Canada

Tenstorrent is changing leadership roles

The Toronto-based startup has grown rapidly, swapping CEO and CTO roles and laying out a compelling AI roadmap and RISC-V.

Tenstorrent, a Toronto-based startup that raised over $200 million in venture funding at a $1 billion valuation in May 2021, has now grown to over 280 employees and is backed by Eclipse Ventures and Real Ventures, among others. The company hired legendary chip designer Jim Keller as CTO in early 2021, an early angel investor and advisor. At the time, he said Tenstorrent had “the best architecture on the market” for AI.

Now, Jim and founder Ljubisa Baic have switched roles. Jim became CEO and founder Ljubiša Baić took over as CTO. In our opinion, these roles are really what each of them have been doing all along, so the change will be barely noticeable. But let’s look at the early impact Jim has already had.

Jim Keller, the famous processor architect, has taken over the role of CEO of Tenstorrent.

Tentorrent

When Jim joined the company, he saw changes he wanted to make to the accelerator roadmap, which had already produced two production chips, specifically the high-performance Greyskull. But in addition to learning about AI while designing Tesla’s FSD, Jim’s forte has always been processor core design. He led the AMD Zen effort that led to incredible market share gains for AMD, particularly in HPC with the EPYC line.

So when you put a CPU designer at the head of a design team, it shouldn’t be surprising that he designs new CPUs. In addition to the high-performance AI accelerators already in development, Jim envisioned the opportunity to tap into the momentum RISC-V is enjoying and potentially lead the market with enhanced cores as a second line of business. His team has come up with processors that can be built as chips, chiplets in accelerators, or perhaps even provided to multiple clients as chiplet IP. Companies like Meta and Google have been lobbying and pushing AMD and Intel for specific additional features for decades. Now they can choose to work with Tenstorrent and use the Tenstorrent RISC-V IP as a solid foundation from which to innovate. You can see below that this is much more than a slide-line business for Tenstorrent, as they plan processors that will run as clients, in edge processors, and in HPC and server applications.

The Tenstorrent RISC-V processor roadmap

Tentorrent

Tenstorrent’s Buda software stack is processor-independent, so the accelerator can be attached to Intel, AMD, or RISC-V processor core clusters. Below you can see part of the roadmap, which, like the AMD MI300 and NVIDIA Grace Hopper, combines AI and CPU cores on a single die or chiplet. So if you’re wondering why Tenstorrent is going the RISC-V route, you can see the value the company creates by doing AI and CPU at the same time.

An example of Jim’s work, a chiplet with 234 cores to accelerate Tenstorrent and a 16-core RISC processor.

Tentorrent

So, you might ask, what is Lubisha working on now? Pretty much the same as before. The founder led the development of the Tenstorrent software stack, which enables the company’s mantra of ease of use, enabling target-agnostic AI development with an open-source stack that developers must both compile, run, and optimize when needed, all in Python and now Python 2.0. Most ambitiously, Ljubisa plans to compile AI models that can scale across an arbitrarily large set of nodes, greatly simplifying the development and deployment of even the largest language models currently in development.

Conclusions

We’ve barely touched the surface of what Tenstorrent does in this short article, and we plan to cover more about Tenstorrent in the near future. We believe that Tenstorrent has both a differentiated AI hardware and software story and a compelling roadmap for RISC-V CPUs. In the nascent chiplet era, Tenstorrent is well positioned for success and we believe that Jim and Lubisha’s new roles are a natural fit.