RISC-V Performance Infra Engineer

Tenstorrent; California (MO); Hybrid; USD 100,000 - 500,000; Full time

Job summary

Tenstorrent seeks a talented engineer to build and run the performance infrastructure used by our RISC-V Software and Performance teams. You’ll implement benchmarking automation, data collection pipelines, and workload capture across silicon, FPGA/emulation platforms and models.

Hybrid role based in Austin, TX or Santa Clara, CA. We value collaboration and curiosity and provide growth for engineers at all levels.

Qualifications

• Build and maintain benchmarking automation and CI pipelines that run SPEC and real-world workloads.
• Own SimPoint generation and management infrastructure.
• Build performance-data collection into databases, with dashboards and visualization tools on top, so trends and regressions are visible without manual effort.
• Develop workload capture and replay tooling (QEMU- and CRIU-based checkpointing).
• Collaborate closely with performance and software engineers on both teams to understand their flows and remove the infrastructure friction from them.

Responsibilities

• Build and maintain benchmarking automation and CI pipelines that run SPEC and real-world workloads across RISC-V silicon, FPGA/emulation platforms and performance models.
• Own SimPoint generation and management infrastructure.
• Build performance-data collection into databases, with dashboards and visualization tools on top, so trends and regressions are visible without manual effort.
• Develop workload capture and replay tooling (QEMU- and CRIU-based checkpointing) so perf engineers can trace and analyze workloads reproducibly.
• Collaborate closely with performance and software engineers on both teams to understand their flows and remove the infrastructure friction from them.

Skills

Linux systems engineer

Python

shell scripting

CI (GitLab)

container experience

Automation at scale

CPU performance interest

Tools

QEMU

CRIU

Job description

Tenstorrent seeks a talented engineer to build and run the performance infrastructure used by our RISC-V Software and Performance teams. You’ll implement benchmarking automation, data collection pipelines, and workload capture across silicon, FPGA/emulation platforms and models.

Hybrid role based in Austin, TX or Santa Clara, CA. We value collaboration and curiosity and provide growth for engineers at all levels.