Chip and server firm SambaNova has unveiled a new chip to support agentic AI workloads.

Dubbed SN50 AI, the company claims the hardware is 5x faster than “competitive chips” and offers a 3x lower total cost of ownership, although it did not name the comparative hardware.

In a statement, SambaNova said the chip also offers 4x more network bandwidth than its previous generation, and up to 256 accelerators can be linked over a multi‑terabyte‑per‑second interconnect.

Built using the company’s Reconfigurable Data Unit (RDU) architecture, SN50 features a three-tier memory design that supports 10 trillion parameter models and can power “thousands of simultaneous AI sessions with consistent high performance.”

SoftBank will be the first company to deploy the SN50 at its AI data centers in Japan, with the hardware set to power low‑latency inference services for sovereign and enterprise customers across Asia‑Pacific.

SambaNova also announced it has entered into a “multi-year strategic collaboration” with Intel to deliver “high‑performance, cost‑efficient AI inference solutions for AI‑native companies, model providers, enterprises, and government organizations around the world.”

The agreement will support the rollout of an Intel‑powered AI cloud built on Intel Xeon‑based infrastructure, with the two companies also planning to integrate their respective infrastructure offerings to power code generation, multimodal applications, and agentic workflows. Intel and SambaNova will also pursue a joint go-to-market strategy across enterprise, cloud, and partner channels.

“AI is no longer a contest to build the biggest model,” said Rodrigo Liang, co‑founder and CEO of SambaNova. “With the SN50 and our deep collaboration with Intel, the real race is about who can light up entire data centers with AI agents that answer instantly, never stall, and do it at a cost that turns AI from an experiment into the most profitable engine in the cloud.”

Kevork Kechichian, EVP, GM, data center group, Intel, added: “Customers are asking for more choice and more efficient ways to scale AI. By combining Intel’s leadership in compute, networking, and memory with SambaNova’s full-stack AI systems and inference cloud platform, we are delivering a compelling option for organizations looking for GPU alternatives to deploy advanced AI at scale.”

In addition to unveiling the SN50 chip, the company also announced that it has raised $350 million in a Series E funding round led by Vista Equity Partners and Cambium Capital, with “strong participation” from Intel Capital as part of the newly announced collaboration.

The round attracted a number of new investors, including Assam Ventures, Battery Ventures, Gulf Energy, Mayfield Capital, Saudi First Data, and Seligman Ventures. The newly raised capital will be used to expand SN50 production, scale SambaCloud, and deepen enterprise software integrations, the company said.

Founded in 2017, SambaNova previously focused on training workloads, but pivoted in 2025 to become an AI cloud services provider. Its Suite product is touted as an on-prem or cloud-based offering that allows companies to train their own generative AI models based on popular foundation models.

The company first unveiled its SambaNova Cloud in September 2024. It provides cloud-based AI inference services using the company’s SN40L AI chip, launched in September 2023 and capable of running models with up to five trillion parameters.