Optimize the inference of model architectures including MHA, GQA, MQA, MLA, sliding-window attention, sparse attention, linear attention, and state space models.
Investigate model-system co-optimization techniques such as speculative decoding, state reuse, and computation reuse.
Build performance models and optimization strategies for long-context inference and agentic workloads, with a focus on throughput, time to first token, inter-token latency, memory efficiency, and deployment cost.
Collaborate closely with model, framework, kernel, library, system software, and GPU architecture teams to influence the design of future models, software platforms, and hardware architectures.
What we need to see
Pursuing an MS or PhD from a leading university in Computer Science, Computer Engineering, Electrical Engineering, Applied Mathematics, or a related field, with an expected graduation date in 2028.
Available to work as an intern for at least three months and at least four days per week.
Strong understanding of Transformer architectures and large language model inference, including core components such as attention, MoE, and KV cache management.
Experience with at least one large language model inference framework, such as TensorRT-LLM, vLLM, SGLang, or a comparable framework.
Proficiency in Python and PyTorch, with strong experimental design, data analysis, debugging, and problem-solving skills.
Experience with at least one GPU programming technology, such as CUDA C/C++, Triton, or a comparable parallel programming framework.
Understanding of GPU architecture, memory hierarchy, Tensor Cores, parallel computing, and performance analysis.
Familiarity with low-precision computing concepts, including quantization error, scaling, rounding, clipping, group or block quantization, and mixed-precision computation.
Ability to evaluate an optimization technique across model accuracy, kernel performance, and end-to-end inference performance.
Good communication, technical writing, organization, and cross-functional collaboration skills.
Ways to stand out from the crowd
Hands-on experience optimizing kernels or end-to-end workloads on NVIDIA GPUs.
Research or engineering experience in model quantization, KV cache optimization, efficient attention, speculative decoding, or long-context inference.
Contributions to inference frameworks, GPU libraries, compilers, or other relevant open-source projects.
Experience profiling and optimizing production-scale LLM inference workloads.
NVIDIA is committed to fostering a diverse work environment and is proud to be an equal opportunity employer. We highly value diversity among our current and future employees and do not discriminate in our hiring or employment practices on the basis of race, religion, color, national origin, gender, gender expression, sexual orientation, age, marital status, veteran status, disability status, or any other characteristic protected by law.
