JPMorgan just picked SambaNova to run AI inside its own walls. Banks are done trusting the cloud with sensitive models.

SambaNova just raised $1B at $11B valuation.

5 months after its last mega-round.

But the funding isn't the story. JPMorgan Chase just selected SambaNova as its inference infrastructure partner — deploying SN40L and SN50 systems for secure, on-premises AI behind the bank's own firewalls.

This is the signal that matters.

The world's largest bank just told the market it will not run sensitive model inference on AWS or Azure. It wants silicon it controls, inside walls it owns, with no third-party access to the data path.

SambaNova CEO Rodrigo Liang put it directly: banks "of the caliber of JPMorgan" are building private, secure infrastructure to run inference on their most sensitive models. He expects this to resonate beyond banking into enterprises and governments that are "just starting their AI journey."

General Atlantic led the round. Intel Capital, BlackRock, Qatar Investment Authority, and T. Rowe Price all participated. The company doubled its valuation in 5 months.

If you are a CISO or infrastructure leader still routing all AI inference through shared cloud APIs, your biggest competitor just drew the line. On-prem AI inference is no longer a niche preference. It is becoming the default for regulated industries.

Audit your inference architecture today. Ask your cloud provider what happens to your model data after the API call. If you cannot answer that question, you are already behind.

SOURCE: https://techcrunch.com/2026/07/08/sambanova-draws-1b-at-11b-valuation-in-series-f-first-close/
VERIFIED: TechCrunch, Reuters, CNBC, U.S. News
SIGNAL: JPMorgan's on-prem inference bet signals that enterprise AI is splitting into two tracks — cloud for experimentation, private infrastructure for production. Your vendor strategy needs to account for both.