Makesh Tarun Chandran

Makesh Tarun Chandran is a senior deep learning performance engineer at NVIDIA. He works on the co-design of generative AI models and GPU architecture, with a focus on the performance and efficiency of large-scale AI inference. He holds a master's degree in computer science and engineering from Pennsylvania State University.
Avatar photo

Posts by Makesh Tarun Chandran

Developer Tools & Techniques

Co-Designing Sparse Attention for Efficient Long-Context Inference

Agentic workloads, long documents, and large codebases are pushing context lengths from tens of thousands of tokens to hundreds of thousands. At those lengths,... 16 MIN READ