SambaNova

SambaNova is an AI chip company founded in 2017 in Palo Alto, California by Stanford computer scientists Kunle Olukotun and Chris Ré together with CEO Rodrigo Liang, a former SPARC processor engineering leader at Sun Microsystems and Oracle. Rather than compete with Nvidia on general-purpose GPUs, SambaNova designs a fundamentally different chip built around a Reconfigurable Dataflow Unit (RDU) — an architecture that maps the structure of an AI model directly onto the silicon, aiming to run large models with fewer chips and lower power than a comparable GPU cluster.

The company sells a full-stack platform rather than a bare chip: its RDU hardware, the SambaNova Suite for enterprise deployment, and SambaNova Cloud, an inference API for developers. Its flagship SN40L processor, launched in 2023, was designed to run models with up to five trillion parameters on a single rack, and the next-generation SN50, unveiled in February 2026, begins shipping in the second half of 2026 with SoftBank as its first deployment partner. SambaNova has leaned into two markets where its efficiency and on-premises story resonate: regulated enterprises and inference at scale. JPMorgan Chase selected it as an inference-infrastructure partner to run secure, on-premises AI inside the bank.

SambaNova has raised heavily to fund the capital-intensive work of designing and manufacturing custom silicon. In July 2026 it raised $1 billion in the first close of a Series F led by General Atlantic at an $11 billion valuation — just five months after a prior round — with backers including Intel, BlackRock, T. Rowe Price, and the Qatar Investment Authority. Intel had earlier explored acquiring the company, but those talks stalled; CEO Rodrigo Liang has since signaled that a public listing, rather than a sale, is the more likely path. SambaNova sits in a crowded field of inference-focused challengers — Groq, Cerebras, and Etched among them — all wagering that the next phase of the AI boom is defined less by training ever-larger models and more by running them cheaply and reliably at scale.