In 1998, I moved to the Silicon Valley and accepted a position with Intel. I recall telling a good friend’s father, an old time valley engineer, that I was going to start work with the famous microprocessor company. He took a deep breath, looked side to side, and after a lengthy dramatic pause, decisively said “They’re the only game in town.” It turned out at the time he was not exactly right. Several factors conspired over the next few years to significantly erode Intel’s pre-eminence and fat margins - notably AMD’s successful execution around a low cost microprocessor, Intel’s own mis-execution of its 64 bit strategy, and countless other mis-steps that were exacerbated by the dot-com bust.

Since that time, Intel has morphed into a bloated organization, anchored down heavily by negative publicity around anti-trust law suits, and overwhelmed by one of the worst global economic cycles anyone can remember. When I look at Intel’s stock price, I cringe when I realize that my 1998 options actually had a strike price that was higher than today’s trading range. And yet, I believe now more than ever - my friend’s father is correct “Intel IS the only game in town”. What makes me think that? Well, last week Fred Pollack, retired Intel Fellow, spoke at our First Technology Conference and he made a very convincing case.

Let me summarize Fred’s analysis in one sentence “Intel only loses when AMD has a competitive CPU microarchitecture and when it is less than 6 months behind Intel on process technology.” Fred basically argued that the conditions for AMD to achieve these milestones were highly unlikely and depend upon some major screw-ups from Intel.

Now Fred knows what he is talking about - after all he was Chief Architect of the original Pentium Pro chip and is up to his neck in industry accolades. But, let’s dive into the merits of his argument (and not his resume). Consider three segments of the market: (1) Server (2) Desktop and (3) Mobile. In the server segment, Fred expects Intel’s Nehalm-EP/EX to eliminate AMD’s DP platform advantage. On the desktop, Pollack noted that several Intel processor variants outperform AMD’s Phenom II (Shanghai) when run at the same clockspeed. And, then on the notebook side - he observed that Intel already has offerings ahead on the process technology curve. He buttressed that last point by noting that Intel was a year ahead on 32nm offerings with product coming out of Intel’s fabs in Q4 of this year.

In other words, Fred believes that it is Intel’s game to lose and that they aren’t going to let that happen again. Specifically, Intel is employing a “Tick/Tock” design strategy where microarchitecture designs are being alternated between two teams (in Haifa, Israel and Hillsboro, Oregon). And, so if you think there is going to be a PC/Mobile boom in 2010 (and 99% of the world thinks that) - there is only one table open at the casino and Intel is the dealer. Ladies and gentleman, exchange your cash for (micro)chips and let the game begin!

Tweet This Post   

Tags: AMD, Intel

This entry was posted on Thursday, June 11th, 2009 at 9:50 pm and is filed under Semiconductor, Technology, Unni Narayanan. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.