Vedaanta Agarwalla

As a senior deep learning software engineer at NVIDIA, Vedaanta focuses on accelerating GPU workloads with a current emphasis on optimizing attention kernels for both training and inference. His previous experience spans ResNet optimizations, GEMMs, and HPC for derivatives pricing in quantitative trading. Vedaanta holds a master’s degree in computer science from the University of Illinois Urbana-Champaign.
Avatar photo

Posts by Vedaanta Agarwalla

Data Center / Cloud Feb 25, 2026

Making Softmax More Efficient with NVIDIA Blackwell Ultra

LLM context lengths are exploding, and architectures are moving toward complex attention schemes like Multi-Head Latent Attention (MLA) and Grouped Query... 10 MIN READ