About Etched
Etched is building hardware for frontier intelligence. We co-design chips, racks, software, and manufacturing to deliver best-in-class throughput and latency across both prefill and decode workloads. Our first products are heavily focused on inference. Backed by hundreds of millions from top-tier investors and staffed by leading engineers, Etched is redefining the infrastructure layer for the fastest growing industry in history.
Job Summary
Join our team and take the lead in illuminating the performance landscape of our cutting-edge ML accelerator. We are seeking a highly skilled engineer to design and develop a sophisticated performance analysis tool, tailored specifically for our hardware. You will be instrumental in creating the essential tooling that enables our ML engineers and customers to understand workload behavior, identify performance bottlenecks, and unlock the full potential of our hardware, accelerating the most demanding ML applications in the world. This is a unique opportunity to shape performance analysis for novel hardware from the ground up.
During your internship, you may:
Build components of our performance analysis and profiling infrastructure.
Collect and analyze performance data from our custom ML accelerators, including hardware counters, execution traces, and memory behavior.
Develop tooling to trace host-side runtime activity, system behavior, and accelerator execution.
Help correlate performance events across CPUs, accelerators, storage, networking, and distributed workloads.
Build analysis and visualization tools that help engineers identify performance bottlenecks and optimize models.
Work alongside hardware, compiler, firmware, and inference engineers to understand performance challenges and develop tools that improve developer productivity.
Representative projects
Implement the data collection framework for hardware performance counters on a custom PCIe-based accelerator.
Develop a user-space service for low-overhead tracing of accelerator activity.
Design and build a correlated timeline view visualizing CPU API calls, driver submissions, PCIe transfers, and accelerator execution units.
Create an analysis pass to detect and quantify memory access inefficiencies or PCIe bandwidth saturation while transacting on a PCIe-attached accelerator.
You may be a good fit if you have
Strong programming skills in C++ or Rust. Experience with Python is a plus.
Solid understanding of computer architecture, including CPUs, GPUs or AI accelerators, memory hierarchies, and parallel programming.
Experience or strong interest in low-level performance analysis, profiling, and performance optimization.
Familiarity with performance analysis tools such as Nsight, VTune, Xprof, Perfetto, or similar tools is a plus.
Experience or strong interest in operating systems, compilers, firmware, drivers, or other low-level systems software.
Passion for understanding how complex systems behave under real workloads and building tools that help other engineers optimize performance.
Strong problem-solving skills and curiosity to learn quickly in a fast-paced engineering environment.
Strong candidates may also have experience with (Nice-to-have qualifications)
Direct experience developing performance analysis or debugging tools.
Experience with ML accelerator architectures (GPUs, TPUs, etc.).
Experience with kernel-mode driver development (Linux or Windows).
How we’re different
Etched believes in the Bitter Lesson. We are the first inference-focused frontier AI system, betting early on transformer and transformer-like architectures and on increasing model sizes. Our addressable market is the entirety of inference, unlike many of our competitors.
We are a fully in-person team in San Jose (Santana Row), and greatly value engineering skills. We do not have boundaries between engineering and research, and we expect all of our technical staff to contribute to both and work across disciplines as needed.