Summary
Balamurugan Marimuthu is a senior software engineer with nine years of experience specializing in distributed LLM inference, model training optimizations, and graph/compiler infrastructure. He builds MLIR/LLVM passes for model distribution, transformer/attention kernel optimizations, and end-to-end performance tuning across CPU, GPU, FPGA and CGRA targets. His background spans deep learning systems, high-performance computing and computer architecture—skills honed through CMU graduate research and production roles at SambaNova and NVIDIA. Notably, he pairs low-level kernel and memory-layout engineering with compiler-driven graph transformations to push compute and interconnect utilization while reducing overheads. He also brings earlier VLSI and EDA expertise, giving him a rare hardware-aware lens on ML system design. Based in Pittsburgh, he combines research rigor with production impact across DL compilers and LLM inference stacks.
9 years of coding experience
7 years of employment as a software developer
Bachelor of Technology (B.Tech.), Electronics and Communication Engineering, Bachelor of Technology (B.Tech.), Electronics and Communication Engineering at Shanmugha Arts, Science, Technology and Research Academy
D.A.V BHEL School
Master of Science - MS, Electrical and Computer Engineering, Master of Science - MS, Electrical and Computer Engineering at Carnegie Mellon University