AI Summary
Join a high-agency team building GPU virtualization software and a GPU cloud. Tackle complex systems problems and optimize performance in a C++ environment.
Key Highlights
Microsecond-level performance optimization in C++
Research and implementation of oversubscription, checkpointing, and distributed GPU clusters
Systems-level debugging and performance issue diagnosis in ML workloads
Technical Skills Required
Benefits & Perks
Competitive equity
Visa sponsorship
Relocation available
Job Description
🚀 Hiring: Software Engineer (C++ Systems) — GPU Virtualization (Onsite SF)
We’re a small, high-agency team building GPU virtualization software to make GPUs dramatically more efficient, plus a GPU cloud that gives us real production traction and feedback. This is an early, high-ownership role tackling hard systems problems end-to-end.
What you’ll work on- Microsecond-level performance optimization in a complex C++ virtualization library
- Research + implementation around oversubscription, checkpointing, and distributed GPU clusters
- Supporting new architectures across the stack
- Systems-level debugging in production
- Diagnosing performance issues in ML workloads
- Optional customer-facing work (as much or as little as you want)
- Strong C++ systems experience (low-level, performance-sensitive code)
- Deep understanding of concurrency, memory, and systems debugging
- Comfort owning production issues and driving fixes quickly
- Strong CS fundamentals and curiosity about how things work under the hood
- San Francisco (onsite)
- Expect 60–65 hrs/week (likely some weekend work)
- Relocation available
- $200K–$250K base + competitive equity
- Visa sponsorship available: TN / Canadian visas / H-1B transfers (no net new H-1Bs)
- 3-step process: behavioral + technical, then deeper evaluation (may include onsite/work trial depending on seniority)