Chandan Sharma

Chandan Sharma#

Chandan

Chandan Sharma is a Member of Technical Staff (MTS) at AMD, where he focuses on advancing the AI/ML software ecosystem for AMD Instinct GPUs, with an emphasis on model training, inference acceleration, and performance optimization.

Previously, he was with Fujitsu Research, where he optimized AI frameworks—including oneDAL, OpenBLAS, PyTorch, Llama.cpp, Scikit-Learn, NumPy, and Statsmodels—for the FUJITSU-MONAKA 2nm, 288-core ARM-SVE CPU, the successor to the Fugaku supercomputer. Earlier, he held roles at Sony and Sandlogic, working on edge AI solutions involving model optimization, quantization, and applications in computer vision, automation, and document intelligence.

Chandan holds an M.Tech in Computer Science & Engineering from IIT Dhanbad and is passionate about building scalable, energy-efficient AI systems at the intersection of hardware and software.

Posts by Chandan Sharma