SambaNova lands $1 billion at $11 billion valuation as AI inference race intensifies

SambaNova Systems has secured a fresh $1 billion in financing at an $11 billion valuation, a dramatic show of investor confidence in one of the more established names in AI chips and infrastructure. The Palo Alto-based company said the first close of its Series F was led by General Atlantic and is expected to expand further as additional backers join in the coming weeks.

The new capital arrives at a pivotal moment for the startup. SambaNova only recently introduced its SN50 chip and follows a $350 million Series E disclosed in February. The company is now positioning itself not simply as another semiconductor vendor, but as a full-stack platform for running large-scale AI inference securely and efficiently, especially for enterprises that do not want to rely entirely on public cloud providers.

CEO and co-founder Rodrigo Liang says the market around the company is moving quickly, with demand surging from banks, sovereign cloud projects, neocloud providers and large organizations building private AI systems. The latest round also underscores how quickly valuation expectations can shift in the AI infrastructure market, where companies tied to compute and inference are attracting outsized interest from strategic and financial investors alike.

Another large round, only months after the last

SambaNova’s new financing follows an unusually fast cadence of fundraising for a hardware company. In February, the company announced its Series E along with the SN50 launch. Roughly five months later, it has already returned to the market with a much larger Series F first close.

That pace reflects the capital intensity of AI infrastructure businesses. Designing chips, building systems, securing supply chains and supporting customer deployments all require significant upfront spending. For startups in this space, a large valuation can be as much about execution confidence and strategic relevance as it is about current revenue.

The company declined to discuss how much of the round remains open, but Liang said more investors are expected to participate before the financing is fully completed. In his view, the second close should wrap up soon.

“In the next few weeks, a few more investors will be coming in, and the second close is likely to finish up,” Liang said.

What SambaNova is selling now

Founded in 2017, SambaNova has spent years building systems designed for high-performance AI workloads. Its current pitch is centered on inference, the process of running trained models to generate outputs in real time. While model training often captures the headlines, inference is where many enterprises expect much of the durable AI spend to land.

According to Liang, the company is focused on what it calls premium inference: serving the largest and most demanding models with a hardware architecture built to handle enormous parameter counts and deliver low latency. The company says its systems can fit multi-trillion-parameter models into a single rack, a feature it argues helps customers run these models faster and more efficiently.

The company launched its SN40L system in 2023, first in cloud environments and later for on-premises use. Its next-generation SN50, announced in February 2026, is scheduled to begin shipping in the second half of 2026. SoftBank is identified as the first deployment partner for the newer system.

A table of the key milestones

Why banks are paying attention

One of the biggest commercial signals in SambaNova’s announcement is its new relationship with JPMorganChase. The bank has selected the company as an inference infrastructure partner, with SambaNova’s SN40L and SN50 systems intended to support secure, on-premises AI workloads.

For Liang, the JPMorganChase decision is meaningful well beyond one customer win. He argued that major banks are no longer comfortable depending only on cloud-based AI services for sensitive workloads, especially as the industry enters a phase in which private, heterogeneous infrastructure becomes more attractive.

Liang said the JPMorganChase deal sends a strong message to the financial sector that institutions should not assume cloud-only architectures are enough for every AI use case, particularly when sensitive data is involved.

The message is consistent with a broader trend in enterprise technology: when AI moves from experimentation to mission-critical deployment, control, security and compliance become more important. Financial institutions, in particular, are often among the first to insist on systems that can run in tightly controlled environments.

Why on-premises still matters

Even as cloud providers continue to dominate many AI rollouts, on-premises and private-cloud deployments remain important for organizations with strict data governance requirements. That includes banks, government agencies, energy companies and industrial operators.

• Data sovereignty concerns can limit use of public cloud services.
• Latency-sensitive applications may benefit from local inference systems.
• Some organizations want direct control over model access and auditability.
• Hybrid infrastructure can reduce dependence on a single vendor or architecture.

SambaNova is betting that these needs will create a large, long-lived market for inference appliances and private AI systems.

Intel relationship deepens, even as exit chatter continues

Intel remains a central part of SambaNova’s story. The chipmaker has backed SambaNova since its Series C and is also participating in the latest round. Five months ago, the two companies announced a multi-year partnership to support AI inference development based on Intel’s Xeon processors. They now co-develop products and sell them together.

That relationship gives SambaNova both technical reach and market access. Liang described the arrangement as a way to combine Intel’s scale with SambaNova’s technology, a partnership that could matter as demand for enterprise AI systems broadens.

Liang said the collaboration allows SambaNova to “leverage the scale of Intel” while advancing its own platform.

At the same time, there has been speculation about whether SambaNova might eventually be acquired. Bloomberg previously reported that Intel had explored a deal valuing SambaNova at about $1.6 billion. Liang did not confirm or deny any transaction interest, but he was notably open-ended about the company’s future.

He said the business continues to receive attention from potential suitors, though he suggested the current trajectory may point elsewhere.

“We’re always being approached,” Liang said, adding that the company could still pursue a public listing if growth continues on its current path.

A crowded market for AI infrastructure

SambaNova is raising money into a market that is both promising and increasingly crowded. AI infrastructure has become a broad battleground that includes model developers, cloud hyperscalers, custom chip designers, server makers, systems integrators and emerging neocloud firms.

Within that landscape, SambaNova is attempting to distinguish itself through specialization. Rather than pitching a general-purpose AI stack, it is focusing on the hardest, largest inference workloads and on customers that need private deployment options.

Liang said the company sees demand coming from three major customer groups:

1. Sovereign cloud projects built with local partners for governments that want control over data and infrastructure.
2. Neocloud providers seeking specialized infrastructure for AI customers.
3. Enterprises building internal AI systems for their own operations.

He believes the largest untapped opportunity is in the enterprise and public sectors, where adoption is still early and many organizations are only beginning to move beyond pilot projects.

Who else is using it

Alongside JPMorganChase, SambaNova cites customers and users including Saudi Aramco, Intel and several Japanese companies. The company’s ability to serve very large, security-conscious institutions is central to its pitch.

That customer mix also suggests a strategic pattern. SambaNova appears to be focusing on organizations with large budgets, complex infrastructure needs and a strong preference for controlled deployments rather than mass-market AI applications.

Why the valuation matters

An $11 billion valuation places SambaNova in rare company among infrastructure startups, especially in a segment where hardware companies usually face slower scaling and heavier capital requirements than software businesses. The number is a strong indicator that investors see the company as more than a niche chip vendor.

Valuation alone, however, does not settle questions about the company’s long-term trajectory. Its future still depends on execution: manufacturing, supply availability, product delivery, customer retention and the ability to translate technical differentiation into durable revenue.

Still, the financing gives SambaNova breathing room. It can continue to invest in product development, expand operations and secure the materials needed to meet expected demand over the next year.

Liang said the new capital will be used not only to grow the business, but also to protect the company’s supply chain so it can fulfill orders over the coming 12 months.

That emphasis is telling. In a market where chip demand can spike rapidly, supply is often the bottleneck. For hardware startups, capital is not just for growth; it is also a tool to lock in production capacity and reduce the risk of missing delivery windows.

Investor roster shows broad conviction

The Series F first close includes both new and existing supporters, signaling that SambaNova is drawing interest from a wide range of institutional backers. General Atlantic led the round, while other participants include Seligman Ventures, T. Rowe Price Associates and Capital Group.

Additional investors listed in the financing include A&E Investment, Assam Ventures, Battery Ventures, Cambium Capital, BlackRock, Kabila Capital, QFO Capital, Qatar Investment Authority, Vista Equity Partners and Volantis.

The diversity of that list matters. It suggests the company’s appeal stretches across venture capital, growth equity, sovereign capital and large asset managers, a blend that is increasingly common in late-stage AI deals.

What the investor mix suggests

• Strategic investors may be looking for long-term exposure to AI infrastructure.
• Growth investors may see a path to enterprise-scale revenue.
• Institutional investors may view SambaNova as a way to participate in the AI hardware buildout without backing a hyperscaler.

In short, SambaNova’s financing reflects a broader market belief that AI inference infrastructure will remain a major spending category as more organizations deploy models in production.

The company’s long-term path: independent, acquired or public?

Asked directly whether the new financing means SambaNova has settled the question of independence, Liang did not offer a definitive answer. Instead, he pointed to the ongoing interest the company receives and said the market remains fluid.

That ambiguity is typical of late-stage private companies in hot sectors. A startup can continue to raise capital, explore acquisition conversations and prepare for public-market options all at the same time. What tends to decide the path is not ideology, but momentum.

Liang suggested the company’s growth profile may eventually push it toward an IPO, even if acquisition interest persists. For now, the immediate priority appears to be scaling product deployment and satisfying the rising demand for inference hardware.

In a sector where every major company is trying to control more of the AI stack, SambaNova’s position is unusual but potentially attractive. It is not trying to compete head-on with large cloud platforms on breadth. Instead, it is trying to own a specific layer of the market where performance, security and deployment flexibility matter most.

Why this round stands out in the broader AI boom

The latest financing is part of a wider wave of funding and consolidation across AI, but it also highlights how the market has matured. Earlier AI investment cycles often centered on model innovation alone. Today, attention is shifting toward the physical and operational foundations that make AI practical at scale.

That includes chips, servers, networking, inference engines and private infrastructure. As enterprises move from experimentation to deployment, they are asking harder questions: How much will it cost to run the model? Can the system stay within security boundaries? Will the hardware be available when needed? Can inference happen fast enough to be useful?

SambaNova’s pitch is built around those questions. Its latest funding round suggests investors believe the answers can still support a large business.

Timeline of SambaNova’s recent moves

What happens next

In the near term, the main watchpoints are simple: whether the rest of the Series F closes quickly, how the SN50 rollout progresses and whether more major enterprise wins follow the JPMorganChase announcement.

The company also faces the practical challenge of converting enthusiasm into shipments. In hardware, especially at the leading edge of AI systems, sales announcements are only the start. Manufacturing readiness, deployment support and reliability all determine whether valuation enthusiasm turns into a sustainable franchise.

For now, though, SambaNova has achieved something that is increasingly difficult even in a frothy AI market: it has convinced large investors and large customers that there is still room for a specialized infrastructure company to grow quickly, command a premium valuation and remain strategically relevant.

If its bet on secure, high-performance inference proves right, the company could become one of the more important suppliers in the next phase of enterprise AI adoption. If not, the scale of this round will only sharpen the pressure to deliver.