GPU Performance tooling engineer
RivosFull Time
Junior (1 to 2 years)
Key technologies and capabilities for this role
Common questions about this position
Strong candidates need deep experience with GPU programming and optimization at scale, the ability to navigate complex systems from hardware interfaces to high-level ML frameworks, and a track record of delivering transformative GPU performance improvements in production ML systems.
Preferred experience includes GPU Kernel Development with CUDA, Triton, CUTLASS; ML Compilers & Frameworks like PyTorch/JAX internals; Performance Engineering such as kernel fusion and profiling with Nsight; Distributed Systems with NCCL and NVLink; Low-Precision techniques; and Production Systems for large-scale training.
Anthropic has a quickly growing team of committed researchers, engineers, policy experts, and business leaders focused on building safe and beneficial AI, with an emphasis on collaborative problem-solving, pair programming, and thriving in ambiguous environments.
Strong candidates are impact-driven, passionate about delivering measurable performance breakthroughs, excited to shape the future of AI infrastructure with world-class teams, enjoy collaborative problem-solving, and care about the societal impacts of their work.
This information is not specified in the job description.
Develops reliable and interpretable AI systems
Anthropic focuses on creating reliable and interpretable AI systems. Its main product, Claude, serves as an AI assistant that can manage tasks for clients across various industries. Claude utilizes advanced techniques in natural language processing, reinforcement learning, and code generation to perform its functions effectively. What sets Anthropic apart from its competitors is its emphasis on making AI systems that are not only powerful but also understandable and controllable by users. The company's goal is to enhance operational efficiency and improve decision-making for its clients through the deployment and licensing of its AI technologies.