Tenstorrent

Tenstorrent is a Canadian artificial intelligence and machine learning technology company specializing in AI architectures focused on scalable deep learning and adaptability. Tenstorrent has developed a proprietary processor core called the Tensix; released two products, Grayskull (machine learning processor) and Wormhole (networked machine learning processor); and is developing the industry's first 8-wide decoding RISC-V core capable of addressing both client and HPC workloads. In 2024, the company plans to release a multi-chiplet solution called Grendel, featuring its own Ascalon general-purpose cores based on its RISC-V microarchitecture and a Tensix-based chiplet for machine learning workloads.

Tenstorrent was founded in 2016 by Ljubisa Bajic, Milos Trajkovic, and Ivan Hamer. Headquartered in Toronto, Canada, Tenstorrent has US offices in Austin, Texas, and Santa Clara, California, as well as global offices in Belgrade, Serbia, and Bangalore, India. Jim Keller, former Intel microprocessor engineer, joined Tenstorrent as chief technology officer (CTO) on January 6, 2021. Keller is a well-known microprocessor engineer best known for his work at AMD and Apple. Keller was previously an angel investor and advisor to the company. In January 2023, the company announced a leadership change with CEO (now previous CEO) Bajic and Keller switching roles. The company has since David Bennett as CCO, Keith Witek as COO, Matthew Mattina as VP of machine learning, Wei-Han Lien as chief CPU architect, Olof Johansson as VP of OS and infrastructure, and Mamoru Nakano as head of sales in Japan. Wei-Han Lien is known for working on Apple's A6, A7 (the world's first 64-bit Arm SoC), and M1 CPU microarchitectures and implementation.

In May 2023, Tenstorrent announced its first public deal with LG, licensing the company's AI and RISC-V CPU technologies for inclusion in its future TV and automotive products. At the RISC-V Days Tokyo 2023 Summer conference, Keller stated his plans for Tenstorrent to compete with NVIDIA, pushing the AI industry forward using RISC-V architecture and chiplet designs. Keller stated:

NVIDIA's GPU responded to the timing when the trend of AI greatly increased. However, GPUs are not optimized for AI processing. We have proven that our AI accelerator exceeds in efficiency. Also, one of the issues with AI processing by GPU is that access to CPU and memory becomes a bottleneck.The generation of Blackhole migration, which integrates RISC-V as a CPU, eliminates these bottlenecks. As a result, more power-saving and low-cost AI computers can be realized.

The company has raised over $230M in five funding rounds, including a $200M Series C round in May 2021 that valued the company at $1B. The round was led by Fidelity Management and Research Company and includes additional investments from Eclipse Ventures, Epic CG, and Moore Capital.