Yotta Labs
Research Engineer Intern - AI Systems
United States +3 · Remote · Internship · Posted Aug 2, 2026
- Location
- United States
Hong Kong
Canada
Singapore - Experience
- 12–16 week internship; no prior years of experience specified.
- Education
- Currently pursuing a BS, MS, or PhD in Computer Science, Computer Engineering, or a related field.
About the job
Develop and optimize AI kernels and LLM systems for heterogeneous accelerators, taking a scoped project from design through profiled code on real hardware.
What you'll do
- Implement and optimize kernels for Attention, GEMM, MoE, and quantization on NVIDIA, AMD, or AWS Trainium.
- Build custom operators and improve inference performance through kernel fusion, scheduling, KV-cache, and memory optimizations.
- Benchmark and diagnose performance regressions, then ship tested and documented code to production or open-source projects.
What they ask for
- Currently pursuing a BS, MS, or PhD in Computer Science, Computer Engineering, or a related field.
- Programming skills in Python and familiarity with C++.
- Understanding of GPU/accelerator fundamentals, including memory hierarchy, parallelism, and occupancy.
- Experience writing CUDA, Triton, ROCm/HIP, or Neuron kernels; class and personal projects count.
Nice to have
- Contributions to open-source AI infrastructure projects such as vLLM, SGLang, PyTorch, or Triton.
- Familiarity with LLM inference internals, performance profiling, or publications at the listed systems and architecture conferences.
Tools
NVIDIA GPUs, AMD GPUs, AWS Trainium, CUDA, Triton, ROCm/HIP, Neuron SDK, PyTorch/XLA
AI focus
AI systems and LLM infrastructure optimization, including accelerator kernels, distributed training and inference, and model-serving performance.
Model-written summary · the original posting is authoritative