Editors Note: From time to time, SemiAccurate will be republishing some older articles by its authors, some with additional commentary, updates and information. We are mainly reprinting some of the oft referenced articles that originally appeared on the Inquirer. Some will have added content, but all will be re-edited from the originals as per contractual obligations. You may see some slight differences between the two versions.
This article has had some of the original links removed, and was published on Monday, October 23, 2006 at 2:42PM.
ONE HAS TO WONDER if the dealings with Stexar will be seen as one of the most damaging missteps of modern semiconductor design or the greatest victory. No, not referring to AMD, Intel or Nvidia but AMD, Intel and Nvidia all could be mortally wounded by the actions they took around the company.
A little history for you, a bunch of really bright Intel folk decided that the house of silicon was no longer strong enough for them, the big bad wolf could come and blow it down. So, they left and built their own house, and from what we hear, it was a very nice house filled with amazingly smart and talented people. There was probably a Starbucks nearby, mainly because they are everywhere.
The projects they were working on were closely tied to AMD, specifically Fred Weber at AMD. They were working on a set to box type chip with multimedia/vectorish instructions, basically an x86 DSP, or at least that is what the rumors were.
They also did a bunch of engineering work for AMD, and all reports said that they did some really inventive and creative things. They also were in line for a lot of money. When the AMD politics went all soap-opera-like around several people, Weber being one of them, Stexar may have lost it’s champion.
The possible mistake was not simply buying Stexar then and there. AMD had good reason to do this, including incorporating Stexar New Instructions into the core K8 CPUs. In any case, it didn’t happen, and to make a long story short, AMD bought ATI and had little need for Stexar. Stexar the company didn’t last very long from there.
If that was AMD’s mistake, what was Intel’s, other than letting the engineers go in the first place? Well it is rumored that some people tried to go back to Intel , and were made some promises by high level Intel execs. This could have resulted in some very interesting things coming out of Intel in a few years.
Sadly, we are told internal politics got in the way, and at least one person had his legs kicked out from under him, and the promises turned out not to be, or at least were rescinded. The Stexar folk (Stexites? Stexy ones?) went off to greener pastures from there, and slipped through Intel’s fingers.
Where does a team of engineers of this caliber go when they are not really welcome at AMD or Intel? Easy, greener pastures was not a euphemism, they went to Nvidia. What are they doing at Nvidia? Well duh, they are making a CPU.
The ATI buy put NV is a really tough spot, they either have to make a CPU, change their entire product line, or wither and die. Given the egos involved, the first one is the only real option, not that they haven’t threatened it in the past. I even think they have a better than average change of success in their venture, and it may be a very interesting product when it rolls out.
Why is this a problem for Nvidia, it sounds like a CPU could be a very good thing for them? The problem is that they are now in direct competition with the two people who make the things their products plug into, AMD and Intel. You can almost hear the gun turrets swiveling toward Santa Clara.
This could be a fatal mistake for Nvidia, or it could be their greatest triumph. In any event, there are some non-trivial hurdles to clear before they can make a CPU, but that is a story for tomorrow.S|A
NOTE: Part 2 can be found here.
Latest posts by Charlie Demerjian (see all)
- ARM adds CMN-600 interconnect and DMC-620 memory controllers - Sep 27, 2016
- Globalfoundries 7nm process isn’t even close to the name - Sep 26, 2016
- ARM upgrades realtime offerings to v8-R and adds Cortex-R52 - Sep 21, 2016
- Everspin and Globalfoundries team up for embedded ST-MRAM - Sep 15, 2016
- Intel’s Xpoint is pretty much broken - Sep 12, 2016