AMD GPUOpen

AMD GPUOpen

@gpuopen.bsky.social

We're the #gamedev home for AMD FSR™, Radeon™ Developer Tool Suite, VMA, AMF, & many more + developer docs.

1.5K followers35 following107 posts

AMD GPUOpen
@gpuopen.bsky.social

The AMD Radeon Developer Tool Suite just got a major update! 🛠️ 🔍 Radeon GPU Analyzer (RGA) 2.15 adds source-to-ISA correlation for Code Objects. 🖥️ Adds AMD CDNA 5 (Instinct MI400 Series), AGPR usage info, & DX12 Mesh Shader support. More on RGA 👇 gpuopen.com/learn/rga-so...

gpuopen.comSource-to-ISA correlation for pre-compiled AMD GPU code objects with RGA 2.15Learn how AMD Radeon GPU Analyzer 2.15 correlates pre-compiled AMD GPU code objects back to source lines, making ISA analysis and optimization far easier.
AMD GPUOpen
@gpuopen.bsky.social

Prepping your game for Microsoft Advanced Shader Delivery support on AMD Radeon GPUs? 🧩 Build, test, deploy — the 3 steps. ⚙️ Use our D3D12 State Object Compiler for testing: build a PSDB to validate before deploying. Get it ⬇️ gpuopen.com/advanced-sha...

gpuopen.comAdvanced Shader Delivery State Object CompilerTo help prepare a game for Advanced Shader Delivery support on AMD graphics hardware, developers can use the AMD Direct3D® 12 State Object Compiler plugin.
AMD GPUOpen
@gpuopen.bsky.social

More on our tetrahedral cage research: the numbers behind the demo video. 🌳 📉 <2GB of BVH memory, vs. up to 80GB for conventional animated BVHs. ⚡ 500M+ animated triangles ray traced per frame at 60+ FPS. See the breakdown ⬇️ gpuopen.com/learn/how-te...

gpuopen.comHow tetrahedral cages significantly reduce BVH memory usageThe AMD tetrahedral cage technique ray traces 25,000 independently animated plants at 60+ FPS while slashing BVH memory and update costs for massive dynamic scenes.
AMD GPUOpen
@gpuopen.bsky.social

ADLX 2.0 is here, connecting AI agents to AMD graphics hardware. 🐍 New Python bindings for AI developers. 🤖 MCP servers for structured AI-to-GPU communication. ⚙️ Query, monitor & tune AMD graphics through natural language. Get started ⬇️ gpuopen.com/learn/adlx-2...

gpuopen.comADLX 2.0: Extending graphics control to AI agents and agentic appsAMD ADLX 2.0 adds AI extensions and MCP servers to help developers build intelligent apps that can monitor, manage, and optimize AMD graphics hardware.
AMD GPUOpen
@gpuopen.bsky.social

Stable lighting, frame after frame: new research on a one-step diffusion model for GI, presented at #ECCV2026. 💡 Geometrically aligned, temporally stable lighting. 🧠 Aligns well with real-world features & illumination. Read more ⬇️ gpuopen.com/learn/tempor...

gpuopen.comTemporally stable generative illumination with a one-step diffusion modelA generative method for real-time global illumination using a single-step latent diffusion model, delivering stable, high-quality lighting without costly iterative processing.
AMD GPUOpen
@gpuopen.bsky.social

Our tetrahedral cage research just got a demo video! 🎥 🔺 Decouples animation cost from triangle count for massive real-time ray tracing. ⚡ See tetrahedral cages used to animate plants in a large-scale terrain at 60+ FPS. Watch it ⬇️ gpuopen.com/learn/ray-tr...

gpuopen.comRay tracing massive amounts of animated geometry using tetrahedral cagesAnimate compact tetrahedral cages and reuse static mini-BLASes to ray-trace hundreds of millions of triangles in real time, dramatically cutting per-frame update and memory costs for dense foliage, gr...
AMD GPUOpen
@gpuopen.bsky.social

AMD FSR & Radeon Anti-Lag 2 plugins updated for #UnrealEngine 5.8! 🆙 FSR Upscaling 4.1.1, now with Radeon RX 7000 Series support. 🎞️ FSR Frame Generation 4.0.1. ⚡ Anti-Lag 2 also updated for UE 5.8. Upgrade your #UE5 project ⬇️ gpuopen.com/learn/amd-fs...

gpuopen.comAMD FSR plugin updated for Unreal Engine 5.8 - AMD GPUOpenThe updated AMD FSR Unreal Engine plugin brings ML-powered upscaling and frame generation to Unreal Engine 5.8, now extending FSR Upscaling support to AMD Radeon RX 7000 Series GPUs.
AMD GPUOpen
@gpuopen.bsky.social

The AMD Radeon Developer Panel (RDP) is now open source, joining others in our Developer Tool Suite on GitHub! 🔍 Greater transparency into how it works. 🛠️ Contribute improvements & integrate into custom workflows. Explore the code ⬇️ github.com/GPUOpen-Tool... #opensource

github.comGitHub - GPUOpen-Tools/radeon_developer_panel: The Radeon Developer Panel (RDP) is a software tool that allows users to capture RGP profiles, RMV traces, RRA scenes, and RGD crash analysis dumps on Ra...The Radeon Developer Panel (RDP) is a software tool that allows users to capture RGP profiles, RMV traces, RRA scenes, and RGD crash analysis dumps on Radeon GPUs. - GPUOpen-Tools/radeon_developer_...
AMD GPUOpen
@gpuopen.bsky.social

Off-screen light not left in the dark: our latest research intros a lightweight model for real-time indirect illumination. 💡 Uses reflective shadow maps to capture off-screen geometry. 🧠 Generalizes scenes never seen during ML training. More info ⬇️ gpuopen.com/learn/lightw...

gpuopen.comLightweight attention-based indirect illumination - AMD GPUOpenIntroducing a lightweight, attention-based neural GI model that reconstructs off-screen indirect illumination using reflective shadow maps, generalizing to scenes not seen during training.
AMD GPUOpen
@gpuopen.bsky.social

MiniDXNN, our open-source MLP library for #DirectX12, is now at v0.4.0! 🔄 Now built on D3D12 Linear Algebra (SM 6.10 preview). 🖼️ New input encoding for neural texture compression. 🎛️ Interactive GUI app trains a texture live on screen. Get started ⬇️ gpuopen.com/learn/minidx...

gpuopen.comMiniDXNN v0.4.0: Interactive neural texture compression on DirectX 12 - AMD GPUOpenMiniDXNN v0.4.0 introduces D3D12 Linear Algebra (SM 6.10) support, input encodings and neural texture compression, plus a real-time GUI app that trains and visualizes GPU-accelerated MLPs on DirectX® ...
AMD GPUOpen
@gpuopen.bsky.social

Smooth surfaces, fewer triangles: new AMD research on “GPU view-adaptive crack-free subdivision of Bézier surfaces”! 🔺 Triangles concentrated where curvature & screen size demand it. ⚡ Built on GPU work graphs, no manual sync needed. Learn more ⬇️ gpuopen.com/learn/gpu-vi...

gpuopen.comGPU view-adaptive crack-free subdivision of Bézier surfaces - AMD GPUOpenLearn how fast, crack-free GPU work graph subdivision for bicubic Bézier surfaces dramatically reduce triangle counts while simplifying implementation and matching hardware-tessellation quality.
AMD GPUOpen
@gpuopen.bsky.social

Post-mortem GPU crash debugging with LLMs is here! 🔍 New AMD Radeon GPU Detective (RGD) MCP Server: 🤖 Connect LLMs to the AMD crash analysis pipeline. ⚡ From crash dump to root cause in minutes. 🔧 Proposes targeted source-code fixes. Read more ⬇️ gpuopen.com/learn/post-m...

gpuopen.comPost-mortem GPU crash debugging with LLMs - AMD GPUOpenThe new AMD RGD MCP Server connects LLM agents to AMD's GPU crash analysis pipeline, turning a single prompt into root-cause analysis and source-code fix suggestions.
AMD GPUOpen
@gpuopen.bsky.social

🔥 AMD FSR "Redstone" SDK 2.3 is now available! 🔥 The big news: FSR Upscaling 4.1.1 for AMD Radeon RX 7000 Series GPUs! 🟥 FSR Frame Generation 4.0.1 - improved input processing. 🟥 FSR Ray Regeneration 1.2.0 - new denoising options. More details ⬇️ gpuopen.com/learn/amd-fs...

gpuopen.comAMD FSR Upscaling 4.1 RDNA 3 Support Now Available in FSR SDK 2.3 Update - AMD GPUOpenAMD FSR "Redstone" SDK 2.3 brings ML-powered FSR Upscaling 4.1.1 to AMD Radeon RX 7000 Series GPUs, along with Frame Generation 4.0.1 and Ray Regeneration 1.2 improvements for RDNA 4 GPUs.
AMD GPUOpen
@gpuopen.bsky.social

Congratulations to our AMD Developer Community on turning 1! 🎉 🙏 Huge thanks to everyone on Discord & our Community Forums. You’ve helped shape what comes next from AMD for developers. 🔁 See what we’ve built together in our Year 1 recap! Join now 👉 gpuopen.com/devcommunity...

AMD GPUOpen
@gpuopen.bsky.social

The AMD Radeon Developer Tool Suite 🛠️ just got a major update! 🔍 View HLSL shader source code directly in RGP. 📌 Extended PIX Marker support in RGP & RGD. ⌨️ New RDP command-line interface for automated pipelines. Full details ⬇️ gpuopen.com/learn/radeon...

gpuopen.comNew AMD Radeon Developer Tool Suite update brings shader source code, Extended PIX Markers, and command-line capture - AMD GPUOpenThe new AMD Radeon Developer Tool Suite release delivers RGP 2.7 with shader source code viewing, instruction‑level divergence metrics, and Extended PIX Marker support, expanded hardware compatibility...
AMD GPUOpen
@gpuopen.bsky.social

New 3-part #WMMA guide for AMD RDNA 4 architecture GPUs! 🧮 Fused GEMMs for deep learning acceleration. ⚡ Wide WMMA for max memory bandwidth utilization. 🔄 In-register matrix transpose techniques. Read all three parts 👇 gpuopen.com/learn/wmma-g...

gpuopen.comWMMA guide for AMD RDNA 4 architecture GPUs - part 1 - AMD GPUOpenPractical guide to fusing GEMMs on AMD RDNA™ 4 architecture, covering WMMA layout, a transpose-by-swapping A/B technique, HIP sample code, and hipBLAS-verified results used in Llama.cpp.
AMD GPUOpen
@gpuopen.bsky.social

AMD support for the Microsoft Agility SDK 721 preview is here! New AMD Developer Preview Edition driver for: 🔗 Partial graphics programs. 💾 UAVs of Depth Stencil resources. More on the SDK update ⬇️ devblogs.microsoft.com/directx/anno... Download the driver ⬇️ www.amd.com/en/resources...

devblogs.microsoft.comAnnouncing AgilitySDK 721 Preview and an addition to LinAlg in Shader Model 6.10 - DirectX Developer BlogOverview Today, we are pleased to announce Agility SDK 1.721-preview with new D3D features along with DXC 1.10.2605.4 adding VectorAccumulate to LinAlg in the existing Shader Model 6.10 preview.  The ...
AMD GPUOpen
@gpuopen.bsky.social

AMD Schola v2.1 is now available! Train RL agents in #UnrealEngine with deeper integration & expanded scale. 🌲 StateTree integration for training & inference. ☸️ Kubernetes support for distributed training. 📊 Stronger Minari workflows. Learn more 👇 gpuopen.com/learn/announ...

gpuopen.comAnnouncing AMD Schola v2.1: state trees, scale, and a richer training stack - AMD GPUOpenAMD Schola v2.1 deepens Unreal Engine integration, adding StateTree support, Kubernetes-oriented distributed training, stronger Minari workflows, and much more to streamline training and inference at ...
AMD GPUOpen
@gpuopen.bsky.social

Microsoft Advanced Shader Delivery now supports more AMD GPUs! Enabled with AMD Software: Adrenalin Edition 26.5.2: 🎮 AMD RDNA 3, 3.5, & 4 GPUs. ⚡ Reduce shader stutter & load times by up to 95%. Learn how to add ASD to your game via AgilitySDK 👇 devblogs.microsoft.com/directx/adva...

devblogs.microsoft.comAdvanced Shader Delivery expands Public Preview with AMD - DirectX Developer BlogAdvanced Shader Delivery expands to Windows 11 PCs with AMD GPUs, cutting Forza Horizon 6 load times by 95% and reducing shader stutter via Xbox Insider.
AMD GPUOpen
@gpuopen.bsky.social

AMD is partnering with Samsung to develop a multivendor DGF extension for Vulkan! Dense Geometry Format (DGF) is AMD's open standard: 🧊 Block-based geometry compression for ray tracing. 🔓 Open-source toolchain & Vulkan extension. Read more ⬇️ gpuopen.com/learn/amd-dg...

gpuopen.comAMD DGF: An Open Geometry Compression Standard - AMD GPUOpenAMD is partnering with Samsung on a multivendor Vulkan extension for Dense Geometry Format (DGF) to help enable dramatically smaller geometry, reduced memory/latency for ray-traced real‑time 3D, and e...
AMD GPUOpen
@gpuopen.bsky.social

And there's more! 🧵 Introducing DGF SuperCompression (DGFS), a new feature in the AMD DGF SDK v1.2: 📦 Smaller disk footprint over DGF. 🔄 Common asset format for DGF & non-DGF devices. ⚡ Fast decode for geometry streaming. Download the SDK ⬇️ gpuopen.com/learn/introd...

gpuopen.comIntroducing AMD DGF SuperCompression - AMD GPUOpenAMD DGF SuperCompression (DGFS) cuts DGF geometry file sizes while preserving exact block reconstruction and enabling fast decode to either DGF blocks or conventional meshlets for cross-device deploym...
AMD GPUOpen
@gpuopen.bsky.social

Introducing MiniDXNN: a new MLP library built for #DirectX12 HLSL, enabling neural rendering without compute API interop: 🧠 Optimized MLP inference kernels. ⚡ Acceleration via cooperative vector APIs. 🎮 Sample apps & full source code. Get started ⬇️ gpuopen.com/learn/minidx...

gpuopen.comIntroducing MiniDXNN: MLP library for DirectX 12 - AMD GPUOpenMiniDXNN is a native HLSL and DirectX 12 library for lightning-fast MLP inference leveraging AMD Radeon™ RX 9000 series matrix cores via cooperative vector APIs, delivering optimized kernels, samples, full source and docs to remove compute interop friction.
AMD GPUOpen
@gpuopen.bsky.social

Learn how GOALS ⚽ delivers competitive performance on AMD Ryzen handheld PCs. Part 1 & 2 cover: ⚡ RDNA GPU power management. 🔧 ADLX SDK integration. 🎮 FSR for dynamic scaling. 📊 Real-world optimization strategies. All the #gamedev details ⬇️ gpuopen.com/learn/how-go...

gpuopen.comHow GOALS delivers sustained, competitive esports performance on handheld PCs - part 1 - AMD GPUOpenThe first part of a developer-first look at how GOALS leverages AMD Ryzen APUs and the ADLX SDK to implement a system that reduces power, fan noise and carbon footprint across legacy and handheld hard...
AMD GPUOpen
@gpuopen.bsky.social

The AMD FSR “Redstone” SDK 2.2 is now available and includes: 🔥 FSR Upscaling 4.1 with improved visual quality. 💥 FSR Ray Regeneration 1.1 with quality and memory improvements. Head over to GPUOpen to download 📥 gpuopen.com/learn/amd-fs...

gpuopen.comWelcome to the AMD FSR SDK 2.2, now available on GPUOpen - AMD GPUOpenThe AMD FSR™ "Redstone" SDK 2.2 update delivers ML-powered FSR Upscaling 4.1 and FSR Ray Regeneration 1.1 optimized for AMD RDNA™ 4 graphics, enabling higher visual fidelity and performance with analy...
AMD GPUOpen
@gpuopen.bsky.social

Introducing AMD Smoldr, our open-source CLI to run #DirectX12 HLSL from simple text scripts. 🛠️ Compile shaders, create resources & pipelines ⚙️ Dispatch compute & ray tracing 🐞 Prototype/debug without C++ Faster iteration for DX12 devs 👇 gpuopen.com/learn/enhanc...

gpuopen.comEnhancing DirectX Testing with AMD Smoldr - AMD GPUOpenSmoldr is an open-source command-line tool that runs DirectX 12 HLSL shaders from simple text scripts, letting you compile, create resources and pipelines, and dispatch compute or raytracing workloads...
AMD GPUOpen
@gpuopen.bsky.social

🎮 AMD x Microsoft at #GDC2026 Partnering to bring powerful new developer technologies to Windows: ✨ DirectX ML integration 📦 DirectStorage 1.4 🛠️ PIX tools updates ⚡ Advanced Shader Delivery 🔧 Agility SDK update support All the #gamedev details ⬇️ gpuopen.com/learn/amd-mi...

gpuopen.comAMD and Microsoft partner on DirectX ML, DirectStorage, and developer tools at GDC 2026 - AMD GPUOpenMicrosoft and AMD partnered at GDC to announce powerful new developer technologies for Windows, including DirectStorage 1.4, PIX tools updates, DirectX ML integration, Advanced Shader Delivery, and su...
AMD GPUOpen
@gpuopen.bsky.social

Build 60 FPS cockpit HMI with Unreal Engine 5 + AMD Ryzen AI Embedded P100. A single #UE5 instance: 🚘 Unifies instrument clusters, maps & apps. 🚘 Delivers gaming-class graphics. With a streamlined dev workflow for in-car experiences. ➡️ More info: gpuopen.com/learn/drivin...

gpuopen.comDriving the future: frictionless automotive HMI development with Epic Games & AMD Ryzen AI Embedded P100 series - AMD GPUOpenAt CES 2026 Epic Games and AMD unveiled the Unreal Engine 5 Next‑Gen HMI Experience powered by AMD Ryzen AI Embedded P100 Series, unifying instrument clusters, maps, and apps into a single high‑perfor...
AMD GPUOpen
@gpuopen.bsky.social

HIP Threads lets C++ teams offload CPU hotspots to AMD GPUs by making minor changes to your existing multithreaded C++ code: ⚙️ No kernel rewrites, no GPU expertise required. 🏎️ Just map threads and see real speedups in days. 👉 Learn more: gpuopen.com/learn/hip-th...

gpuopen.comHIP Threads: GPU power for teams without GPU experts - AMD GPUOpenHIP Threads lets C++ teams eliminate CPU hotspots by running familiar multithreaded code on AMD GPUs, no kernel rewrites or GPU expertise required, delivering real speedups in days.
AMD GPUOpen
@gpuopen.bsky.social

See how Youthcat Studio rewrote the threading for Dyson Sphere Program: ✅ Custom core binding + dynamic task allocation + improved synchronization 🟰 more performance. ➡️ Discover how they are continuing to optimize for AMD Ryzen #Threadripper processors: gpuopen.com/learn/dyson-...

gpuopen.comDyson Sphere Program - new multithreading dev log & full AMD Ryzen Threadripper PRO breakdown - AMD GPUOpenYouthcat Games overhauled the Dyson Sphere Program game's multithreading system, boosting performance by up to 88% through custom core binding, dynamic task allocation, and enhanced thread synchroniza...
AMD GPUOpen
@gpuopen.bsky.social

This holiday season, our radiant Team Redstone engineers have upscaled the AMD FSR SDK by generating some new additions 🟥 Joining AMD FSR Upscaling, meet AMD FSR... 🪟 Frame Generation 🎇 Ray Regeneration ☀️ Radiance Caching #gamedev unwrap here: gpuopen.com/learn/amd-fs...

gpuopen.comAMD FSR™ Redstone for developers and the neural rendering future - AMD GPUOpenThe AMD FSR Redstone SDK enables developers to integrate neural rendering technologies, including ML-powered upscaling, frame generation, denoising, and radiance caching, for next-generation gaming ex...
AMD GPUOpen
@gpuopen.bsky.social

AMD Schola v2 drops today! 🔥 Our #opensource #UnrealEngine 5 plugin makes it easy to create AI that learns by playing your game — just like a human would. Perfect ✅ for: 🚀 Smarter NPCs 🎯 Realistic opponents 🌍 Dynamic game worlds gpuopen.com/amd-schola/?... #GameDev #GameAI

gpuopen.comAMD Schola - AMD GPUOpenAMD Schola is a library for developing reinforcement learning (RL) agents in Unreal Engine and training with your favorite python-based RL Frameworks.
AMD GPUOpen
@gpuopen.bsky.social

🦾 Watch a robot arm teach itself to work! We trained a virtual X-ARM 5 in #UnrealEngine using AMD Schola — starting simple (reach one spot) and progressing 🚀 to complex (find any colored target). 🎯 It learns by trial and error, just like we do! gpuopen.com/learn/traini... #robotics

gpuopen.comTraining an X-ARM 5 robotic arm with AMD Schola and Unreal Engine - AMD GPUOpenTrain a robot arm with reinforcement learning in AMD Schola using Unreal® Engine, progressively increasing task complexity to adapt to changing conditions.