News
Guide
Is AI Making Your Organization More Productive or More Dependent?
Key Takeaways
- SambaNova closed the first tranche of a $1B Series F at an $11B post-money valuation.
- JPMorganChase selected SambaNova for secure, on-premises AI inference.
- IT leaders gain new options for scalable, sovereign on-prem AI.
SambaNova announced the first close of a $1 billion Series F round at an $11 billion post-money valuation.
General Atlantic led the round, joined by Seligman Ventures, T. Rowe Price Associates and Capital Group. Battery Ventures, BlackRock, Intel Capital, Qatar Investment Authority and Vista Equity Partners also participated.
SambaNova disclosed that JPMorganChase selected the company as an inference infrastructure partner, deploying SN40 and SN50 systems for secure, on-premises AI inference.
According to SambaNova, proceeds will fund capacity expansion, product development and scaled deployments for enterprises, neo-clouds, sovereign AI customers and service providers globally.
On-Prem AI Inference & the Full Stack
Where enterprises run inference — and on what hardware — now shapes cost, data control and strategic value. The case for purpose-built, on-premises inference is strengthening as inference costs climb, data sovereignty mandates tighten and ROI gaps widen.
"At JPMorganChase, AI infrastructure has to meet a very high bar for performance, control and reliability. We're excited to deploy SambaNova's RDU architecture and looking forward to testing its speed and security for on-prem inference in our demanding enterprise AI workloads."
- Darrin Alves
CIO, Infrastructure Platforms, JPMorganChase
The ROI Case for Sovereignty
An MIT Technology Review survey of more than 2,000 senior executives across 13 countries found that companies deeply committed to AI and data sovereignty reported roughly 5X higher AI ROI from generative and agentic AI deployments compared to organizations with weaker governance controls.
Security and resilience ranked as the top driver, cited by 85% of respondents. Nearly all, 95%, said they plan to establish their own AI and data platforms within three years.
Inference Is the Cost Problem
Training creates a model. Inference runs it at scale on every request.
As generative AI moves deeper into production, inference is becoming a significant cost driver across coding assistants, long-context agents and multi-step workflows. Existing chip architectures face memory bandwidth constraints that limit how economically large models can run.
Full-Stack Architecture & Tiered Deployment
Production AI systems require far more than GPUs. Behind any user-facing interface sits a stack of orchestration layers, vector databases, retrieval pipelines, monitoring and governance controls, and high-performance storage and networking.
As Shailesh Manjrekar of Fabrix.ai told VKTR, much of the real cost sits below the surface: "Inference cost is visible. What's less visible is all the operational drag around it, pipelines that fail silently and retry, messy, redundant data ingestion and agentic workflows with no real observability. That's where the money actually goes missing."
Many organizations are adopting tiered inference, routing routine tasks to smaller local models while reserving frontier models for complex reasoning. Vertically integrated infrastructure is gaining traction among enterprises with composable architectures and sovereignty requirements.
Recent SambaNova Developments
SambaNova has more than doubled its valuation in five months. CEO Rodrigo Liang said proceeds will fund deployments across enterprises, neo-clouds, sovereign AI and service providers.
JPMorganChase joins Meta, Hugging Face, SoftBank and TEPCO Systems as customers selecting SambaNova's RDUs for on-premises inference.
The raise follows a pivotal February 2026 triple announcement: a $350 million-plus oversubscribed Series E led by Vista Equity Partners, the launch of its 5th-generation SN50 RDU chip — marketed as 5X faster and 3X lower TCO than Nvidia's Blackwell B200 for agentic AI — and a multi-year collaboration with Intel pairing Xeon CPUs and GPUs with SambaNova's inference systems.