CUDA Engineer

Fuse Energy
London
Workplace: RemoteFull timeFunction: Software EngineeringExperience: 4+ yearsSkills: []

Design and optimize low-level CUDA code powering transformer inference workloads. Build custom CUDA kernels, profile and remove bottlenecks across occupancy, memory throughput, and warp divergence, and apply techniques like kernel fusion, memory-hierarchy tuning, quantization-aware and mixed-precision arithmetic. Implement efficient caching for autoregressive decoding, benchmark against baselines, and maintain internal CUDA libraries with tests to ensure performance and correctness.

Loading

Loading job details...

Preparing the role view and application actions.

FursaFursa
Fuse Energy
Fuse Energy
1 month ago

CUDA Engineer

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI
Last checked: 20 hours agoStatus: Live

Job Summary

Design and optimize low-level CUDA code powering transformer inference workloads. Build custom CUDA kernels, profile and remove bottlenecks across occupancy, memory throughput, and warp divergence, and apply techniques like kernel fusion, memory-hierarchy tuning, quantization-aware and mixed-precision arithmetic. Implement efficient caching for autoregressive decoding, benchmark against baselines, and maintain internal CUDA libraries with tests to ensure performance and correctness.
Location: London
Workplace: Remote
Employment Type: Full time
Job Function: Software Engineering
Seniority: Mid level

Key Responsibilities

  • •Write and optimize custom CUDA kernels for transformer inference operations.
  • •Profile kernels to identify and eliminate bottlenecks in occupancy, memory throughput, and warp divergence.
  • •Apply kernel fusion and optimize memory access patterns to improve throughput and latency.
  • •Implement quantization-aware kernels, mixed-precision arithmetic, and caching for autoregressive decoding.
  • •Benchmark against baselines, tune kernel launches for target GPU architectures, and add tests and documentation for CUDA libraries.

Pay and Benefits

Perks:EquityAnnual BonusMeal Allowance

Key Requirements

  • •4+ years writing production CUDA code and shipping 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 to diagnose compute-bound vs memory-bound bottlenecks.
  • •Experience with kernel fusion, memory coalescing, warp divergence avoidance, and quantised/mixed-precision kernels.
Experience:4+ years
Tech Stack:CUDACUDA C++GPUCUDA kernelsPTXSASS

Company Brief

Fuse Energy
Develops and delivers renewable energy solutions, focusing on clean power projects and services to enable decarbonization and grid flexibility for commercial and utility customers.
Industry: Renewable Energy
Website