Computex Wars: AMD vs. Intel vs. VIA - and some unexpected implications

  • At Computex this week we saw a new round of the battle of CPU manufacturers: Intel came out swinging hard, which further validated AMD’s ATI acquisition strategy. But AMD may have to make more tough decisions, including taking the company private. VIA also had a series of announcements, but I am starting to wonder if they are even relevant anymore.   

    Computex is one of the major industry shows due to its proximity to where it seems the majority of PCs and parts are manufactured. Over the last couple of years, even though Intel was clearly the larger vendor, AMD has come into the show with a significant amount of strength and Intel was on the defensive. This year, while AMD continued to show penetration into protected Intel accounts like Dell and Toshiba, Intel hit back hard with announcements that showcased its growing platform play.  

    Intel’s Series 3 chips look impressive, bracketing the PC space from value to performance machines, and promise advancement in performance and value across the segment. The company’s only discord was that a few vendors pulled back from the performance-enhancing Turbo Memory offering due to Vista compatibility problems (Intel and Microsoft still appear to be having issues). The coolest Intel-based product was the Toshiba R500, which looks like it will be the most advanced notebook on the market for some time.  

    AMD, on the other hand, trotted out Barcelona - the quad-core Intel killing part they hope will take back the performance lead. While somewhat hampered by rumors that Barcelona was delayed, the company followed up with solid examples in the server space and vendors who had completed validation.  In addition, AMD responded, partially, to Intel chipset moves with their own DTX small form factor specification. The most significant product announced around AMD was the Alienware Hangar 18 Media Center, which is the only entertainment PC I’ve ever seen that is complete from a hardware design perspective. It is expensive, but if you want to put one in your living room, this is the only one I think makes any sense.   

    VIA also tried to focus us on its own small form factor offerings, but even though VIA clearly led both Intel and AMD into this space, now that the big guys have entered, it is increasingly hard for them to be taken seriously. They simply don’t appear to have the resources to compete, but that doesn’t mean they are giving up yet. The nicest product out of VIA was the Packard Bell Nanobook, more on that further down.  

    Validating the ATI move

    Intel is more and more playing the platform card. The 3-series chipsets cover the PC space from low cost to performance and provide a strong value advantage connected directly to Intel processors. The compelling aspect to manufacturers and buyers here is both an integrated, and packaged, solution from a single vendor as well as a strong product value.

    AMD’s traditional response would have been to counter with solutions from ATI and Nvidia that were not as coordinated and less optimized. To stay competitive in terms of price, AMD would have to take a hit on its processor revenues and watch margins erode as the company struggled to maintain market share.  And, over the short term, that is likely still going to be the case until AMD can gain the benefits of a fully integrated ATI.   

    But long term, AMD gains the opportunity to respond to Intel and the company is furiously working across the two divisions to leapfrog Intel with a more advanced platform solution next year.  Intel has always been inferior to Nvidia and ATI in graphics and by acquiring this technology, AMD gained a paper advantage over Intel that could translate into a real one once they integrate the offerings.  You’ll begin to see the result late this year but the true benefits won’t show up until 2008 when we’ll know if this strategy is paying off.  

    Now, don’t get me wrong, Intel sees this coming, which is why it is rapidly ramping their own graphics efforts and pushing secondary technologies like Turbo Memory to maintain their competitive edge. AMD will be responding with a competitive solution called HyperFlash in 2008 that may actually work well with Vista. Flash storage may be one of the new battlegrounds that open up between the two firms.   

    But, had AMD not acquired ATI, they could have been at a severe competitive disadvantage by year. Recovery could have been unlikely - which is speaking directly to the necessity of the ATI move.  

    ATI/Nvidia:  Unintended consequences

    For some time, both firms were neck and neck with one gaining advantage for part of a year and the other taking it back with a vengeance. Currently, the battle at the high-end is difficult to call, but Nvidia clearly got there first (though they have had a lot of Vista issues as a result). It is these high-end parts that both create a foundation for marketing superiority and the strongest margins in the segment. Losing both is likely to continue to be painful, so this makes it increasingly difficult for AMD to maintain their relationship with Nvidia, though the collaboration with them seems to be working amazingly well given that fact.   

    But this is a major war between Intel and AMD, ranging from litigation through products and acquisitions and has had both companies in a lot of pain for months. Long term, if both Intel and AMD graphics integration strategies are successful, Nvidia may not even be left with the top-end of the discrete graphics market, which could be a rather significant problem for them.   

    Figuring out the end game here between all three players is going to be interesting, because neither Intel nor AMD headquarters (the ATI division being the exception) want Nvidia to die and would rather have them as a partner than a hard competitor.    

    Privatizing AMD

    The Gartner Group has begun to agree with the idea of privatizing AMD so that it could better compete with Intel. After similar moves at Chrysler and partially at Palm, there does seem to be a general trend in this direction because full privatization avoids the excessive costs of being a corporation in terms of reporting and financial compliance. I’ve seen estimates indicating that a private company can often be more, sometimes a lot more, profitable than a public company and vastly more agile and secretive as well.  Not to mention, because the firm doesn’t have to disclose profit or loss, the executives can operate strategically and not focus on the quarterly numbers that their incorporated peers have to focus on and are thus less likely to do foolish things like repetitive layoffs to appease financial analysts.  

    Given the benefits and the fight ahead, were I part of AMD’s executive management, I’d seriously be looking at this myself. Though I don’t do this often, I agree with Gartner that this outcome is likely and the result would put AMD in vastly better shape to take on Intel in the long term.

    Via still fighting, doing cool stuff, but less relevant

    Via, along with partners, announced the Nanobook   which, in many ways appears to be a vastly better deal for the money than the recently announced Palm Foleo. It is modular in design and is likely to be more useful as well (though also about $200 more expensive.) They have had this small form factor desktop and laptop space almost to themselves ever since Transmeta excited the segment. This was somewhat eclipsed by Intel’s $200 laptop announcement, unfortunately.  

    But, in all this time, they’ve gained little traction and now, increasingly Intel and AMD are starting to focus on this area. While I don’t know how much resource AMD has put on this, Intel has put 1200 engineers on their UMPC efforts and expects to have systems smaller and less expensive systems than this Via based offering on the market shortly.

    One other cool thing Via announced was the mobile-ITX platform, a tiny motherboard that was about the size of a credit card for phone sized x86 devices. They coupled with a partnership announcement with a telephone carrier and Gigabyte suggesting a partnering strategy compensate for the Intel advantage. If the partners are strong enough, and they aren’t initially, this strategy could work against Intel. But that’s a big “If”.  

    Wrapping Up:  Why should you care?

    We are going to get some incredibly nice mobile computers over the next 24 months. With the OEMs being focused on design and the chip companies being focused on battery life, shrinking the component size, and driving the cost out of those components, I think it will be hard to compare today’s hardware to future products. Add to this the increase in LED backlit displays, flash hard drives, and fuel cells that may actually work and we have some amazing times ahead. 

    Rob Enderle is one of the last Inquiry Analysts.  Inquiry Analysts are paid to stay up to date on current events and identify trends and either explain the trends or make suggestions, tactical and strategic, on how to best take advantage of them.  Currently he provides his services to most of the major technology and media companies.