Neocloud Executives Say Nvidia Penalizes AI Clouds Using AMD or TPU Chips
According to interviews with artificial intelligence startup executives, Nvidia is applying pressure on "neocloud" operators to exclusively use its hardware, allegedly penalizing those that deploy rival chips or networking gear. The retaliation reportedly includes withholding early allocations of Nvidia’s latest graphics processing units and discouraging or withholding support for the startups’ initial public offerings or venture capital fundraising rounds.
This high-pressure strategy does not appear to extend to hyperscale cloud providers, which possess sufficient purchasing power to negotiate independently. Instead, smaller AI cloud builders are feeling the impact firsthand, prompting several executives to consider quietly designing their infrastructure around alternative hardware like AMD graphics processors or Google tensor processing units. The findings were reported by SemiAnalysis, which specializes in semiconductor and data center industry analysis.
From the sources (3 posts)
@semianalysis_SHOCKING: Many neocloud executives we spoke with feel that if they have non-NVIDIA networking gear on their cluster, or if their cloud has an AMD GPU or TPU offering, NVIDIA will retaliate. They feel that retaliation includes not giving ear
@semianalysis_Note that this doesn't apply to hyperscalers, as they have more buying power. But for neoclouds, executives feel NVIDIA has been using high-pressure tactics to keep them NVIDIA-only.(2/3)
@semianalysis_Some neocloud executives are even starting to consider offering TPUs or AMD GPUs quietly, to avoid getting pressured by NVIDIA and to avoid the feeling that NVIDIA will retaliate.(3/3)