Ardavan Pedram
San Francisco, California, United States
Summary
👤
SeniorArdavan Pedram is a Principal Architect at NVIDIA and a renowned leader in algorithm-architecture codesign for AI and HPC accelerators, bridging academia and industry in the San Francisco Bay Area. His NSF-funded work on parametric accelerator platforms and hardware-software co-design gave rise to the PRISM project, a platform for rapid investigation of efficient scientific computing and machine learning. At Samsung Electronics America, he directed next-generation AI/HPC accelerator codesign, led hybrid sparse core architectures for CNNs and LLMs, and conducted deep workload profiling to balance vector and tensor units. He serves as a Research Fellow and Director of the Prism Project at Stanford, shaping advanced ML accelerator research and education while maintaining strong industry impact. Earlier roles at Cerebras, Movidius, and UT Austin’s TACC involved building ML and linear algebra accelerators, demonstrating a consistent track record of turning theory into scalable, production-ready hardware solutions.
11 years of coding experience
Github Skills (25)
hardware9
accelerators8
machine-learning7
reduction3
manifold2
cluster-computing2
algebra2
eigenvalues2
sparse2
fenics2
lattice2
arbitrary-precision2
optimization2
kd-tree2
mpi1
Programming languages (2)
Github contributions (5)
cs217/cs217.github.io
Jan 2020 - Jan 2023
Course Webpage for CS 217 Hardware Accelerators for Machine Learning, Stanford University
Contributions:39 commits, 42 pushes in 3 years 1 month
stanfordstanford-universityhardware-acceleratorsmachine-learningaccelerators
LAProc/LAP_SIM
Dec 2014 - Dec 2014
Contributions:13 commits in 2 days
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.