Ai Engineering 3 min read

$1B Series F Lands SambaNova SN50 Inference at JPMorgan

SambaNova Systems has raised a $1 billion Series F round at an $11 billion valuation to scale production of its SN50 enterprise AI inference infrastructure.

On July 8, SambaNova Systems announced the first close of a $1 billion Series F financing round led by General Atlantic, valuing the AI chipmaker at $11 billion post-money. This capital injection arrives just five months after a Series E round and marks a total reversal from December 2025 rumors of a $1.6 billion Intel acquisition. The funding is earmarked to scale production for on-premise AI inference infrastructure targeting regulated enterprise markets.

JPMorgan Chase and Enterprise Deployment

Alongside the funding, JPMorgan Chase announced it will deploy SambaNova’s SN40 and SN50 systems to run secure, on-premises AI workloads. The bank’s selection validates the company’s full-stack approach, where regulated industries avoid public cloud GPU clusters while maintaining high throughput over proprietary data. This deployment models a broader industry shift toward ASIC-native cloud architectures for production environments that process sensitive customer telemetry.

Hardware Efficiency and the SN50 Appliance

SambaNova’s valuation surge maps directly to enterprise demand for its flagship SN50 chip, released in February 2026. Designed specifically for inference workloads rather than general-purpose model training, the SN50 operates at approximately 20 watts. The hardware ships to enterprise data centers inside the SambaRack SN50, a 16-chip appliance designed for drop-in rack compatibility.

MetricSambaNova SN50Nvidia B200 Equivalent
Power Consumption~20 wattsHardware dependent
Relative Throughput3x base1x base
Peak Speed5x fasterBaseline

The Intel Strategic Alliance

Intel Capital participated in both the Series E and F rounds, taking a 9% stake in SambaNova following antitrust clearance in May 2026. The two companies are co-developing AI hardware that integrates Intel Xeon processors with SambaNova’s Dataflow architecture. Intel CEO Lip-Bu Tan has chaired SambaNova’s board since its 2017 founding. Rather than an acquisition, the current relationship functions as a multi-year manufacturing alliance to automate AI infrastructure deployments across shared enterprise customer bases.

Market Context and On-Premise Inference

The AI inference sector is projected to reach $250 billion by 2030, heavily fragmenting between merchant silicon providers and custom ASIC designers. SambaNova positions its architecture directly against Nvidia for full-stack inference. Regulated industries like banking and government increasingly require dedicated infrastructure to process data locally, eliminating the shared tenancy and data egress risks associated with external APIs.

If you are evaluating inference hardware for highly regulated environments, the SN50 provides a measurable power and throughput advantage over standard GPU clusters. The primary technical tradeoff remains ecosystem compatibility, requiring your production models to compile efficiently against SambaNova’s proprietary Dataflow environment rather than relying on standard CUDA pipelines.

Get Insanely Good at AI

Get Insanely Good at AI

The book for developers who want to understand how AI actually works. LLMs, prompt engineering, RAG, AI agents, and production systems.

Keep Reading