Summary
Muhammed Ozturk is an ML and HPC-focused researcher-engineer with 11 years of experience, currently a Member of Technical Staff on AMD’s ML Framework team and a PhD candidate in Computer Science at the University of Utah. He designs and optimizes performance-critical tensor and ML kernels across Kokkos, SYCL, CUDA/HIP, and specialized accelerators including Cerebras CS-2 and AMD ROCm. His work spans hardware/software co-design for LLM and HPC kernels, with collaborations and internships at AMD Research, Intel, IBM, Berkeley Lab and Ohio State, and publications at top venues. Muhammed combines deep systems-level optimization (loop- and architecture-aware GPU kernels) with practical runtime and library integration (MIOpen, CK), targeting both performance and power efficiency. Notably, he has hands-on experience minimizing data movement on a 2D mesh AI accelerator of 850,000 processing elements and applying tensor factorization to compress transformers. Based in California, he bridges academic research and production ML stack development to accelerate scientific discovery.
11 years of coding experience
4 years of employment as a software developer
The University of Utah
Lisans Derecesi, Computer Science and Engineering, Lisans Derecesi, Computer Science and Engineering at Bilkent University
COLLEGE, Japanese, COLLEGE, Japanese at ABK COLLEGE - The Asian Student Cultural Association - Tokyo/Japan
Bachelor's degree-Exchange, Computer Science, Bachelor's degree-Exchange, Computer Science at National University of Singapore
Doctor of Philosophy - PhD (Transferring to UofU, Advisor Changed), Computer Science and Engineering, Doctor of Philosophy - PhD (Transferring to UofU, Advisor Changed), Computer Science and Engineering at The Ohio State University
İngilizce, Japanese