Tenstorrent $1.5B Raise Doubles Valuation to $6.2B

The Tenstorrent $1.5B raise is one of the largest semiconductor rounds of the year: the chip startup announced on Monday that it has raised $1.5 billion at a $6.2 billion valuation. The Series E was led by Trace Capital, a new firm launched by Quiet Capital co-founder Ben Mahdavi, and Innovation X, with participation from Cadence Ventures, Eclipse, Fidelity, SanDisk and T. Rowe Price. The round more than doubles the company's valuation from the $2.6 billion it fetched in late 2024, according to reporting by TechCrunch.

The Santa Clara, California-based company said the new capital will go toward scaling production of its Galaxy servers — racks that combine dozens of Tenstorrent chips to power chatbots, coding assistants and video generators — with an emphasis on chips tailored for serving open-weight AI models. The Tenstorrent $1.5B round arrives as demand for inference capacity keeps climbing and developers hunt for alternatives to Nvidia's tightly controlled ecosystem.

A bet on open-weight inference

Tenstorrent builds AI chips and software on RISC-V, an open-source chip architecture that lets companies design processors without license fees and reduces dependence on proprietary ecosystems such as Arm's, according to Reuters. That openness is the heart of the company's pitch: rather than locking customers into a closed stack, Tenstorrent wants to be the neutral hardware layer for the growing universe of open-weight models.

Jim Keller, Tenstorrent's CEO and one of the industry's best-known chip architects, told Bloomberg that the company decided to remain independent and raise more capital after it demonstrated that its Galaxy servers could run DeepSeek models — a proof point that its hardware can serve the frontier open models developers actually want to deploy. The Tenstorrent $1.5B raise gives the company the balance sheet to scale that hardware from showcase deployments into serious production volume.

Reuters reported that the round also drew backing from Cathie Wood's ARK Invest, Morgan Stanley and Tether Investments, alongside SanDisk and T. Rowe Price. The breadth of the syndicate — venture firms, strategics and institutional capital — signals that open-model inference has moved from a niche bet to a mainstream infrastructure thesis. The Tenstorrent $1.5B valuation step-up suggests investors now treat independent inference hardware as a durable category rather than a speculative trade.

The Nvidia-challenger wave

Tenstorrent is the latest mover in a wave of AI hardware startups trying to chip away at Nvidia's dominance. Earlier this month, TechCrunch reported that Etched, another AI chip maker, was fielding investment offers at valuations of $40 billion or more — a number that shows just how much capital is now chasing non-Nvidia compute.

Interest in Tenstorrent itself has been building for months. About five months ago, Bloomberg reported that Intel and Qualcomm had expressed acquisition interest in the 10-year-old company at valuations of as much as $5 billion, according to TechCrunch. Keller's team chose to stay independent — and Monday's Tenstorrent $1.5B raise at a $6.2 billion valuation now prices the company above those reported takeover offers.

The company's last funding round raised more than $693 million in December 2024 at a $2 billion pre-money valuation, led by Samsung Securities and AFW Partners, according to Reuters. The Tenstorrent $1.5B round extends a rapid funding cadence: roughly $2.2 billion raised in under two years, all pointed at manufacturing scale and software maturity for its open RISC-V stack.

Why it matters for AI agents

For the agent ecosystem, the significance of the Tenstorrent $1.5B round is downstream of the hardware itself. Autonomous agents run on inference: every tool call, web-browse step and multi-turn reasoning loop burns tokens. Cheaper, more abundant inference — especially on open-weight models that companies can run in their own data centers — directly lowers the cost of deploying agents at scale.

That dynamic is already visible in the market. Nvidia itself has leaned into open models for enterprise agents, and startups racing to build agent infrastructure are acutely sensitive to per-token economics. If Tenstorrent's Galaxy servers can serve open models at materially lower cost than the incumbent stack, the Tenstorrent $1.5B bet could show up not in benchmark tables but in the unit economics of the next generation of agent platforms.

The open question is execution: scaling chip production is unforgiving, and Nvidia's software moat remains deep. But with $1.5 billion in fresh capital and a valuation that now exceeds reported acquisition bids, Tenstorrent has bought itself the runway to find out. The Tenstorrent $1.5B round is, at its core, a wager that the future of AI compute looks more like open infrastructure than a single company's walled garden.

Further reading: TechCrunch on the Tenstorrent raise and Reuters' coverage of the round. Related coverage on genznewz.com: Anthropic's own numbers show junior tech hiring has collapsed.