« Rich and Poor |
Main
| Saved by the Cable Guy »
October 25, 2005
Intel Losing Power
Posted by Dana Blankenhorn
The headline is a pun.
Intel's new chip road map, announced today, implements a commitment to lower-power processors the company announced earlier.
But in some ways the headline is not a pun. Intel really is losing some of its power, the power to define markets, the power to control them. Much of this power is being lost to AMD.
But Intel is willing to lose this power in order to rationalize its chip line and prepare for the real chip end game, which begins once a Chinese chip company decides to abandon the Intel road map and go out on its own.
Seen from that angle, everything Intel is doing looks prudent. Pruning the line, emphasizing lower-power in higher-speed processors, it's all very reasonable based on Moore's Second Law, the fact that, as designs become more complex, they become exponentially more difficult to implement.
Which leads to a question about Otellini. I know he's a high-bandwidth guy, very smart. I also know he's the first non-engineer to helm Intel. But is it time, perhaps, to name a Chief Engineer with real power, and a real public profile, who can defend the many moves like this sure to come, both in English and in a language that other engineers will respect?
Just asking.
Comments (5)
+ TrackBacks (0) | Category: Business Strategy | Consulting | Moore's Lore | Semiconductors
POST A COMMENT
- RELATED ENTRIES
- › The Legend of Dennis Hayes
- › Evolution Changes Its Mind (Again)
- › Welcome to 1966
- › What Must Craigslist Do?
- › No Such Thing as Free WiFi
- › The Internet As A Political Issue
- › Google Images Ruled Illegal
- › Fall of Radio Shack
Dana Blankenhorn has been a business journalist for over 25 years and has covered the online world professionally since 1985. He founded the "Interactive Age Daily" for CMP Media, and has written for the Chicago Tribune, Advertising Age, and dozens of other publications over the years.
1. Jim on October 25, 2005 03:08 PM writes...
This has everything to do with AMD and nothing to do with China. AMD has made huge inroads in the server market, particularly in multiprocessor architectures. Intel has been forced to address AMD's strengths...and is cancelling one chip and pushing out another to respond to AMD. The only way this is a reaction to China is that if Intel succeeds in squashing AMD, it will be in better shape to battle China in 10 years.
Permalink to Comment2. bbbl67 on October 25, 2005 04:05 PM writes...
Yeah, gotta agree with Jim, not sure how you come up with a relationship to China. All of this activity is only about how Intel is going to battle AMD. China doesn't produce x86 processors, so it doesn't factor into the battle; and in fact, ironically it was only yesterday that AMD licensed the Chinese government to distribute and produce some of AMD's own low-cost designs of x86; so AMD feels comfortable enough distributing its designs to the Chinese. So none of these strategic moves by Intel have any significance to China. Yeah sure, there may be a long term impact from the Chinese or the Indians, or whatever, but those are off in the background. Way off in the background.
The only thing that's relevant here is that Intel is in a major panic state right now. It finds itself saddled with several technologies that are several years behind AMD's, and so it doesn't know where to start answering AMD from. Yes, that's right several *years* behind AMD: the technology it's thinking of introducing in 2007 or 2008 are things that AMD already introduced back in 2003! The only advantage it's still got over AMD is in manufacturing technology, as it can usually move to the next miniaturization stage before AMD can. But AMD's disadvantage against Intel is not nearly as great as Intel's disadvantages against AMD.
Permalink to Comment3. Matt on October 25, 2005 04:08 PM writes...
Was it coincidence that Intel started development of the Itanium the same year that AMD won the cross license case?
NO, the purpose of the IA-64 platform was to replace the X86 platform, to perpetuate Intel’s monopoly in the PC market.
We are all supposed to be using Itaniums by now, Intel stopped true development on X86 back in the late 90's - why innovate on a dying platform?
They got caught by the K7, and slapped together the P4, with its illogical, extremely long pipeline, whose sole purpose was marketing ("megahertz sells").
But the main problem that Intel had with the K7, is that is gave a viable alternative to Intel products. Since IA-64 required a massive amount effort to switch over too, Intel was relying on its status as a monopoly, to say "you better go to IA-64 - because we aren't going to continue down the x86 path" - and they didn't.
The K7, and then the K8 allowed the industry to reply "that's OK - we don't need IA-64 thank you".
Intel was caught in a catch 22, develop better X86, and IA-64 is dead, don't develop better x86 and we fall behind AMD.
They chose to compromise, modify X86 enough to keep it viable. But in order to do this they added kluge after kluge. First was the outrageously long pipeline - this created great clockspeeds, at the cost of performance. Then hyper threading a kluge to try and recover from the pipeline kluge. Now they are left with an architecture that is unmaintainable, at least in the sense of creating something new that will be competitive.
So they are left with only 1 alternative, go back to the P3 (Pentium M is based on P3), and start innovating from there. And that's exactly what they are doing.
So were does that leave Intel. First IA-64 is in its death throes, and they are at least 2-3 years behind in X86.
Permalink to Comment4. Paul on October 25, 2005 04:14 PM writes...
Intel is in trouble and very few people are willing to admit that since those involved with the capital markets stand to lose if Intel shares drop in value. Analysts tend not to directly state that Intel is losing the race with AMD.
Xerox is another company that once owned a market. The competition chipped away at them with better products and lower prices. History is repeating itself. Intel basically fell asleep. I think the technical term for this is "Fat, Dumb and Happy."
The other issue is that with AMD's lawsuit against Intel, Intel is being very careful about how they negotiate with their customers. The lawsuit will take a long time to play out in the courts, but it's most immediate effect is to lessen Intel's anti-competitive behavior
Permalink to Comment5. Stan on October 25, 2005 06:35 PM writes...
China?? That's just another excuse for Intel falling behind. Intel pays too much attention to trying to maintain the status quo (along with its stock price) and not enough attention trying to provide its end customers more bang for the buck.
That is intel's one and otellini problem.
Permalink to Comment