The company is looking to scale its AI-infrastructure platform and serve a rapidly growing customer base by optimizing its core GPU virtualization stack. This involves improving microsecond-level performance in complex C++ systems and building impactful low-level GPU software.
Requirements
- Elite C++ expertise (Rust also sufficient but they will be working in C++).
- Experience optimizing NIC/C++ performance.
- Ability to trace performance issues across the stack.
- Experience working on low-level systems in production.
- Experience with compilers, networking protocols, or kernels.
Responsibilities
- Performance optimization of the C++ virtualization library.
- Research into oversubscription, checkpointing, and distributed GPU clusters.
- Supporting new architectures with deep understanding across the system.
- Systems-level debugging in production environments.
- Diagnosing performance issues in machine learning workloads.
Other
- On-site work policy
- Typically 60 - 65 hour work weeks which will likely require some weekend work.
- Relocation packages are available.
- 2+ years of experience in C++ systems engineering.
