🎉 Another New Memory Overclocking World Record! 🎉 6576.8 MHz (13153.6 MT/sec) by Saltycroissant 🇨🇦 using Intel Arrow Lake, GIGABYTE Z890 Aorus Tachyon Ice, and @corsair.com Vengeance DDR5. Chart: skatterbencher.com/memory-overc... Details: hwbot.org/benchmarks/m... Validation: valid.x86.fr/j4et68

I'm gonna be attending the graphicsprogrammingconference.com in Breda, Netherlands this year. My first ever conference! Any chance I'll be able to meet anyone from here?

Ticket booked, looking forward to it!

Tickets are up! graphicsprogrammingconference.com/tickets/

Solved it by the ancient tradition of Just Reading The Code. Turns out continuously taking the RCU lock by inserting thousands of elements into a BPF_MAP_TYPE_LRU_HASH from within a NMI is Not Good for your system. Rolled our own (simpler) version directly in eBPF.

We're releasing our Neural Radiance Caching benchmark soon. Here's what you need to know 🧵 For clarity - all shading you see in the video is inferred through our high quality Neural Radiance Cache at the first intersection point (primary rays).

Wrapping C/C++ libraries in Rust doesn’t have to be painful. We’ve written up our approach to simple, reproducible build.rs setups using just cc + bindgen—no CMake, no autotools, no drama. If you care about FFI and clean builds, this is for you. 📖 Read more: www.evolvebenchmark.com/blog-posts/h...

Hey everyone! My talk on how we approach NPR rendering in @unrealengine.bsky.social at Telltale Games that I gave at the Graphics Programming Conferences is posted: www.youtube.com/watch?v=VoUJ...

We’re committed to helping developers, researchers, gamers and organizations accurately assess GPU performance. Here is an updated version of Evolve Benchmark Technical Guide—a more authoritative, transparent, and forward-looking resource. #GPU #game www.evolvebenchmark.com/technical-gu...

🦖 Evolve 1.0.3 Update LIVE! Check out our new Hall of Fame – flex your GPU scores! 🔧 Critical fixes: Temporarily disable benchmark interrupts for the Intel driver-configured frame limiter Authentication flow optimized...... 📝 Full notes: www.evolvebenchmark.com/release-note... #GPU

Nice to see our benchmark get some traction!

Breda features a highly customizable sun/atmosphere editor 🌇. This allows us to tweak many physics-based parameters 🔬 which influence the look of our atmosphere. In this video, we go over each of the settings to display their effect on the scene.

A while back i gave a talk on building and shipping a rust ui engine. It’s now free to watch here: www.infoq.com/presentation... #rustlang

🚪 The GDC Expo doors are opening! Find us at the AWS booth to: 👉 See Evolve in action 👉 Test-drive our cinematic Photo Mode 📸 Capture GPU stress test masterpieces! 📍 #GameDev #GDC2025

🕹️ GDC 2025 | The GPU Benchmark Revolution That Redefines Game Development Jasper, Chief Technology Officer at Traverse Research, will reveal how Evolve Benchmark will empower the next generation of benchmarking at #GDC2025: ⏳Book your appointment now：→ meetings-eu1.hubspot.com/jasper-bekke...

In our path tracers we keep track of a world space hash grid, which contains radiance values associated with ray directions and positions 🌞. Based on the path spread, we can stop tracing rays and instead directly use the cached radiance, still achieving an accurate result 🎯 at a lower cost.

🚨 Unlock a €23.99 Next-Gen Benchmark Tool for FREE! Want to stress-test your GPU with #EvolveBenchmark Advanced version? ✅ World’s first #Workgraphs Evaluation ✅ Real-time monitoring of GPU voltage/temp/energy ✅ A benchmark that crushes traditional tools Tag your hardware-obsessed friends! #gpu

The render graph debug composer is an essential component of the Breda framework. It allows us to view any resource we use during render graph construction 🏗️. In this video, we select a few textures which are used while rendering our final image.

Breda is our in-house rendering framework. This video highlights the efficient workflow we use for development. There are three steps when adding a new compute shader. (1) Setup resources with our render graph. (2) Define the shader in a yaml config file. (3) Write simple shaders using bindless.

This is some early footage of our work on neural radiance caching (NRC). With some aggressive learning rates 🎓 we achieve quite responsive results! Here we tweak the radiance coming from an object in the room next to us 💫. NRC is getting us closer to responsive and stable radiance caching 🌇.

@dachsjaeger.bsky.social thanks for the shout out of GPC, reallynice to see it get mentioned! 🫡

Here's a recording of my Tiny Glade presentation from the Graphics Programming Conference 2024: www.youtube.com/watch?v=jusW... It's an any% speedrun of our GPU-driven rendering, shadows, global illumination, water, and DoF, involving a few weird tricks.