SambaNova raises $1B at $11B valuation for AI inference chips, JPMorgan signs on

The article exposes the tension between surging demand for AI inference — especially from banks and governments demanding private, on-premises infrastructure — and the hyperscaler-dominated cloud model that has powered most AI growth so far. SambaNova positions itself as the “premium inference” provider purpose-built for multi-trillion-parameter models that fit onto a single rack, but faces the operational challenge of scaling its supply chain fast enough to capture what CEO Rodrigo Liang calls “a huge amount of revenue” still waiting in enterprise and sovereign markets.

SambaNova raised $1 billion at an $11 billion valuation led by General Atlantic, just five months after a $350M Series E. The funding will secure supply chain capacity for the next 12 months. The company deepened its partnership with Intel — now co-developing products — and landed JPMorganChase as an inference-infrastructure partner, deploying SN40L and next-generation SN50 systems for secure on-premises inference. Liang sees three customer segments: sovereign clouds, neoclouds, and enterprises building private AI infrastructure. The SN50, unveiled in February 2026, will begin shipping in the second half of 2026 with SoftBank as its first deployment partner.

For builders, the key takeaway is that the next wave of AI growth will shift from training and cloud inference to private, secure, on-premises inference for enterprises and governments. SambaNova‘s capital raise and customer wins signal that heterogeneous infrastructure choices are becoming real alternatives to full cloud dependency. The company keeps the door open to acquisition or IPO, but the immediate priority is delivering hardware that handles frontier-scale models efficiently while securing the supply chain to meet what Liang describes as an “incredible wave of demand.”