8 дней назад
GPU Performance Engineer (Machine Learning)
Вакансия из Hirify Global, списка международных tech-компанийДля мэтча и отклика нужен Plus
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Описание вакансии
Текст:
TL;DR
GPU Performance Engineer (CUDA/Machine Learning): Building highly optimized CUDA kernels and production-grade GPU implementations for low-latency inference across neural networks, tree-based models, and other structured workloads with an accent on kernel optimization, memory layouts, and hardware-aware execution. Focus on profiling and benchmarking GPU performance, improving inference latency and throughput, and translating quantitative models into efficient compute pipelines.
Location: Onsite in Bala Cynwyd, Philadelphia Area, Pennsylvania, United States
Company
is a global quantitative trading firm using machine learning, scientific research, and advanced technology to develop systematic trading strategies.
What you will do
- Design, implement, and optimize custom CUDA kernels for latency-critical inference workloads.
- Develop fine-grained GPU implementations tailored to neural networks, tree-based models, and other structured model architectures.
- Analyze quantitative research models and computational bottlenecks to identify parallelization and hardware-efficiency opportunities.
- Translate mathematical models into production-grade, high-performance compute pipelines in collaboration with quantitative researchers.
- Optimize inference through kernel tuning, memory-layout design, execution strategies, I/O optimization, and precision tradeoffs.
- Profile and benchmark GPU performance, improve production latency and throughput, and contribute to GPU architecture decisions.
Requirements
- Strong proficiency in writing and optimizing CUDA kernels.
- Solid programming experience in C/C++.
- Deep understanding of GPU architecture, including memory hierarchy, SIMT execution, occupancy, and latency/throughput tradeoffs.
- Ability to reason about numerical stability, precision, performance tradeoffs, and hardware-efficient model design.
- Strong problem-solving skills and comfort working with low-level systems.
Nice to have
- PhD in mathematics, physics, computer science, engineering, or a related quantitative field.
- Background in linear algebra, probability, numerical methods, or scientific computing.
- Experience with quantitative research teams, financial models, or real-world inference optimization beyond baseline frameworks and libraries.
- Familiarity with PTX-level behavior, tensor cores, architecture-specific tuning, ONNX Runtime, TensorRT, Triton, TVM, or similar systems.
- Experience with neural networks, LightGBM, Mamba architectures, kernel fusion, custom operators, model compilation, or graph-level optimization.
Culture & Benefits
- Intellectually driven and highly collaborative quantitative trading environment.
- Close collaboration among researchers, engineers, and traders.
- Opportunity to solve complex problems involving global markets, machine learning, and advanced quantitative research.
- Immediate start availability.
Будьте осторожны: если работодатель просит войти в их систему, используя iCloud/Google, прислать код/пароль, запустить код/ПО, не делайте этого - это мошенники. Обязательно жмите "Пожаловаться" или пишите в поддержку. Подробнее в гайде →