Nvidia has rebuffed speculation that its next-generation Fermi GPU is targeted primarily at the high-performance computing market (HPC) market.
“Such speculation is insane,” Nvidia’s Sanford Russell told TG Daily. “We have made Fermi modular and flexible to address all possible markets.”
Nvidia CEO Jen-Hsun Huang expressed similar sentiments during a recent Q&A session for reporters and analysts. According to Jen-Hsun, Fermi architecture represents the “foundation” of the world’s first computational GPUs – which will include a range of products, including GeForce, Quadro and Tesla.
“GPU computing has reached an important tipping point. GPU computing is for real,” said Jen-Hsun. “It is not just used by scientists or hobbyists, but by mainstream developers and companies that have adopted a parallel computing model. GPUs are everywhwere – in notebooks, PCs, servers and cloud computing systems.”
Jen-Hsun also noted that Nvidia was focused on helping the industry advance with Fermi and had little interest in replacing the CPU.
Nevertheless, chip expert David Kanter has hypothesized that Nvidia intends for its graphics processors to become more like CPUs. Fermi – said Kanter – is aimed at the HPC market because it is “leaving behind many of the more charming, but baroque, elements of GPU architecture.”