Anthropic
Performance Engineer, GPU
About this role
Anthropic seeks a GPU Performance Engineer to optimize large-scale GPU systems powering Claude and advanced language models. You'll develop custom kernels, distributed architectures, and performance innovations across the full hardware-software stack to maximize efficiency at unprecedented scale.
What you'll do
- Design and implement GPU kernel optimizations using CUDA, Triton, and tensor core techniques
- Architect distributed communication strategies for multi-node GPU clusters
- Optimize end-to-end training and inference pipelines for frontier language models
- Build performance modeling frameworks to predict and improve GPU utilization
- Profile production systems to identify and eliminate performance bottlenecks
- Collaborate with hardware vendors to influence future accelerator capabilities
What they're looking for
- GPU programming and optimization (CUDA, Triton, CUTLASS)
- ML frameworks and compilers (PyTorch, JAX, XLA, torch.compile)
- Distributed systems (NCCL, NVLink, collective communication)
- Kernel fusion and memory bandwidth optimization
- Low-precision quantization (INT8/FP8) and mixed-precision techniques
- Performance profiling tools (Nsight)
- Large-scale system design and cluster orchestration
- Model parallelism and distributed training infrastructure
Benefits
- $280,000–$850,000 annual salary
- Hybrid work policy (minimum 25% office presence)
- Work on state-of-the-art AI systems with real-world impact
- Collaborative environment with world-class researchers and engineers
- Opportunity to influence hardware and software evolution
Opens the application — the Jobs AI extension fills it for you. Set up autofill
Opens the official application on the employer’s site. No login required.
Anthropic
Anthropic builds Claude, an AI assistant, and is hiring for engineering roles across infrastructure, data systems, and security that support both AI research operations and the company's internal technology needs. The company seeks infrastructure engineers, systems integrators, data scientists, and security specialists to build production-scale systems for training data pipelines, financial operations, developer productivity measurement, research infrastructure, and server firmware security.
- Website
- anthropic.com
Likely interview questions
- Walk us through a specific GPU kernel optimization you've implemented in production. What was the performance bottleneck, and how did you measure the improvement?
- Describe your experience with CUDA or Triton kernel development. Have you worked with tensor core optimizations or custom operators?