OpenAI signs $20B multi-year compute deal with Cerebras
OpenAI's early-2026 $20 billion multi-year agreement with Cerebras for compute capacity and related services was the structural piece that re-rated Cerebras from niche wafer-scale vendor to credible NVIDIA second source — and underwrote the May 2026 IPO.
The deal commits OpenAI to Cerebras inference capacity at substantial scale over multiple years. For OpenAI, the strategic value is supplier diversification — being able to walk away from NVIDIA terms because there's a real alternative on the floor. For Cerebras, it's a multi-year revenue floor that justified the IPO valuation.
The wider read: hyperscaler-grade AI compute is no longer single-sourced from NVIDIA. AMD's MI350/MI450 generation, Google's TPU v7, Cerebras WSE, Groq's LPU, and SambaNova's SN50 all hold real production-grade commitments from at least one Tier-1 buyer in 2026. The "NVIDIA monopoly" framing of 2024 is now structurally wrong even if NVIDIA still ships the majority of units.