What is Tenstorrent's core business?

Tenstorrent is a next-generation computing company that designs and builds high-performance computers for artificial intelligence. This includes AI accelerator hardware, workstations, servers, and an open-source software stack.

What kind of products does Tenstorrent offer?

Tenstorrent offers a range of AI hardware products including Blackhole® accelerator cards, TT-QuietBox™ liquid-cooled AI workstations, Tenstorrent Galaxy™ scale-out AI servers, and flexible IP licensing for specific AI workloads.

TT-Forge™ is Tenstorrent's MLIR-based open-source compiler. It's built on their existing AI software stack and designed to compile and optimize models from frameworks like PyTorch, JAX, and ONNX for execution on Tenstorrent hardware. It's currently in public beta.

What is Tenstorrent's philosophy regarding software and hardware?

Tenstorrent advocates for an open future in AI computing. Their IP is transparent, architectures are open, and software is open source. This allows users to edit, select, fork, and own their silicon future, promoting flexibility and preventing vendor lock-in.

Where are Tenstorrent's main offices located?

Tenstorrent is headquartered in the U.S. with offices in Austin, Texas, and Silicon Valley. They also have global offices in Toronto, Belgrade, Seoul, Tokyo, and Bangalore, reflecting their international presence and talent base.

Tenstorrent Review — AI Hardware

Tenstorrent is a pioneering next-generation computing company dedicated to building advanced computers specifically designed for AI workloads. Headquartered in the U.S. with key offices in Austin, Texas, and Silicon Valley, alongside a global footprint spanning Toronto, Belgrade, Seoul, Tokyo, and Bangalore, Tenstorrent convenes top experts in computer architecture, ASIC design, RISC-V technology, advanced systems, and neural network compilers. Their product portfolio encompasses accelerator cards, powerful liquid-cooled workstations capable of running models up to 120 billion parameters, and sovereign, scale-out servers for production AI deployments. A cornerstone of their approach is the TT-Forge™ MLIR-based open-source compiler, which optimizes models from frameworks like PyTorch, JAX, and ONNX for their hardware. Emphasizing transparency and developer empowerment, Tenstorrent champions open architectures and open-source software, aiming to give users ownership and flexibility over their AI silicon future.

Tenstorrent

About Tenstorrent

Core Capabilities

Main Tasks

Accelerating AI/ML Workloads

AI Model Training

AI Model Inference

Hardware-Software Co-design

Custom AI IP Development

MLIR-based Compilation

What this tool is best suited for

Shortlist Tenstorrent against top options

Key Features

MLIR-based Open-Source Compiler (TT-Forge)

RISC-V Based Custom AI Accelerators

Scalable & Sovereign AI Infrastructure

Use Cases

Large-scale AI Model Training and Inference

AI Research and Development Workstations

Custom AI Acceleration for Edge and Embedded Systems

Pros

Cons

Frequently Asked Questions

Reviews & Ratings

Write a Review

Blackhole® Accelerator Cards

TT-QuietBox™ Workstations

Tenstorrent Galaxy™ Servers

Specs

Core Tasks

Categories

Use Tenstorrent For

Tenstorrent vs Alternatives

Alternative Tools

Tensor.art

Parallel Domain

Unity AI

Graphcore

Habana

Kneron

Limitless AI

Mythic AI

Data Interface