Graphics Software Engineer. Currently at Plastic, previously at AMD, Intel, Microsoft, gamedev, television... Developer of VMA, D3D12MA, D3d12info projects.
New Radeon Developer Tool Suite adds RadeonDeveloperPanelCLI allowing cmd-line control over all features: profiling, memory tracing, GPU crash capture, driver experiment toggles, etc. IMHO that's a big thing for test automation and possibly for agentic AI!
gpuopen.com/learn/radeon-dev…
Vulkan Memory Allocator new release 3.4.0 with many additions, improvements, and fixes. If you use code from the "master" branch, you already have them, but if you rely on the official Releases, you should upgrade. This one gathers >1 year of development.
github.com/GPUOpen-Libraries…
DirectX 12 Agility SDK 1.721.0-preview has been released. Interesting to see texture layouts based on GUIDs added, next to the existing `UNKNOWN`, which may help texture memory compatibility between GPU and CPU or between different GPUs.
devblogs.microsoft.com/direc…
I implemented multiple fixes & improvements in Vulkan/D3D12 Memory Allocator libraries, on my fork for now. If you use any of these, please give it a try so we validate they work fine before I turn them into official releases.
github.com/sawickiap/VulkanM…github.com/sawickiap/D3D12Me…
Attending #digitaldragons2026 conference. It's always inspiring, with interesting and funny discussions and thoughts. For example, on the pre-party we concluded that soon we should start calling "craft code" (like we have craft beer) a code written by human developers not AI 😀
On the first talk, one about rendering optimizations, the speaker (Jan Mróz) compared throughput and bottlenecks between GPU blocks to coveyor belts in Factorio and Satisfactory 😀
New DirectX 12 Agility SDK. Just preview version, but I hope these additions will reach the retail version soon. I'm excited about ID3D12CommandListAsyncCommands::CopyBuffer/TextureRegions, clears, and FillBuffers - finally available without implicit barriers around them!
Now available: Shader Model 6.10 and Agility SDK 1.720 public preview! This drop introduces first‑class matrix ops in shaders for ML/neural rendering workloads, new Batched Asynchronous Command List APIs and more! Details: devblogs.microsoft.com/direc…devblogs.microsoft.com/direc…
[Polish] Kto z Warszawy i okolic, w środę 2026-03-25 zapraszam na KNTG Polygon: wykład Michał "Lap-Luck" Biesek: Programowanie GPU i Compute shadery w Godot, a potem mój: Jak działa GPU? Architektura i programowanie współczesnych kart graficznych facebook.com/events/15924682…
A wacky idea for a game (an indie game or a game jam entry) that makes you feel like like an LLM: You're shown a short chat session with back-and-forth with a "user". Your task is to provide the next word of a response. Just one word. Then, you're switched to a different chat. 🧵
This would simulate how LLMs work: no chance to form a persistent memory of a chat, just general knowledge you have the current context window = generating next token. The difficulty of this would also demonstrate why these calculations are so expensive.
Not sure if the “user” side should be other players or fixed scenarios. But surely, LLM responses should be scored by the quality/correctness as well as speed.
Looking forward to seeing what DX12 news will be announced at GDC! I won't attend this year, but hopefully they'll share it publicly soon after. I'm especially excited about updates to GPU tools (PIX).
GDC 2026 is almost here, and we’d love to see you there next week! The DirectX team has a lot to share, with 4 sessions spanning a huge range of topics! For more on our talks, check out devblogs.microsoft.com/direc…
I developed a skill for Agentic AI that can query for parameters of the GPU installed in the local system, using vulkaninfo and D3d12info tools. Any feedback is welcome.
github.com/sawickiap/gpu-inf…