Back to Careers
Hardware Performance Engineer
San Francisco, CA (In-Person)
General Diffusion’s mission is to decouple intelligence from silicon. We believe the path to AGI requires a universal translation layer that makes compute fungible across any architecture—from H100s to TPUs to neuromorphic chips.
About the role
As a Hardware Performance Engineer, you will build HP1 (Hardware Profiler Agent), the sensory system of our OS. You will write the low-level probes that interrogate silicon for its true capabilities, bypassing marketing spec sheets to find the raw limits of memory bandwidth, FLOPs, and thermal headroom.
What you might work on
- Writing micro-benchmarks to reverse-engineer the undocumented behavior of new accelerators.
- Building the HP1 telemetry pipeline to stream real-time metrics from thousands of devices.
- Analyzing performance regressions across different driver versions and firmware updates.
- Collaborating with vendors (AMD, Intel, Cerebras) to debug silicon errata.
- Optimizing the "cold start" time of our profiling suite to run in milliseconds.
What we’re looking for
- Low-level systems programming experience (C, C++, Assembly).
- Experience with performance analysis tools (Nsight Systems, rocprof, VTune).
- Understanding of PCIe, NVLink, and interconnect topologies.
- Fearlessness in the face of undocumented hardware and closed-source drivers.
- A hacker mindset—you enjoy breaking things to see how they work.
Our culture
- Silicon Neutrality. We build for the world where compute is a commodity, not a monopoly.
- Radical Efficiency. We believe software bloat is an existential risk to AGI.
- Deep Work. We value long periods of uninterrupted focus over endless meetings.