SambaNova Systems announced the first close of a $1 billion Series F on July 8 at an $11 billion valuation, led by growth-equity firm General Atlantic with a notably deep bench of institutional capital: BlackRock, the Qatar Investment Authority, Vista Equity Partners, T. Rowe Price, Capital Group and Battery Ventures all participated. A second close is expected in the coming weeks.
The pace is the story here as much as the size. SambaNova closed a $350 million Series E just five months ago, in February 2026 -- meaning its valuation has roughly tripled in under half a year, a compounding rate that puts it alongside the fastest-appreciating AI infrastructure names of 2026. The company builds custom AI chips, the SN40L and SN50, designed specifically for inference workloads rather than training, positioning it as an alternative to Nvidia's GPU dominance for enterprises that want to run AI models on their own infrastructure.
That inference-specific bet just landed a marquee enterprise customer: JPMorganChase is deploying SambaNova's chips for on-premise AI inference, a validation that matters more than the funding number for a chip startup trying to prove it belongs in production at a bank's scale, not just in benchmarks. SambaNova now competes most directly with Cerebras -- which priced its own IPO earlier this year and popped 108% on debut -- and Groq, both chasing the same thesis that inference, not training, is where the next wave of AI-chip demand concentrates as models move from research labs into daily enterprise use.
An $11 billion valuation for an inference-chip company is a meaningful marker against Cerebras' public trading range and Groq's last private round, and it underscores how much capital is now flowing specifically into the compute layer beneath the compute layer -- alternatives to Nvidia that promise lower cost-per-token at production scale. For infrastructure-focused investors, the SambaNova-JPMorganChase relationship is the more durable signal than the valuation multiple: enterprise inference contracts, not benchmark wins, are what will separate this generation of AI-chip challengers from the last one.
The bear case is that SambaNova is still racing supply chain and manufacturing capacity against much larger, better-capitalized rivals -- Nvidia's balance sheet dwarfs anything SambaNova can raise, even at $1 billion a round. What to watch next: the terms and size of the pending second close, whether other regulated-industry customers follow JPMorganChase's lead, and how SambaNova's inference benchmarks hold up against Cerebras and Groq as all three scale production this year.