At the beginning of March I pointed to a story in which Michael Hara, SVP of Investor Relations for NVIDIA, was quoted saying that it would “make sense” for NVIDIA to start developing its own x86 platform. Odd for an investory relations guy to be saying that, but there you go.
This week Michael Feldman writes at HPCwire that NVIDIA is now contradicting that earlier statement.
But in the TechRadar article this week, NVIDIA’s Perez [NVIDIA's director of public relations] contradicted Hara’s earlier statement saying “We didn’t need x86 for our first 15 years and we won’t need it for our next 15.”
Again, pretty strong language for a non-technical leader at a technology company. I’m not buying in either way until I hear it from someone with an engineering degree, because I think NVIDIA really does have a problem here.
Michael’s larger story is actually about the medium term threat that NVIDIA faces as a maker of discrete components in someone else’s system
At least on the client side of the GPU computing business, one visible threat NVIDIA-style GPU computing is the CPU-GPU hybrid processor model envisioned by both Intel and AMD. Although the first products won’t have an HPC play, all signs are pointing toward an eventual convergence of CPUs and GPUs below the chipset level. If this turns out to be the case, future general-purpose microprocessors for high-end computing may have GPU-like smarts already built in, making discrete parts somewhat redundant.
NVIDIA is full of really smart guys, but lots of smart guy companies have gotten complacent with their position in the marketplace only to be made irrelevant by a technology zig (or zag) they didn’t see coming.