2 часа назад
CUDA Engineer (AI)
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
CUDA Engineer (AI): Writing and optimising low-level GPU code for transformer inference workloads with an accent on custom CUDA kernels, memory hierarchies, and performance profiling. Focus on kernel fusion, quantisation-aware and mixed-precision computation, autoregressive decoding, and maximising throughput across GPU architectures.
Location: Remote; London, England, United Kingdom and United States
Company
is an energy startup developing integrated energy generation, hardware, grid infrastructure, real-time power trading, distributed energy systems, and high-performance compute infrastructure for AI workloads.
What you will do
- Write and optimise custom CUDA kernels for transformer inference operations.
- Profile kernels and eliminate occupancy, memory throughput, and warp divergence bottlenecks.
- Apply kernel fusion, optimise memory access patterns, and manage GPU memory hierarchies.
- Implement quantisation-aware and mixed-precision kernels, plus caching for autoregressive decoding.
- Tune kernel launch configurations and benchmark performance against existing baselines.
- Write regression tests, maintain internal CUDA libraries, and contribute to coding standards and documentation.
Requirements
- 4+ years of production CUDA development experience with shipped performance-critical kernels.
- Deep understanding of GPU microarchitecture, including warps, occupancy, register pressure, and memory hierarchy.
- Strong CUDA C++ skills, including streams and asynchronous execution.
- Hands-on experience profiling compute-bound and memory-bound bottlenecks.
- Experience with kernel fusion, memory coalescing, warp-divergence avoidance, quantised kernels, and mixed-precision arithmetic.
- Strong understanding of parallel algorithm design and numerical precision trade-offs.
Nice to have
- Experience with transformer or attention kernels and autoregressive decoding.
- Experience building high-performance GPU libraries from scratch.
- HPC, latency-critical performance engineering, or multi-GPU and multi-node kernel optimisation experience.
- Ability to read PTX/SASS to validate kernel efficiency.
Culture & Benefits
- Remote work with global hiring; benefits vary by location.
- Competitive salary with equity eligibility.
- Biannual bonus scheme.
- Fully expensed technology matched to work needs.
- Private health insurance.
- Breakfast and dinner allowance for office-based employees.
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 →