SambaNova announced a $1 billion oversubscribed funding round at an $11 billion valuation, led by General Atlantic with participation from Seligman Ventures, T. Rowe Price, Capital Group, and existing backers including Intel Capital, BlackRock, and QIA. The round builds on a $350M+ Series round earlier in 2026. CEO Rodrigo Liang credited the valuation jump to customer momentum: "having 40 or 50 customers of a certain size was meaningful to our investors."
Simultaneously, SambaNova announced JPMorganChase as a customer deploying SambaNova's SN40 and SN50 RDUs (Reconfigurable Dataflow Units) for secure on-premises AI inference via a multi-year agreement. Liang noted that enterprises increasingly want inference workloads run privately on-prem due to data privacy concerns and model size, which SambaNova's disaggregated architecture supports. SambaNova also disclosed a $3.5B order from Vector Core Compute (a venture from Vista Equity Partners and Cambium Capital) for deployment over three years.
For architects: SambaNova is positioning disaggregated inference—combining SambaNova decode acceleration with NVIDIA Blackwell GPUs for prefill—as a cost-competitive enterprise alternative to GPU-only clusters. The JPMorganChase win validates on-prem inference demand, and SambaNova expects profitability next year with a low burn rate (no data center buildout).