Marc Koester

Staff at DuPont(TM)

Buffalo, New York, United States
email-iconphone-icongithub-logolinkedin-logotwitter-logostackoverflow-logofacebook-logo
Join Prog.AI to see contacts
email-iconphone-icongithub-logolinkedin-logotwitter-logostackoverflow-logofacebook-logo
Join Prog.AI to see contacts

Summary

🤩
Rockstar
🎓
Top School
Marc Koester is a Staff software engineer based in Buffalo, New York with 12 years of experience building high-performance systems. At DuPont he combines practical engineering with technical leadership, applying deep backend expertise to real-world industrial problems. He contributes to open-source compiler work, notably optimizing the ILGPU JIT for .NET to improve kernel generation, loop unrolling, and memory handling for NVIDIA GPUs. That low-level performance focus suggests a strong grounding in compilers, numerical code generation, and systems optimization beyond typical application development. Marc studied at Technische Universität Darmstadt and brings a pragmatic, performance-first mindset to cross-disciplinary engineering teams. He is the kind of engineer who bridges research-level optimization techniques and production-grade reliability.
code12 years of coding experience
bookTechnischen Universität Darmstadt
github-logo-circle

Github Skills (10)

compiler-optimization10
cuda10
loop-unrolling10
irr10
c-language10
cprogramming-language10
gpu10
optimisation10
intermediate-code10
optimization10

Programming languages (9)

C#HCLTypeScriptJavaC++JavaScriptGoHTML

Github contributions (5)

github-logo-circle
m4rs-mt/ILGPU

Jun 2017 - Dec 2022

ILGPU JIT Compiler for high-performance .Net GPU programs
Role in this project:
userBack-end Developer
Contributions:29 releases, 558 reviews, 1352 commits in 5 years 7 months
Contributions summary:Marc appears to be involved in optimizing the ILGPU JIT Compiler. Their contributions include implementing and refining arithmetic simplification strategies, improving loop analysis and unrolling, and extending the IR to support new types of operations. They worked on enhancing the efficiency and performance of kernel code generation, particularly for NVIDIA GPUs, by optimizing branches and memory handling.
amdcompilerkernelsgpgpu-computinggpgpu
m4rs-mt/ILGPU.SharpDX

May 2017 - Feb 2019

Contributions:5 commits, 13 pushes, 1 branch in 1 year 8 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.
Request Free Trial
Marc Koester - Staff at DuPont(TM)