Tenstorrent company profile: valuation, funding & revenue

This Tenstorrent company profile covers the company's valuation, revenue, funding rounds, investors and history. Tenstorrent works in semiconductors, with a focus on AI processors and RISC-V CPU IP. It was founded in 2016 and is headquartered in Toronto, Ontario, Canada. Its latest valuation on record is US$6.2 billion (post-money) (October 2026). Every figure links to a public source, and anything the company hasn't disclosed is marked as such.

Revenue

Funding history

Tenstorrent timeline

Key dates from the sources on this page, oldest first. Metric snapshots are listed after the funding events.

• 2016 — Tenstorrent is founded in Toronto by Ljubisa Bajic, Ivan Hamer and Milos Trajkovic.
• May 2021 — Series C: More than US$200M, valuing the company at US$1B. Investors: Fidelity (lead).
• Dec 2024 — Series D: About US$693M, valuing the company at About US$2B pre-money / US$2.6B post-money (reported). Investors: Samsung Securities and AFW Partners (leads); Bezos Expeditions, Fidelity, LG Electronics.
• Oct 2026 — Series E: US$1.5B, valuing the company at US$6.2B. Investors: Trace Capital and Innovation X (leads); ARK Invest, Fidelity, T. Rowe Price, Morgan Stanley, Sandisk, Tether Investments.
• December 2024 — Total funding: More than US$1 billion since 2016.

Investors

Overview

Tenstorrent designs AI processors and high-performance RISC-V CPUs, and licenses its chip designs to other companies. Its AI chips, named Grayskull, Wormhole and Blackhole, are sold in cards, workstations and servers for running and training AI models, with an open-source software stack that developers can inspect and modify. Alongside its own products, Tenstorrent licenses its RISC-V CPU and AI designs to companies building their own chips, including automakers and consumer electronics makers. It is one of Canada's most prominent deep technology companies and one of the best-known challengers to Nvidia.

Ljubisa Bajic, Ivan Hamer and Milos Trajkovic founded Tenstorrent in Toronto in 2016. Bajic had worked on chip design at AMD and Nvidia. In 2020, Jim Keller, one of the most celebrated chip architects in the industry, joined as chief technology officer and became CEO in 2023. Keller led or contributed to landmark designs at Digital Equipment Corporation, AMD (including the Zen architecture), Apple (the A-series chips), Tesla (its self-driving chip) and Intel. His involvement gave Tenstorrent unusual credibility and helped attract investors and customers.

Products and technology

• AI processors: Grayskull, Wormhole and Blackhole chips, built from many "Tensix" cores that combine compute, memory and networking, designed to scale across many chips with Ethernet connections.
• Systems: PCIe cards, desktop workstations and multi-chip servers such as the Galaxy system.
• RISC-V CPUs: high-performance CPU designs, including Ascalon, based on the open RISC-V instruction set.
• Open-source software: TT-Metalium and TT-Forge, low-level and compiler software released openly so developers aren't locked into proprietary tools.
• IP licensing: chip designs licensed to partners, including LG and Hyundai, and to Japan's Leading-edge Semiconductor Technology Center for national AI chip projects.

Why RISC-V and openness matter

Most CPUs use instruction sets owned by Intel and AMD (x86) or licensed from Arm. RISC-V is an open standard that anyone can use without paying royalties, which appeals to companies and countries that want more control over their chip technology. Tenstorrent combines RISC-V CPUs with its AI cores and releases its software openly, positioning itself as an alternative to closed ecosystems. Keller has argued that open hardware and software will win over time, as Linux did in operating systems. For customers building their own chips, licensing Tenstorrent's designs can be faster and cheaper than designing from scratch.

Funding

Tenstorrent raised more than US$200 million in a Series C in May 2021 at a US$1 billion valuation. Hyundai Motor Group and Samsung invested in 2023. In December 2024 it closed a Series D of about US$693 million, led by Samsung Securities and AFW Partners, with Bezos Expeditions, Fidelity and LG Electronics participating. Crunchbase reported a US$2 billion pre-money valuation, and Bloomberg reported a valuation of about US$2.6 billion, consistent with a post-money figure. Total funding exceeds US$1 billion. We haven't found a newer priced round.

Business model

Tenstorrent earns revenue from selling AI hardware, from licensing CPU and AI intellectual property, and from design services for partners' custom chips. Licensing deals can bring upfront fees and royalties without the cost of manufacturing. The company doesn't publish revenue, so we show it as not disclosed.

Competition

In AI chips, Tenstorrent competes with Nvidia, AMD, Intel and startups such as Etched, Groq, SambaNova and Cerebras. In CPU IP, it competes with Arm and other RISC-V designers such as SiFive. Its differentiation is the combination of high-performance RISC-V CPUs, scalable AI cores, open-source software and a licensing business that serves companies wanting their own chips. Our guide to AI chip startups compares the field.

Tenstorrent at a glance

Tenstorrent is a Toronto-founded AI and RISC-V chip company started in 2016 by Ljubisa Bajic, Ivan Hamer and Milos Trajkovic and led by Jim Keller. It sells AI processors and licenses CPU and AI designs to partners such as LG and Hyundai. It raised about US$693 million in a December 2024 Series D at a valuation reported at about US$2.6 billion, and in October 2026 a US$1.5 billion Series E at US$6.2 billion.

Timeline

• 2016: founded in Toronto.
• 2020: Jim Keller joins as CTO.
• May 2021: raises more than US$200 million at US$1 billion.
• 2023: Keller becomes CEO; Hyundai and Samsung invest; licensing deals expand.
• December 2024: raises about US$693 million in a Series D led by Samsung Securities and AFW Partners.
• October 2026: raises a US$1.5 billion Series E at US$6.2 billion, led by Trace Capital and Innovation X.

Canada's role in AI hardware

Canada is best known in AI for research pioneers such as Geoffrey Hinton and Yoshua Bengio and companies like Cohere, but it also has a deep semiconductor talent base, partly from ATI Technologies, the Toronto graphics chip company acquired by AMD in 2006. Many of Tenstorrent's early engineers came from that ecosystem. Tenstorrent shows how Canadian hardware expertise can compete globally, although much of its growth, manufacturing and sales activity now takes place internationally.

Sovereign chip programmes

Governments increasingly want domestic capabilities in advanced chips for economic and security reasons. Japan, South Korea, India and European countries have launched programmes to design or manufacture AI chips locally. Tenstorrent's willingness to license designs and work with national projects, such as its partnership with Japan's semiconductor initiatives, opens markets that companies selling only finished chips may struggle to reach.

Risks

• Nvidia's ecosystem: developers and customers are heavily invested in CUDA.
• Scale: competing on volume and cost with giant chipmakers is difficult.
• Execution: delivering successive chip generations on time requires flawless engineering.
• Licensing dependence: partners' chip projects can be delayed or cancelled.

Questions for buyers and partners

• How mature is the software stack for the models you run?
• How does performance per dollar compare with GPUs at your scale?
• What licensing terms and support apply for custom chip projects?

Latest developments (October 2026)

Tenstorrent raised a US$1.5 billion Series E at a US$6.2 billion post-money valuation, announced on 19 October 2026 and led by Trace Capital and Innovation X, according to Dealroom and Let’s Data Science, with Reuters confirming the terms. Investors include ARK Invest, Fidelity, T. Rowe Price and Tether. Earlier in 2026 the Jim Keller-led company reportedly held acquisition talks with Intel and Qualcomm before choosing to raise. The money will fund volume production of its Galaxy servers. See our AI chip startups guide.

What to watch

Watch new chip generations, licensing deals with automakers and governments, any new funding round or IPO preparation, and adoption of its open-source software.

Takeaways

Chip buyers and designers have an open, licensable alternative to closed ecosystems. Founders can see how a world-class technical leader changes a startup's trajectory. Investors should weigh strong IP against Nvidia's dominance.

Sources

Every figure on this page comes from the public sources below. Where sources disagreed, we used the company's own announcement or a primary news report. Private-company data changes; check the date beside each figure.

1. Dealroom
2. Crunchbase News
3. AI Business
4. Tenstorrent