Maciej Dziuban is a pragmatic graphics and driver engineer with eight years of experience building and optimizing GPU drivers for Intel and AMD. He specializes in low-level compute and graphics stacks—contributing to Intel’s compute-runtime for Level Zero/OpenCL and developing AMD’s OpenGL, Vulkan UMD and video frameworks—where he couples architecture-specific code with careful performance tuning. Regularly using tools like RenderDoc, Radeon GPU Profiler and Windows Performance Analyzer, he focuses on correctness-first implementations and measurable CPU/GPU optimizations. His work spans hardware-aware changes (Skylake, Coffee Lake, Broadwell) and build-system adjustments, showing comfort across systems code, drivers and toolchains. Based in Gdańsk, he brings a disciplined, no-shortcuts ethos to complex debugging and feature delivery in production GPU software. An unusual strength is translating microarchitectural constraints into clean, maintainable driver abstractions that survive cross-vendor transitions.
8 years of coding experience
4 years of employment as a software developer
Informatyka, Informatyka at Gdańsk University of Technology
Intel® Graphics Compute Runtime for oneAPI Level Zero and OpenCL™ Driver
Role in this project:
Back-end Developer & Systems Architect
Contributions:300 commits, 4 PRs in 3 years 8 months
Contributions summary:Maciej made adjustments to the "intel/compute-runtime" repository, which is an Intel Graphics Compute Runtime for oneAPI Level Zero and OpenCL Driver, focusing on low-level driver and hardware interaction. The commits indicate work on adjusting the CapabilityTable, including extracting and implementing specific functionalities for the hw_info_config and utilizing image compression. The user’s work required modifying source files for different architectures such as Skylake, Coffee Lake, and Broadwell, implying architecture-specific driver development. They also worked on CMake build file and included the use of preemption.
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.