GPU Performance Engineer | Experienced Hire
Core
Build highly optimized CUDA kernels for low-latency inference in quantitative trading workloads where off-the-shelf runtimes are insufficient.
Role type
Senior IC GPU Performance Engineer
Builds
Production-grade GPU implementations for compact neural networks, tree-based models, and structured inference workloads
Domain
Quantitative trading / High-performance computing
Deliverable
production ML models
Required skills
CUDA kernel optimization, C/C++ programming, GPU architecture knowledge (memory hierarchy, SIMT), numerical stability analysis, performance profiling
Preferred skills
PhD in quantitative field, linear algebra/probability background, experience with financial models, PTX-level tuning, familiarity with ONNX Runtime/TensorRT/Triton, knowledge of state space models (Mamba)
Technologies
CUDA, C/C++, PTX, ONNX Runtime, TensorRT, Triton, TVM, LightGBM, Mamba architectures
Responsibilities
Design and optimize custom CUDA kernels for latency-critical inference, analyze computational bottlenecks in research models, collaborate with researchers to translate math models into high-performance pipelines, profile and benchmark GPU performance, contribute to GPU architecture decisions
Seniority
Senior, hands-on IC
