Tenstorrent is an AI chip company led by chief executive Jim Keller, a veteran processor architect. It designs AI processors and servers, with a focus on inference: the stage at which a trained model answers requests. Its stated targets are serving open-weight models and video generation, and it sells its Galaxy servers as an alternative to Nvidia GPUs.
Funding and valuation
In October 2026, TechCrunch reported that Tenstorrent raised $1.5 billion at a $6.2 billion valuation, more than double the $2.6 billion it fetched in late 2024. The round was led by Trace Capital and Innovation X, with Cadence Ventures, Eclipse, Fidelity, SanDisk and T. Rowe Price among the backers named. Five months earlier, TechCrunch noted, Intel and Qualcomm had reportedly shown acquisition interest at valuations of as much as $5 billion; we have not independently verified that report. Keller told Bloomberg that the company chose independence after showing that Galaxy servers could run DeepSeek models.
Why open-weight inference
Open-weight models, whose trained parameters are published, can be hosted by anyone. A buyer who wants to run one itself needs hardware that is affordable and not tied to a single vendor’s software. Tenstorrent’s pitch is to be that option. The company had not, in the coverage we reviewed, published performance numbers tied to the new round, so its video-generation focus should be read as a product direction rather than a measured result.
Competitive position
Nvidia’s advantage is as much software as silicon, as we explain in why Nvidia dominates AI. Big cloud providers build their own accelerators, covered in why Big Tech builds its own AI chips, so a startup selling to everyone else must win on price, availability or openness.
