Research Engineer – AI Systems

This is a fully remote position, open to applicants in United States.

📋 Description

• Create and implement high-performance kernels for Attention, Mixture of Experts (MoE), General Matrix Multiplication (GEMM), collective communication, and quantization.

• Enhance kernels for platforms such as NVIDIA, AMD, and AWS Trainium.

• Construct custom operators and optimize graphs using Neuron SDK, PyTorch/XLA, Torch Dynamo, and Neuron Compiler.

• Boost the performance of vLLM, SGLang, TensorRT-LLM, and tailored inference runtimes.

• Develop scalable distributed training and inference solutions across numerous accelerators.

• Contribute to open-source initiatives, disseminate technical insights, and engage with the developer community.

⛳️ Requirements

• Expertise in AI programming languages including Python and C++.

• Comprehensive understanding of GPU architecture and performance enhancement techniques.

• Familiarity with CUDA, Triton, ROCm/HIP, or AWS Neuron.

• Strong knowledge of AI frameworks (e.g., PyTorch, Dynamo, LMCache), model architectures, and profiling tools (e.g., Nsight, ROCm Profiler, or Neuron Profiler).

• Excellent problem-solving abilities and capacity to thrive in a collaborative, remote setting.

• A degree in computer science, engineering, or a related discipline is advantageous.

🏝️ Benefits

• Competitive salary with equity options.

• A flexible, remote working environment that promotes innovation and independence.