Ruiz said in the speech, in which he added that larger rival Intel's practices have stifled the PC industry's growth. "Knowing Apple, why would they want to be held hostage like everyone else has been?''

I just don't see this happening, either now or any time in the next 3 - 5 years.  There's just not enough that separates AMD and Intel to make having two product lines competing against each other worth the fuss. 

To be honest, I really don't think that Apple users want a choice, especially a subtle technical choice.  And I can't see Apple being able to explain to their customers what the difference between an AMD CPU and an Intel CPU.  To be honest, it took Apple years to make the leap to Intel even though the only place where PowerPC microprocessors were faster than Intel and AMD was in Apple's own mind and in their commercials.

And anyway, until AMD comes out with something that can come close to the Core 2 Duo, why should Apple bother? 

Topic: Processors

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

15 comments
Log in or register to join the discussion
  • Hector is starting to smell

    "..Knowing Apple, why would they want to be held hostage like everyone else has been?.."

    Just is what is wrong with that dude? He is starting to sound like that clown from Venezuela, Hugo Chavez.
    Prognosticator
  • Reason is simple: INTEL will BK in 2008

    dude, you need to have some foresight, companies plan ahead, AMD will launch K8L in mid 2007, which will be 60% faster than Core 2 and push Intel back by 3 years. If you don't go AMD now, it will be too late at K8L time. Understand?
    sharikou
  • AMD and Apple

    Perhaps you have it backwards? Why on earth would Apple NOT consider AMD? Apple learned to its dismay about single source. We Mac users looked forward to what the dual G5's would be able to accomplish, something akin to having the original Cray "supercomputer" on our desktop. But neither M or I could resolve heat or power consumption problems, which just made a laptop impossible. If AMD has a good product WHY NOT use it? Wintel is still very much alive.
    One guarantee in life is that things will continue to change.
    gregbarker
    • Intel probably gave Apple a sweetheart deal

      Intel is right down the street from Apple. Jobs and Intel CEO are
      friends. Apple is different from your typical Wintel company. For
      Intel to win the Apple contractwas a feather in Intel's cap, for
      prestige more than anything else. Also it was a good thing for
      Apple as well. They finally got a processor company thats main
      business is making processors. IBM made great processors but
      not a full line and development was spotty, like the promise of 3
      ghz in one year which didn't materialize.

      I wouldn't doubt that someday AMD and Apple do business.
      Especially since you can see AMD courting them, already. You
      might even see Mac clones again. The whole industry is
      changing and partners are changing as well.
      MacGeek2121
  • The PowerPC really WAS faster.

    [i]To be honest, it took Apple years to make the leap to Intel even though the only place where PowerPC microprocessors were faster than Intel and AMD was in Apple's own mind and in their commercials.[/i]

    Oh, and on a few other places, like most unbiased benchmarks (the infamous Adobe AfterEffects one was [i]one[/i] one of these: more on that later), and in the Top 500 Supercomputers Lists of the time where most of the top supercomputers used Power chips, with the #1 most powerful computer on the planet using PowerPC (not Power4 or Power5). Oh, and the fastest AMD chip-based system was barely in the Top 20, and the fastest Intel was way below that. Oh, and the PowerPC systems consistently got higher scores with fewer chips running at slower clock speeds than the Intel chips and, to a lesser extent, the AMD ones.

    This has changed recently with the Core chips. Intel finally did it right, after all this time. Until then, in [i]every single case[/i], pairing any Intel chip against its nearest equivalent Motorolla 68x or PowerPC chip at the same clock speeds (or even with the Motorolla chips running at half the clock speed), the Intel lost every time, all the way back to their respective 8-bit CPUs up to and including the Pentium 4 Xeons compared to the PowerPC G4s, or Itanium²s compared to the PowerPC G5s. The AMD came closer to parity with the PowerPC, but never quite matched it.

    As for the AfterEffects benchmark used by Charlie White, well, I have bad knees and am middle-aged and out-of-shape, but even I could win a short race against an Olympic sprinter if I was allowed to use both of my legs and the Olympic sprinter had to hop on one foot. That&rsquo;s effectively what the AfterEffects benchmark did. The Mac version of AfterEffects would use only one CPU no matter how many were installed in the machine, and Charlie compared dual-CPU Macs vs. dual-CPU Intel machines. Of <I>course</i> the Macs lost!

    When a dual-CPU G4 was benched against an absolutely identical Mac with only one G4, AfterEffects ran nearly as fast on the single-CPU machine, with the slight (single-digit percentage) speed advantage on the dual-G4 machine totally explainable by some of the OS itself and other background tasks running on the other CPU. <I>None</I> of AfterEffects was even <i>acknowledging the existence</I> of the second G4.

    The guy who noticed this also provided a way to fool AfterEffects into using the second CPU (basic technique: set up network rendering on the dual-CPU machine by itself, with the second CPU as a render node). With that done, the dual G4 machine nicely trounced the dual Intel machine that was running at a higher clock speed, as would be expected by anyone who actually knows those CPU families.

    The Intel Netbust (spelling intentional) architecure was, well, a bust, even by Intel&rsquo;s standards. Even the 386, which was itself slower than the Motorolla 68030 (remember the nearest-equivalent-chip-to-chip comparisons over the years that I mentioned earlier?), was [i]way[/i] faster than the Netbust 486s running at the same clock speed. The Pentium 4s only wound up faster overall by being clocked at insanely high speeds (but requiring insanely deep pipelines, with resulting slowdowns, to achieve those clock speeds). Even a rider lawn mower could outrun a sports car if you ran its engine at fast enough RPMs (assuming the thing didn&rsquo;t explode or otherwise fall apart). But nobody would say that the lawn mower engine was more [i]efficient![/i]

    Intel&rsquo;s most efficient chips for the past few years have been the one that eschewed the Netbust architecture and gone back to the 386 architecture: the Pentium M used in Centrino laptops would easily keep pace with a Pentium 4 clocked at a [i]full gigahertz faster![/i] A 1.8GHz Pentium M would match a 2.8GHz Pentium 4, for instance!

    The new Core chips likewise use and improve on the 386 architecture, improving what made it so efficient (compared to Netbust, anyway). They have finally beaten AMD, and I do believe that they may well be the first Intel chips to actually be more efficient than the current PowerPC chips (there are no real equivalent PowerPC chips to compare them to, yet). As I said, Intel finally did it right.
    Joel R
    • Anyone at ZDnet wanna 'splain to me where my paragraphs went!?

      This happens on occasion: all my paragraph breaks go bye-bye!

      Testing: this is a separate paragraph.

      What's the deal here!? Can you guys not even write a decent blog commenting system!? And we're supposed to trust you on reporting on the entire industry!?
      Joel R
  • The PowerPC really WAS faster (attempted repost)

    [i]To be honest, it took Apple years to make the leap to Intel even though the only place where PowerPC microprocessors were faster than Intel and AMD was in Apple's own mind and in their commercials.[/i]

    Oh, and on a few other places, like most unbiased benchmarks (the infamous Adobe AfterEffects one was [i]one[/i] one of these: more on that later), and in the Top 500 Supercomputers Lists of the time where most of the top supercomputers used Power chips, with the #1 most powerful computer on the planet using PowerPC (not Power4 or Power5). Oh, and the fastest AMD chip-based system was barely in the Top 20, and the fastest Intel was way below that. Oh, and the PowerPC systems consistently got higher scores with fewer chips running at slower clock speeds than the Intel chips and, to a lesser extent, the AMD ones.

    This has changed recently with the Core chips. Intel finally did it right, after all this time. Until then, in [i]every single case[/i], pairing any Intel chip against its nearest equivalent Motorolla 68x or PowerPC chip at the same clock speeds (or even with the Motorolla chips running at half the clock speed), the Intel lost every time, all the way back to their respective 8-bit CPUs up to and including the Pentium 4 Xeons compared to the PowerPC G4s, or Itanium&sup2;s compared to the PowerPC G5s. The AMD came closer to parity with the PowerPC, but never quite matched it.

    As for the AfterEffects benchmark used by Charlie White, well, I have bad knees and am middle-aged and out-of-shape, but even I could win a short race against an Olympic sprinter if I was allowed to use both of my legs and the Olympic sprinter had to hop on one foot. That&rsquo;s effectively what the AfterEffects benchmark did. The Mac version of AfterEffects would use only one CPU no matter how many were installed in the machine, and Charlie compared dual-CPU Macs vs. dual-CPU Intel machines. Of <I>course</i> the Macs lost!

    When a dual-CPU G4 was benched against an absolutely identical Mac with only one G4, AfterEffects ran nearly as fast on the single-CPU machine, with the slight (single-digit percentage) speed advantage on the dual-G4 machine totally explainable by some of the OS itself and other background tasks running on the other CPU. <I>None</I> of AfterEffects was even <i>acknowledging the existence</I> of the second G4.

    The guy who noticed this also provided a way to fool AfterEffects into using the second CPU (basic technique: set up network rendering on the dual-CPU machine by itself, with the second CPU as a render node). With that done, the dual G4 machine nicely trounced the dual Intel machine that was running at a higher clock speed, as would be expected by anyone who actually knows those CPU families.

    The Intel Netbust (spelling intentional) architecure was, well, a bust, even by Intel&rsquo;s standards. Even the 386, which was itself slower than the Motorolla 68030 (remember the nearest-equivalent-chip-to-chip comparisons over the years that I mentioned earlier?), was [i]way[/i] faster than the Netbust 486s running at the same clock speed. The Pentium 4s only wound up faster overall by being clocked at insanely high speeds (but requiring insanely deep pipelines, with resulting slowdowns, to achieve those clock speeds). Even a rider lawn mower could outrun a sports car if you ran its engine at fast enough RPMs (assuming the thing didn&rsquo;t explode or otherwise fall apart). But nobody would say that the lawn mower engine was more [i]efficient![/i]

    Intel&rsquo;s most efficient chips for the past few years have been the one that eschewed the Netbust architecture and gone back to the 386 architecture: the Pentium M used in Centrino laptops would easily keep pace with a Pentium 4 clocked at a [i]full gigahertz faster![/i] A 1.8GHz Pentium M would match a 2.8GHz Pentium 4, for instance!

    The new Core chips likewise use and improve on the 386 architecture, improving what made it so efficient (compared to Netbust, anyway). They have finally beaten AMD, and I do believe that they may well be the first Intel chips to actually be more efficient than the current PowerPC chips (there are no real equivalent PowerPC chips to compare them to, yet). As I said, Intel finally did it right.
    Joel R
    • Argh! It did it aGAIN!!

      Is it because I used your square-bracket pseudo-HTML tags?

      Testing: [b]bold[/b] and [i]italics[/i].

      Testing HTML Named Entities: I have &frac12; a mind to never patronize anything from ZDnet ever again. Just putting my 2&cent; in.

      There should be four paragraphs in this post.
      Joel R
    • The PowerPC really WAS faster (attempted repost)

      Maybe if you'd get a real computer and quit living on the old dinosaur reputation of Apple, you wouldn't be reliving "ground hog day" and keep repeating yourself over and over. Dude, you've made like 4 posts and it's all SSDD (same s**t different day) actually, its really like SSSS (same s**t same s**t) BORING!! My experience with the Mac has been extremely lackluster in both graphics labs and laptops. My 4 year old IBM T-Pad kicked ass with several different graphics programs running simultaniously from Corel Draw, Adobe Photoshop, Jasce Paintshop, (yes, there are things I prefer in each program over the others)as well as several IM programs to keep in contact with associates during my projects, plus antivirus and firewall. Co workers next to me running a Mac G4, or Mac I-Book with a single Adobe program open would often crash and lose work. Now there's quality!! I just can't wait for Big Blue and little brother Lenovo to catch up with that kind of reliability. I'll take my T-Pad anyday over Apple. by the way, on average, my T-Pad runs about 60+ processes running with OS stuff, network, applications and fun stuff, some of which are very processor and memory intensive. I have yet to have a crash and lose work. NEENER! NEENER! NEENER!
      carrie.girl@...
      • Well, I noticed you have no paragraphs either.

        And, for the record, I posted these posts using a Wintel machine. Pentium 4 Nutbust, natch. And I was originally only going to post one post, but ZDNet Forums software messed up (running on a Windows server, most likely). You replied to my REpost.

        I did note that you said that your T-Pad runs lots of applications, "plus antivirus." Well, Macs don't NEED antivirus nearly as much, if at all. NEENER! NEENER! NEENER!

        Seriously, I use both, daily. Our business depends on both. They each have their advantages.

        But, my post was addressing one simple fact, and that was about the CPUs, not the computers built on them, nor the operating systems, nor the apps. PowerPCs were faster than Pentium 4s. Period. Nothing you said has anything to do with that. Can you refute what I actually said?
        Joel R
      • Carrie? Ya moron.

        There is NO REASON WHATSOEVER that ANY website should be
        difficult with ANY browser or operating system that is standards
        compliant. In case you (and ZDNet and many others) forgot, the
        internet is SUPPOSED to be platform agnostic. It's Microsoft that
        has wrecked the health of the internet, DICTATING its own crap
        and not playing along with others (like Sun's Java, for instance).

        I tell ya, the internet was doing juuuust fine before Microsloth
        woke up and smelled the coffee. Then, by hook or by crook, they
        attempted to reinvent it in their own image... and WRECKED it.
        dropzone@...
        • To be honest ...

          ... I've used this site with all three of the major browsers and never had a problem.
          Adrian Kingsley-Hughes
  • Intel BK?

    To the deluded person who claims Intel will be bankrupt in 2008, I offer this link:

    http://articles.moneycentral.msn.com/Investing/SimpleStrategies/SpotBankruptcyCandidatesAndSteerClear.aspx
    kukito
    • Try "Reply to Message" instead of "Reply to Story" next time.

      So that your Reply would be listed under the message that you&rsquo;re Replying to.

      Anyway, you are correct that Intel isn&rsquo;t going bankrupt any time soon; not now that they&rsquo;ve gotten their act together and are finally competing on quality, performance, and value, and not just their name (&ldquo;Intel Inside&rdquo;).

      They do take AMD seriously, as a serious threat, even when it had a much smaller market share. They do still remember Zilog. Remember them? The company that almost destroyed Intel, beating them soundly at their own game back in the 1980s? The chip (Z-80) that Radio Shack chose over Intel (8080, 8085) for most of the TRS-80 line? The one that 90+% of CP/M machines (the dominant OS of the day) wound up using?

      If Zilog had been as dishonest as Intel was at the time, or Intel had been as honest as Zilog, Intel would be nothing more than a chapter in computer industry history books now. When IBM came knocking on the chipmakers&rsquo;s doors, saying that they wanted a cheap 16-bit internal, 8-bit external CPU for a personal computer that they were going to build just to see if it was worth competing against Apple, both Zilog and Intel had such chips in development (Z-8002 and 8088, respectively), but neither were ready, though the non-crippled full 16-bit internal and external versions of both chips were ready (Z-8001 and 8086, respectively). The Z-8001 was frankly superior to the 8086, just as the Z-80 was to the 8080 and 8085. But, Zilog told IBM the truth, that the Z-8002 wasn&rsquo;t quite ready yet, and had some minor problems. Intel lied, and told IBM that the 8088 was ready to go, though in reality it had similar minor problems.

      Were it not for that lie, Intel would&rsquo;ve been dead within months. The IBM PC contract saved them, and the entire computer industry was changed forever by that lie.
      Joel R
  • The PowerPC really WAS faster (Reply)

    It was faster! Plain and simple better architechture. Better bit execution because of better architechture. A BMW is going to run better than a Yugo. Even if the Yugo beet the BMW down the line. The ride their is no where near as nice.
    osbd