Etched
Performance Tools Intern
San Jose · On-site · Internship · Posted Aug 3, 2026
About the job
Develop performance profiling and analysis tools for Etched’s custom ML accelerators, helping engineers and customers understand workloads and identify bottlenecks.
What you'll do
- Build performance analysis, profiling, data collection, and visualization tools for custom accelerators.
- Collect and analyze hardware-counter, execution-trace, memory, and system-performance data; correlate activity across system components.
- Collaborate with hardware, compiler, firmware, and inference engineers to improve performance analysis and developer productivity.
What they ask for
- Strong programming skills in C++ or Rust.
- Solid understanding of computer architecture, including processors or accelerators, memory hierarchies, and parallel programming.
- Experience or strong interest in low-level performance analysis, profiling, and optimization.
- Experience or strong interest in operating systems, compilers, firmware, drivers, or other low-level systems software.
Nice to have
- Familiarity with tools such as Nsight, VTune, Xprof, or Perfetto.
- Experience developing performance or debugging tools, with ML accelerator architectures, or with kernel-mode driver development.
Tools
C++, Rust, Python, Nsight, VTune, Xprof, Perfetto, PCIe
AI focus
Build tools to analyze and optimize performance on custom ML accelerators used for inference workloads.
Model-written summary · the original posting is authoritative