SambaNova Systems has locked in $1 billion in fresh capital at an $11 billion valuation, marking the first close of its Series F funding round — a striking acceleration for a Palo Alto-based AI chip startup that only completed a $350 million Series E just five months ago. The SambaNova AI funding round was led by General Atlantic, with a broad coalition of institutional investors joining in, and CEO Rodrigo Liang says more are still coming.
The speed of this latest raise is hard to ignore. Five months between a $350 million Series E and a $1 billion Series F — at a valuation that has climbed dramatically — reflects just how fast institutional appetite for AI inference infrastructure is intensifying. General Atlantic took the lead position, joined by a roster of heavy-hitting investors including BlackRock, Qatar Investment Authority, Vista Equity Partners, T. Rowe Price Associates, Capital Group, Battery Ventures, and several others ranging from Seligman Ventures to Volantis.
Liang told TechCrunch the round isn’t fully closed yet. “In the next few weeks, a few more investors will be coming in, and the second close is likely to finish up,” he said. That means the final figure could climb further.
The company was founded in 2017 and is now nine years old. Intel has backed it since its Series C and participated in this latest round as well — a detail that takes on added meaning given the two companies’ deepening commercial relationship.
SambaNova’s hardware story centers on its SN50 chip, unveiled in February 2026 — the same month as the Series E close. The chip is purpose-built for agentic inference, and it’s due to begin shipping to customers in the second half of 2026. SoftBank is the first confirmed deployment partner for the SN50 system, a significant early anchor for a product not yet in general circulation.
Its predecessor, the SN40L, launched in September 2023 and became available for on-premises deployment from November 2023. Both the SN40L and the SN50 systems are now central to SambaNova’s enterprise pitch.
What sets SambaNova’s positioning apart is its focus on what Liang calls “premium inference” — running the largest frontier models, which now span trillions of parameters, as fast as possible. The company’s architecture allows it to fit multi-trillion-parameter models onto a single rack, which is a meaningful operational advantage when enterprises are trying to run sensitive models without routing data through public clouds.
The relationship with Intel has evolved well beyond a capital stake. The multi-year SambaNova-Intel partnership, announced five months ago, centers on co-developing AI inference products based on Intel’s Xeon chip — and then taking those products to market together. That’s a joint go-to-market arrangement, not just a licensing deal.
“That gives us a great relationship with them that lets us leverage the scale of Intel with the technology we have,” Liang said. For SambaNova, Intel’s manufacturing and distribution reach represents a significant amplifier for its own hardware stack. For Intel, aligning with a specialist inference company helps it compete in the AI infrastructure market where it has faced pressure.
The partnership also helps explain why Intel participated in the Series F. This isn’t passive financial exposure — it’s a strategic co-investor with active skin in SambaNova’s commercial success.
JPMorgan Chase has selected SambaNova as its inference-infrastructure partner, deploying both the SN40L and SN50 systems to run secure, on-premises AI inference inside the bank. For a company targeting enterprise and financial sector adoption, this is a marquee win.
Liang framed it explicitly as an industry signal. “Having JPMorgan Chase decide they’re going to use SambaNova for their inference solution is a big deal,” he said. “It sends a message to the banking industry that it’s time not to completely depend on cloud services. These banks want heterogeneous infrastructure.”
The logic is straightforward: financial institutions operating with highly sensitive models — credit risk, fraud detection, trading strategies — cannot always route inference workloads through third-party cloud environments. On-premises inference infrastructure addresses that compliance and security requirement directly. The JPMorgan decision suggests the largest banks are now ready to build that capability in-house, and are selecting specialized vendors to do it.
Beyond banking, Liang argued the pattern will spread. Enterprises and governments, he said, are “just starting their AI journey,” with most early AI investment concentrated among frontier model developers. That leaves substantial commercial opportunity still untapped across industries that haven’t yet built their own inference stack.
SambaNova’s market targets are specific. The company pursues three customer types: sovereign clouds (government-funded private cloud infrastructure), neoclouds (next-generation cloud providers), and enterprises building private AI deployments for internal use. JPMorgan represents the third segment. Saudi Aramco and several Japanese firms are also named as customers alongside Intel.
The new capital will go directly toward supply chain security. Liang described it as essential: “We’re using that capital to secure the supply chain,” he said, framing it as the key constraint on SambaNova’s ability to fulfill orders over the next 12 months. With the SN50 set to ship in the second half of 2026, the timing of the funding is deliberate — building inventory and materials pipeline ahead of expected deployment demand.
SambaNova’s independence hasn’t always been guaranteed. A December report from Bloomberg News noted the company had been in acquisition talks with Intel at a valuation of roughly $1.6 billion — a figure that now looks remarkably different against the current $11 billion post-money valuation.
Asked directly whether the Series E and F closes signal a commitment to staying independent, Liang was careful not to close any doors. “We’re always being approached,” he said. He acknowledged that momentum and growth would most likely drive the company toward “being public at some point” — but left acquisition possibilities explicitly on the table.
That ambiguity is itself a strategic posture. A company with JPMorgan Chase as an infrastructure partner, a co-development agreement with Intel, SoftBank as a hardware deployment partner, and backing from BlackRock and the Qatar Investment Authority has significant optionality. Whether SambaNova reaches a public listing or attracts a premium acquisition offer, the Series F positions it to negotiate from strength rather than necessity.
SambaNova Systems raised $1 billion in the first close of its Series F funding round, at a valuation of $11 billion. Additional investors are expected to join before the second close.
General Atlantic led SambaNova’s Series F funding round. Other participants include BlackRock, Qatar Investment Authority, Vista Equity Partners, T. Rowe Price Associates, Capital Group, Battery Ventures, and several others.
SambaNova has a multi-year partnership with Intel to co-develop AI inference products based on Intel’s Xeon chip. The two companies now build products together and take them to market jointly. Intel has been a backer of SambaNova since its Series C round.
JPMorgan Chase selected SambaNova as an inference-infrastructure partner to run secure, on-premises AI inference using both the SN40L and SN50 systems.
Article produced with the assistance of artificial intelligence and reviewed by the editorial team.