Tenstorrent
Semiconductors · RISC-V AI chip (Jim Keller)
- HQ
- Toronto, CA
- Founded
- 2016
- Website
- tenstorrent.com
Semiconductors · RISC-V AI chip (Jim Keller)
| Title | Stage | AI score |
|---|---|---|
| Sr. Engineer, Software - AI Compiler Sr. Engineer, Software - AI Compiler role at Tenstorrent focused on developing TT-Forge, an MLIR-based compiler for Tenstorrent hardware, optimizing AI models for training and inference. | Serve | 8 |
| Sr. Engineer, Kernel Development and Optimization Sr. Engineer, Kernel Development and Optimization at Tenstorrent, focusing on designing, implementing, and optimizing performance-critical kernels for AI hardware, including matrix multiplication and attention primitives. The role involves host-side orchestration, parallelization, developing benchmarks and tests, and collaborating with compiler, runtime, ML, and hardware teams to integrate kernels into production systems. Experience with C++, low-level software, concurrency, and data-driven optimization is required. | Serve |
| 7 |
| Sr Engineer, Server Inference The role focuses on developing software for state-of-the-art AI inferencing on Tenstorrent's hardware, including designing APIs, deploying workloads, and benchmarking inference speed. It involves optimizing end-to-end ML inference on custom silicon and building scalable software interfaces. | Serve | 7 |
| Experienced Technical Recruiter Experienced Technical Recruiter for an AI company, focusing on hiring deeply technical roles in AI, ML, systems, hardware, or infrastructure. The role involves full-cycle ownership, proactive sourcing, and acting as a trusted advisor to hiring managers. | — | 5 |
| Sr. Staff Engineer, Driver Tenstorrent is seeking a Sr. Staff Engineer, Driver to focus on the user-mode driver and interface layer for their AI hardware. This role involves designing and evolving high-performance APIs, defining driver interfaces, collaborating with kernel and hardware teams, driving performance engineering, and supporting external integrations, including in safety-critical use cases. | — | 0 |