Tenstorrent, the company led by legendary chip architect Jim Keller, the brains behind AMD's Zen architecture and Tesla's original autonomous chip, has launched its first hardware. Grayskull is a RISC-V alternative to GPUs that is designed to be easier to program and scale and reportedly excels at handling runtime sparsity and conditional computation.
Following this, Tenstorrent has also introduced its Grayskull-powered DevKits: the standard Grayskull e75 and the more powerful Grayskull e150. Both are inference-only hardware designed for AI development and come with TT-Buda and TT-Metalium software. The first is for running models immediately, while the second is for users who want to customize their models or write new ones.
The Santa Clara-based technology company's landmark launch comes on the heels of a partnership with Japan's Cutting Edge Semiconductor Technology Center (LSTC). Tenstorrent's RISC-V and IP Chiplet will be used to build a next-generation 2nm AI accelerator, with the ultimate goal of revolutionizing AI performance in Japan.
By the power of Grayskull!
The Grayskull e75 model is a low-profile, half-length PCIe Gen 4 board with a single Grayskull processor running at 75W. The more advanced e150 model is a standard-length, 3/4-height PCIe Gen 4 board containing a Grayskull processor that runs at up to 200W and balances power and performance.
Tenstorrent processors comprise a network of cores known as Tensix Cores and come with network communication hardware so they can communicate with each other directly over networks, rather than through DRAM.
Grayskull DevKits support a wide range of models, including BERT for natural language processing tasks, ResNet for image recognition, Whisper for speech recognition and translation, YOLOv5 for real-time object detection, and U-Net for image segmentation .
The Grayskull e75 and e150 DevKits are now available for purchase at $599 and $799, respectively.