Ritesh Patel
Senior Deep Learning Architect at NVIDIA
Sacramento, California, United States
Summary
👤
Senior🎓
Top SchoolRitesh Patel is a Senior Deep Learning Architect with 15 years of experience optimizing ML performance across GPUs and large-scale clusters, currently focusing on end-to-end training performance for LLMs at NVIDIA. He has led performance model development and software-hardware co-design at Google and Intel, building predictive cost models for core ML ops (GEMM, convs, fusions) and influencing TPU/GPU architecture decisions. Ritesh combines low-level kernel implementation expertise with system-level profiling to drive tangible speedups—authoring cycle-accurate simulators and highly optimized compute shaders earlier in his career. His background includes building control systems for precision manufacturing and translating hardware behavior into actionable software optimizations, a blend that helps him spot non-obvious bottlenecks across the stack. Based in Sacramento, he brings deep practical experience in performance pathfinding, sharding/scheduling strategies, and operation fusion techniques that accelerate production-scale ML training.
15 years of coding experience
14 years of employment as a software developer
Master of Science (M.S.) Electrical and Computer Engineering, Master of Science (M.S.) Electrical and Computer Engineering at University of California, Davis
English, Gujarati
Github Skills (15)
transformers10
transformer-models10
large-language-models10
cuda9
data-parallel9
huggingface8
gpu7
python7
nvidia7
floating-point6
pytorch6
deep-learning6
machine-learning6
inference4
jax4
Programming languages (2)
Github contributions (5)
rapatel/TransformerEngine
Sep 2025 - Jun 2026
A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit floating point (FP8) precision on Hopper, Ada and Blackwell GPUs, to provide better performance with lower memory utilization in both training and inference.
Contributions:7 pushes, 2 branches in 9 months
rapatel/Megatron-LM
Dec 2025 - Jun 2026
Ongoing research training transformer models at scale
Contributions:23 pushes, 8 branches in 6 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.