Ilia Sergachev

Zurich, Zurich, Switzerland

Join Prog.AI to see contacts

Summary

🤩

Rockstar

🎓

Top School

Ilia Sergachev is a researcher-engineer with 11 years of experience building commercial industrial and scientific devices end-to-end, combining analog/digital electronics, optics and embedded computing. With a PhD in Physics from the University of Basel and based in Zurich, he blends rigorous research skills with practical systems engineering. He is an active open-source contributor to high-profile ML and hardware projects—improving performance in XLA, TensorFlow and JAX and extending LiteX and PCIe cores for FPGA platforms. His work sits at the intersection of low-level hardware integration and high-performance software, from cuDNN/CUDA graph optimizations to FPGA memory and PLL tuning. Colleagues rely on him for hard-to-debug performance wins and for translating research ideas into reliable production hardware and firmware.

11 years of coding experience

University of Basel

German, French, Russian, English

Stackoverflow

Stats

1reputation

0reached

0answers

0questions

Github Skills (46)

verilog10

c-language10

xilinx10

python10

back-end-development10

gpu-programming10

machine-learning10

c1110

hardware10

ml10

mle10

c1710

kernel10

lit10

lite-server10

Programming languages (10)

C++CRustLLVMScalaVerilogMLIRJupyter Notebook

Github contributions (5)

openxla/xla

Sep 2022 - Jan 2023

A machine learning compiler for GPUs, CPUs, and ML accelerators

Role in this project:

Back-end Developer & Performance Engineer

Contributions:173 reviews, 54 commits, 194 PRs in 4 months

Contributions summary:Ilia contributed to the XLA compiler, focusing on performance optimizations and preventing inefficient computations. Their work involved generalizing logic to avoid redundant calculations in fusions, particularly related to expensive operations like reductions. They implemented and adjusted tests to ensure the correctness of these optimizations. The user also added support for CUDA device information retrieval, including L2 cache size and floating-point operation counts, for performance analysis.

compilercommunity-drivenmachine-learningmodular

enjoy-digital/litex

Jun 2019 - Jul 2022

Build your hardware, easily!

Role in this project:

Embedded Systems Engineer

Contributions:59 commits, 29 PRs, 12 comments in 3 years 2 months

Contributions summary:Ilia primarily focused on fixing and improving core aspects of the LiteX hardware build system. Their commits addressed issues related to interrupt handling, CSR and AXI bus integration, and support for features like incremental implementation in Xilinx Vivado. They demonstrated proficiency in modifying low-level system integration code, highlighting their understanding of hardware design and build processes. In addition, the user contributed to compilation warning fixes in the simulation modules.

vhdlsystem-on-chipedasochardware

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.

Request Free Trial