Between AMD and Intel – the Hard Plaice, Nvidia is certain to find itself more than a little squeezed. Our information is that Intel has far from given up on its Larrabee project and of course AMD is already sampling its Fusion chips, courtesy of GloFo (GlobalFoundries) and TSMC.
IBM, of course, is a very big company and has its own Microelectronics division. It also has a serious grudge against Hard Plaice (Intel) and has had so since the Robert Noyce Building in Boca Raton – a joint collaboration between it and the Hard Plaice – fell apart in the early 1990s.
Together they produced quite a good few things but that’s all forgotten now. IBM uses X86 processors and of course its Microelectronics division based in East Fishkill still has 200 AMD engineers learning all they can about stuff.
But since AMD span off its fabrication plant facilities suddenly the little X86 player isn’t as much of an ally as it was. GloFo is something of a competitor, IBM still hates Intel and Nvidia really has got some rather nice technology, although it is in something of a cul-de-sac because of Hard Plaice’s Larrabee and AMD.
OK – but what else would IBM get out of snapping up Nvidia apart from exercising its undoubted spite against the Hard Plaice and adopting some good technology?
The truth is that IBM Microelectronics comes out with some fantastic ideas but it is very press unfriendly in the 21st century. Sam Palmisano’s mind is on software and services and it’s hard to dissuade him from the Lou Gerstner mindset he’s stuck in.
To most observers, Nvidia’s foray into high end computing appears to be at the expense of its place in the graphics world churn. But Nvidia’s Jen-Hsun Huang is a very clever and one pointed man. And it could just be that he’s been thinking far ahead of the guys at Intel and AMD. There’s a Venn diagram here, but it’s a very complicated one indeed.