#GPU
200
2026-10-01
deepseek-ai/DeepGEMM-Ascend
S
DeepGEMM-Ascend: clean and efficient matrix multiplication kernel library for Huawei Ascend NPUs
2026-09-11
deepseek-ai/DeepSelect
S
DeepSelect: TopK kernels for DeepSeek Sparse Attention (DSA) and Samplers
2026-09-04
sohu-mptc/FlashRec
A
FlashRec is a CUDA-graph engine for generative recommendation: wide beam search (3–5 SID steps, n=50–512+) over a trie-constrained catalog, in-process FP8 serving, and ranked beams on /v1/chat/completions.
2026-10-16
ahwootech/BRUTAL
A
A free, open source C# framework for making games, or your own game engine. Direct access to native APIs and Vulkan, with the low-level libraries around them. The foundation under Kitten Space Agency and RimWorld's Waylith engine.
2026-10-01
deepseek-ai/DeepEP-Ascend
A
A high-performance communication library for machine learning training and inference on Huawei Ascend NPUs.
2026-10-16
OpenMOSS/MOSS-Image
A
Official repository for MOSS-Image, a 5B-parameter base DiT model with high-quality text rendering, layout generation, and efficient fine-tuning.
2026-09-23
Show HN: InstinctFlash – High-Performance Serving Runtime for Robotics Models
A
Hey HN, Guanming here, cofounder of General Instinct. We just released InstinctFlash, a high-performance serving framework for robotics models. It’s licensed under AGPL-3.0. On Jetson Thor, we see speedups about 1.2x to 7.9x from runtime optimizations alone and up to 33.78x for LingBot-VA when we combine those runtime optimizations with a distilled few-step diffusion scheduler, going from the original 25 visual / 50 action steps to 2 / 4 steps. Across 50 Robotwin2.0 tasks, we evaluated 1,153 episodes per configuration, LingBot-VA with InstinctFlash at 2 visual / 4 action steps achieved a 90.5% success rate, compared with 92.1% for the baseline at 25 visual / 50 action steps. Here’s an optimized 5B world action model, running in real time on a Jetson Thor: https://youtu.be/nku65iyL5Fw InstinctFlash currently supports 8 VLA / world-action model families, including pi0.5 and NVIDIA Cosmos Policy, across RTX 4090 / 5090 and Jetson Thor. Just give it your fine-tuned checkpoint and InstinctFlash handles the rest, exposing the accelerated model through a Python runtime or an OpenPI-compatible WebSocket server. We started working on this because we kept running into the same problem while deploying robot policies, the models were getting much better, but inference was often way too slow for the control loop we actually wanted. For pi0.5, mixed-precision GEMMs and CUDA graphs speed up computation and reduce launch overhead. For Cosmos, caching avoids redundant computation across diffusion steps. World-action models’ diffusion denoising step depends on the previous one which motivated our work on few-step distillation. Right now, InstinctFlash contains 6 aspects of optimization. - Graph: CUDA graph capture, memory planning and separating prefill from repeated execution. - Cache: Reusing KV and conditioning state across diffusion steps and prediction calls. - Attention: Specialized attention paths for different model architectures. - Kernels: Fused operations and kernels tailored to specific backends and tensor layouts. - Precision: FP8 and mixed-precision execution. - Model: Few-step distillation for diffusion and action generation. Teams at Samsung, Siemens, and other robotics startups have used InstinctFlash for model acceleration on VLAs, WAMs, and diffusion-based world models. Now we are opening up access to you. Feel free to try it here: https://github.com/General-Instinct/InstinctFlash More implementation details and benchmarks: https://general-instinct.com/blog/instinctflash-edge-inferen... Would love to hear your feedback!
2026-11-04
shanselman/gputaskman
S
A GPU task manager for Windows, Linux, and WSL: per-process VRAM, GPU activity, and LLM model/KV-cache insights (Windows AMD and NVIDIA).
2026-11-09
SamG-Coder/solstice
S
Nanite and Lumen style rendering in the browser: virtualized geometry and dynamic global illumination in Rust and WebGPU
2026-08-27
0xBakeer/qwen38-flash-next-spark
A
Run Qwen3.8-Flash-Next (180B) on a single DGX Spark by keeping its 51B n-gram embedding table on NVMe
2026-09-22
maanHimself/OpenDLSS-NR
A
A Vulkan reimplementation of NVIDIA's DLSS 5 Neural Rendering network, bit-exact against the original.
2026-10-25
alleyxd/quantpicker
A
Measures which GGUF quant of a model to run on your GPU: real speed and quality loss (KL divergence), not estimates.
2026-09-18
RenaudRohlinger/node-webgl
A
Headless WebGL 1 & 2 for Node.js, powered by ANGLE (Metal, D3D11, Vulkan). Render any WebGL or three.js scene without a browser.
2026-11-04
fuyichen0725/blinkmimic
A
Whole-body motion tracking for the Unitree G1 in about 3 minutes on one GPU.
2026-10-22
ahtx/kindi
A
A custom LLM inference engine for the NVIDIA Quadro K420: Ternary-Bonsai-2-2B at 20+ tokens/s with an OpenAI-compatible server
2026-09-22
Show HN: Mini-AGI – Dynamic continual learning model trained on 8GB VRAM
A
Sorry for the pretentious name, I know, I know.. It just contains all the pieces I would like to see a AGI model to have, and I can't stand the temptation. Before throwing rocks at me, please take a glance at the Readme, and I hope it will cover your mood a little bit. So, first of all it does work and you can see the sample from the whole training run here: https://raw.githubusercontent.com/volotat/mini-AGI/refs/head... Here is the scaling law graph I have so far, and it looks very promising: https://github.com/volotat/mini-AGI/blob/main/assets/scaling... The model was built under my deep dissatisfaction so we cannot really train even moderately big models (1B+ scale) on the consumer's hardware. We can inference and fine-tune them for sure, but I would like to have full control over what the model sees over the training run, so it is fully aligned with my interests, not some corporations. I was thinking about for some time and come up with two interesting ideas I thought worth pursuing: MoE with a lot of experts that gets added and pruned from the model while it trains, where only a small subset of of experts are actually in use at any particular moment + batch 1 training on the single continuous stream of data. First allows us to be bounded only by the disk space in terms of number of parameters and load and unload experts only when they are needed. The second (if figured out and it turns out to be doable) allows us to get aways with small VRAM capacity because we do not need to store big randomized batches and their respective gradients. I started brainstorming with Claude and after some time we found an approach that seems to be promising, and low and behold, a few weeks pass and you can see the results yourself. Obviously, I did use AI in the process of making this project and I am pretty sure it would be completely impossible for me to do something like this without it, so I hope it is more than justified. The model is still running over the first of 7.8B characters corpus I selected for training, so the weights are not out yet, and it's about a couple weeks of waiting until they are cooked at the current reading speed. And yeah, the model just read continuous interleaved passages from the dataset, each by 32K characters long each as a single stream. Just as you or I would do. The set up seems to be really simple so you can git clone the project, run it and observe everything for yourself. Thanks for your attention.
2026-10-17
gpu-verifier/sentryward
A
Automatic formal verification for warp-specialized GPU kernels.
2026-10-04
Almosst-DEV/Navi48-MacOS
A
Proof-of-concept native macOS support for the AMD RX 9070 XT (Navi 48): bring-up kext, RADV Darwin port, Metal-on-Vulkan bundle
2026-09-11
kevinzakka/mjbatch
B
A Python library for running thousands of MuJoCo simulations in parallel on CPU
2026-11-09
LiteLDev/Arke
A
Runs Java Edition shader packs written for Iris and OptiFine on Bedrock's Direct3D 12 renderer.
2026-10-28
K240/hdUE
A
A same-process Unreal Engine Hydra render delegate with Lumen, Path Tracer, and Houdini Solaris integration
2026-10-23
Acelogic/Snapintosh
B
EL2 hypervisor and toolchain that boots macOS on the ASUS A16 (Snapdragon X2 Elite), with an Adreno Metal/IOGPU stack
2026-10-08
nullmoth/nvidia-macos-driver
B
Metal driver for NVIDIA GeForce RTX cards on macOS 15 Sequoia (Intel / OpenCore). Free, source included.
2026-09-11
fromgof/TridentVision
A
Implement yolov26m using a complete FPGA, without any C language code. Requires 160K LUTs and 386 DSP cores (based on Xilinx XC7K325T).
2026-09-03
Blueforcer/ComfyUI-DLSS5-Enhancer
A
NVIDIA DLSS 5 Neural Rendering (NGX feature 18) as ComfyUI nodes. Enhance image batches and video files with the real neural renderer, with optional upscaling.
2026-08-28
tonyd2wild/GLM-5.3-Flash-NVFP4-DFlash2-2x-DGX-Spark
B
GLM-5.3-Flash (NVFP4) on 2x NVIDIA DGX Spark - vLLM TP2, 262K context, MTP. World-first deploy recipe: 7 day-0 bugs found and fixed, patched sm121 image, probes and full report.
2026-10-30
mstlvc/Radiance
A
Lightweight, bloat-free Windows display utility for NVIDIA digital vibrance, DDC/CI hardware monitor control, Direct3D 11 color vision overlays, and custom crosshairs.
2026-10-13
tritant/ComfyUI_NVIDIA_RTX_VSR_Long_Video
A
NVIDIA RTX Video Super Resolution for ComfyUI, optimized for long videos with disk-backed processing and low RAM usage. Supports AI upscaling, denoising, deblurring, H.264/HEVC NVENC encoding, audio preservation, and an integrated video preview.
2026-09-28
linuxkid473/ppcosxkvm
A
PowerPC Mac OS X Tiger on Apple Silicon with an emulated ATI Radeon 9700 PRO: Quartz Extreme, Core Image and OpenGL via Metal
2026-09-16
slorber/gpu-lexer-textarea
A
Demo of using gpu-lexer with the new OpaqueRange web API (Chrome 152+) to create a universal syntax highlighter within an editable textarea
2026-08-30
jlrouzies-fr/DLSS5-Feeder
B
DLSS 5 neural rendering in D3D11/D12/Vulkan games that ship without any DLSS — feeds a synthetic DLAA contract (ReShade depth + motion vectors) to the DLSS 5 add-on via a private D3D12 device.
2026-10-05
MohammadHaishemKhawaja/hashyy
A
A full, unpruned 177B MoE model at 11.5 tok/s on one 12 GB GPU. Expert streaming for llama.cpp, with the measurement harness and every dead end.
2026-10-05
noahdunnagan/mcopt
B
mcopt, a Minecraft mod: a native Metal renderer for Minecraft: Java Edition on Apple Silicon Macs (alpha)
2026-09-24
ColinLeung-NiloCat/NiloPathTracing
A
RTX Path Tracing + DLSS Ray Reconstruction for Unity URP. One-click setup, native NiloToon support.
2026-09-23
NoizAI/YuE2-Turbo
A
Fast, concurrent inference for YuE2. Same model and recipe: 1.68× faster per song, 3.31× more songs per GPU.
2026-09-08
Deep-Learning-Profiling-Tools/ptx-optimization
B
Evidence-led NVIDIA PTX optimization skill: profiling, exact-cubin patching, validation, and measured tuning
2026-10-28
xiaoxiao150/vram-switcher
A
GPU coordination for local AI companions on AstrBot / DeepSeek Harness: persona LLM, IndexTTS voice and ComfyUI selfies take turns on one 12 GB card. No proxy.
2026-10-26
alrojo/fp4Attn
A
CroissAttn: FP4 attention forward and backward, with analysis of stochastic rounding, zero-sum softmax gradients, and sampled centering.
2026-10-03
juliankang4/spark-scope
A
Read-only dashboard and rack panel for DGX Spark / ASUS GX10 / MSI EdgeXpert nodes running vLLM or SGLang
2026-09-06
aufkrawall/dlssdl
A
Check NVIDIA NGX update servers for DLSS SR / RR / FG and Streamline updates and download them with proper consumer file names (egui, D3D12 renderer)
2026-10-25
lillyguisnet/sam3-segmentation
A
Agent skill: outline, mask and count objects in photos with Meta's SAM 3 on your own NVIDIA GPU, with review, exports and a viewer. Works with Claude Code, Codex, Cursor, Pi and other Agent Skills clients.
2026-10-16
PaddlePaddle/FlashMask
A
Efficient Mask-Aware Distributed Attention for Long-Context Training
2026-10-15
MiaAI-Lab/sparkLimiter
A
Cap the GPU clock on NVIDIA DGX Sparks across all your machines, persistently
2026-10-13
GaelQuinta/minecraft-metal-shaders
A
Native Metal shaders for Minecraft Java on Apple Silicon Macs: shadows, volumetric clouds, god rays, reflective water, HDR. A fork of mcopt.
2026-09-16
ismailivanov/TressFX-Godot
A
Real-time hair and fur for Godot 4: a native GDExtension port of AMD TressFX
2026-09-01
faisalkindi/DLSS5oneclick
B
One-click setup of the leaked DLSS 5 neural-rendering build for any DX11/DX12 game on RTX 20–50, with or without DLSS. ReShade + RenoDX add-on (or OptiScaler engine); DLSS5-Feeder + LumeniteFX for games without DLSS; dlss5-bridge for DX11. Rust, single exe.
2026-10-30
Huy-Quang-Dao/instantTrack
A
Whole-body motion tracking for the Unitree G1 in 15 minutes on one GPU (SuperMimic-style world model + InstantMimic-style GPU loop, mjlab)
2026-10-23
alexturkey/voxcpm2-turkce-ses-tasarimi
A
Turkish voice design with VoxCPM2 on a 4 GB GPU: describe a voice, pick a variant, read Turkish text. Patched audio.cpp, local UI, 7 synthetic voices.
2026-10-15
hnlyf168/SimdYolo
A
Pure C#/.NET YOLO inference: SIMD CPU, Vulkan/Metal GPU, direct .pt and ONNX; detection, segmentation and pose.
2026-10-13
IFreemz/msfs2024-glass-cockpit-dlss-fix
A
ReShade add-on that removes DLSS and FSR ghosting from glass cockpit displays (PFD, ND, ECAM) in Microsoft Flight Simulator 2024, on a monitor and in VR
2026-10-04
Yamz-Labs/kyojin
A
Kyojin: the Yamz inference engine for AMD Strix Halo (ROCm, gfx1151), built on ExLlamaV3. Runs 300B-class MoE models on one 128 GB mini PC.
2026-10-04
christopherowen/dgx-spark-memory-saver
A
Recover usable memory on DGX Spark with a 64 KiB kernel: an opt-in NVIDIA UVM page-table packing patch, build tools, and measured validation.
2026-10-02
niklasp/gpuslider
A
A slider drawn by shaders: WebGPU, or WebGL 2 where there is none. A 4.7 KB core, everything else a plugin.
2026-09-26
bertaye/agentic-cuda-optimizer
B
A langgraph based workflow with a C++ CUDA harness to optimize CUDA kernels
2026-09-26
Show HN: Agentic CUDA Kernel Optimizer
B
Hello; I was working on optimizing some CUDA kernels and I thought may be it is a good oppurtunity learn langgraph as well. I created a simple C++ CUDA Test Harness and handed that to AI agents. They can run kernels, get benchmarks, and even can profile via nsight
2026-09-07
Datura-ai/lium-localmaxxing
A
Reproduce Lium's LocalMaxxing leaderboard runs (7 boards, 20 Verified) on a rented Lium GPU in one script — vLLM/llama.cpp recipes, measured cost per million tokens
2026-08-28
MiaAI-Lab/GLM-5.3-Flash-NVFP4-Dual-DGX-Spark
B
2x DGX Spark recipe for LibertAIDAI/GLM-5.3-Flash-NVFP4 (NVFP4, Ray TP=2, multimodal)
2026-08-27
nuskey8/SpriteAfterimage
A
High-performance afterimage effect for Unity 2D using GPU instancing
2026-08-23
bbaudio-2025/Comfyui-MMH3-UltimateUpscale
B
Upscale long, high-resolution MiniMax H3 video on a VRAM-limited GPU in a single node.
2026-10-17
Noniv/web-ocean
B
Real-time WebGPU ocean renderer written from scratch, no engine. FFT waves, whitecaps, sun glitter, physically based sunset sky with clouds, free camera. Built with Claude Code (Opus 5.5). Opens from a single 61 KB HTML file in Chrome.
2026-09-26
architectds/collabosm
B
Run Qwen3.8-Flash-Next (125B-A6B MoE) on ONE Colab A100-80GB High-RAM: pinned runtime, weights from HF, OpenAI-compatible endpoint, measured 2.8k-3.9k t/s prefill / 90-97 t/s decode, plus a verified pinned-RAM KV tier.
2026-09-10
glarsson/fast-rtxvsr
B
Standalone NVIDIA RTX Video Super Resolution CLI - GPU decode -> VSR -> encode in one pass, no ComfyUI
2026-09-01
v2space-labs/shader-for-interfaces
B
Agent Skill for designing, building, debugging, and validating focused GPU effects in product interfaces.
2026-08-29
StabRise/ScaleDP-TS
B
ScaleDP-TS processes PDFs and images entirely in the browser: PDF rendering, text detection, OCR and named-entity recognition, composed as a pipeline. Inference runs on onnxruntime-web over WebAssembly or WebGPU.
2026-08-27
seanyourhighness/vllm-sm12x-nvfp4-dflash2
B
All-NVFP4 vLLM v0.27.1 + DFlash2 K7 speculative decoding for Blackwell (SM120/SM121): NVFP4 target/draft/KV, 262K context, optimized language-only serving.
2026-10-24
fan-ziqi/rstop
B
A GPU-aware system monitor for the terminal, built in Rust. CPU, processes, storage, networking, containers and services. Linux, macOS and Windows.
2026-10-03
kindlingai/kindling-spark-os
B
Minimal read-only OS image for GB10 (DGX Spark / ASUS GX10) boxes
2026-09-25
ruc-system/quiver-ae
B
A low-latency, high-throughput, and billion-scale GPU-SSD vector search system.
2026-09-11
zmodelerlover/dlss5-neural-amd
B
ReShade add-on running the DLSS-NR network on AMD GPUs via HIP, RenoDX-shaped. Tested on PCSX2. Published to be forked. Discord: https://discord.gg/wYhvS3JSHM
2026-11-02
KohakuBlueleaf/Tokenomos
B
Local inference for small VLMs: host work per step stays flat, so the GPU is the only bottleneck
2026-10-22
guscanales/rigup
B
Turn a gaming PC or a MacBook/Mac Mini into a local AI server with one install. A web dashboard installs and switches the models: chat and coding on both, image and video on NVIDIA. Tuned for 12 to 24 GB NVIDIA cards and 16 to 128 GB Apple Silicon Macs.
2026-10-20
lxe/dsx
B
DeepSeek-V4.1-Flash-1017 (552B MoE) on 2x RTX 3090 + 128 GB RAM: Rust + CUDA engine and 2-bit quant. 30-37 tok/s decode.
2026-10-14
obraia/gausify-orbit
B
Turn orbit videos (including AI-generated ones) into Gaussian Splats, 100% locally: desktop app (Tauri) and a WebGPU version that runs in the browser.
2026-09-02
mohamedhossammohamed/m4-prefill-engine
B
Hardware-saturated Metal inference engine targeting the LLM prefill bottleneck on Apple M4 (~3.7x True GPU Speedup).
2026-11-02
Betelgeuse-AI/Betelgeuse
B
A custom fork for frontier-scale mixture-of-experts inference on an RTX 3060 with 12 GB VRAM, 16 GB system RAM and local NVMe.** Betelgeuse keeps dense computation on the GPU and brings routed experts into a bounded VRAM pool only when needed.
2026-10-29
jameshaworthcs/diagonal-blend
B
Two photos, one seamless diagonal: Rust and wgpu in the browser, WebGPU with a WebGL2 fallback. Nothing is uploaded.
2026-10-26
Leonxlnx/procedural-fire
B
A fully procedural, physically based campfire in WebGPU: 3D combustion sim, blackbody rendering and synthesized sound. No assets.
2026-10-23
JMS1717/ming-voice
B
Low-latency local speech-to-speech voice assistant with Ming-3.0-audio-tiny on AMD Radeon GPUs (ROCm on WSL2)
2026-10-22
willwaltz/globes
B
Glass globes of water, simulated live on the GPU with WebGPU and three.js
2026-10-14
AndrewTriesToCode/bg3mfx
B
MetalFX upscaling and frame generation for Baldur's Gate 3 on Mac
2026-09-27
ScottieFox/caustic-volume
B
Real-time water caustics in the browser, built on three.js. Single-file lite and sandbox versions.
2026-08-29
Holozing1/vram-guard
B
Detect and work around silently-corrupting VRAM on consumer GPUs. No ECC, no page retirement, no error - the driver just returns wrong data.
2026-08-24
maderix/apple-gpu-dvfs
B
Apple Silicon GPU DVFS reverse engineering: voltage corners, CLPC frequency control, Metal benchmarks
2026-11-09
LuckyIYI/fah-metal
B
FAH hand pose on Apple silicon in plain Metal: 115-234 fps on M5, with MANO mesh and hand-object contact
2026-10-28
johnfink8/sfnx
B
CUDA engine that splits every layer of a production quality MoE LLM across three PCIe V100 cards delivering 382 tok/s on a used $1,800 box.
2026-10-26
ArjixWasTaken/penguin-engine
B
Animated desktop wallpapers for Linux (Wayland): plays Wallpaper Engine scene, video and web wallpapers. Rust + wgpu.
2026-10-12
UndeadBulwark/Proton-ZLUDA
B
Run CUDA / GPU-PhysX Windows games on AMD Radeon under Steam on Linux, via ZLUDA on ROCm. Drag-and-drop Proton compatibility tool + version converter.
2026-10-04
bhouston/three-dlss-nr
B
A Three.js (TSL / WebGPU) port of OpenDLSS-NR, the open-source reimplementation of NVIDIA's DLSS 5 neural rendering network
2026-09-05
MiaAI-Lab/Qwen3.8-Flash-Next-Single-DGX-Spark
B
Qwen3.8-Flash-Next on ONE DGX Spark (TP=1)
2026-10-29
patch-gg/crowdbricks
B
Batched NVIDIA MotionBricks: generate the motion of hundreds of characters in one model call.
2026-10-28
physicsdolphin/nv-uefi-unlocker
B
UEFI pre-driver NVIDIA GPU unlocker for Windows (Rust): SEC2 scrubber -> FWSEC/WPR2 -> signed Booter -> V67 four-write payload, read-back verified. Verified on RTX 4060 Laptop 28e0.
2026-10-02
0xSero/glm53-flash-offload
B
GLM-5.3-Flash EXL3 on one 24 GB RTX 3090 + DDR4: elastic GPU expert cache, zero-copy experts, AVX2 CPU tier, OpenAI API
2026-09-28
yunfanye/openhouse-3d
B
Turn real-estate listing photos into a 3D house model and a cinematic walkthrough video. Photo camera solving, procedural Blender (bpy) modeling, virtual staging, headless Cycles rendering and photo-vs-render comparison films.
2026-09-16
LevinAi-arch/rtx-5070ti-laptop-160w-power-limit
B
Research and proof-of-concept for raising the NVIDIA RTX 5070 Ti Laptop GPU power limit from 140W to 160W on Lenovo Legion Pro 5 Gen 10.
2026-09-16
serialexperimentslainnnn/afmf-linux
B
Vulkan layer to enable AFMF (AMD Fluid Motion Frames) For linux
2026-09-11
tonyd2wild/DeepSeek-V4.1-Flash-vLLM-DGX-Spark
C
DeepSeek-V4.1-Flash (552B MoE, MXFP4 experts, 1M ctx) on four NVIDIA DGX Sparks with vLLM TP4: Engram-on-disk patch, sm121 kernel build, launchers, measured numbers
2026-09-04
shtjww/llm-inference-capacity-handbook
B
A desk reference for LLM inference (open-source edition): given GPU compute, how many QPS can a model handle? Three-layer model × three walls × queueing theory × open-loop load testing
2026-09-02
hbofz/NeSLE
B
GPU-native NES emulator + RL stack: thousands of parallel NES envs on one CUDA GPU, training agents on Super Mario Bros.
2026-09-02
ntqueryinformation/FeedKit
B
One-click installer/uninstaller for DLSS5-Feeder: brings DLSS 5 neural rendering to games without DLSS. Fetches everything fresh from upstream at install time.
2026-08-29
maisterr/phone-llm-fp8
B
FP8 quantization + vLLM serving of pipecat-ai/phonellm-alpha-1 on Modal
2026-08-24
Redhatvale/Nvidia-Tesla-v100-nvfp4-pcie
B
Qwen3.8-27B in native NVFP4/FP8 on 2x PCIe Tesla V100-32GB (SM70): the PCIe runbook for v100-skinny + 1Cat-vLLM, with the 3 fixes that make it work without NVLink. 61-74 tok/s decode, MTP speculative decoding, OpenAI-compatible.
2026-11-02
alla2001/material-painter
B
PBR texture painter: paint, bake mesh maps and design materials with a node graph. Rust + wgpu.
2026-10-24
joshuaswarren/omarchy-egpu
B
eGPU setup, driver status, diagnostics, and measured results for Omarchy Linux
2026-10-16
VantaCode420/SmartHdrManager
B
Automatic Windows HDR switching for games, YouTube HDR and RTX Video HDR
2026-10-16
alienflm/MacTok
B
[CVPR 2026 Highlight] Official PyTorch implementation of MacTok, a masked 1D continuous image tokenizer that compresses images into just 64 or 128 tokens.
2026-10-15
OpenImagingLab/FlashVSR_Plus
B
This is the official repo of FlashVSR+: Efficient Video Super-Resolution via Hybrid Attention and Reward-Yield Distribution Matching.
2026-09-14
mdj2812/sunshine-vdisplay
B
NVIDIA + KDE Wayland virtual display for Sunshine/Moonlight streaming on Linux
2026-10-17
plotarmordev/drama-tensorfold
B
Fish Audio Drama 3 text-to-speech on the TensorFold native CUDA engine for DGX Spark (GB10): recipe, engine patches, benchmarks. Built with Fish Audio.
2026-10-06
lonewolf0622/HelixSR
C
Drop-in FSR 3.1 upscaler for Direct3D 12 that runs DLSS's network (Model E) on AMD GPUs
2026-10-03
AdrielSantana/three-bodies
B
The three-body problem, live, in Bend: physics on the CPU, every pixel on the GPU.
2026-09-25
yoheinakajima/glance-speedlab
B
A measurement-first lab for low-latency local camera inference with Glance.
2026-09-19
Peeter95/code-map
B
Fly over a codebase as a zoomable GPU treemap (2D and 3D city), built with Makepad
2026-09-17
dollarwong/ninfer-v100
B
NInfer on Tesla V100 (sm_70): Qwen3.8-27B deployment, build/convert scripts, systemd service, and cross-engine benchmark (vs llama.cpp). Fork of Neroued/ninfer with sm70 vision-encoding speedup.
2026-09-17
gensyn-ai/open-transformers
B
Reproducible LLM pretraining and audit-replay harness behind the OPEN model collection
2026-09-15
XiaoQAQ10086/FrameGen-Manager
B
One-click DLSS Frame Generation deployer for RTX 20 and 30 series GPUs, with game library scanning and auto-update. One-click DLSS Frame Generation deployer for RTX 20/30 series.
2026-09-08
nerficg-project/efficient-gaussian-appearance
B
Faster-GS with JIT-compiled view-dependent appearance models and a WebGL viewer, described in the paper "Compact Neural Appearance Models for Efficient Gaussian Splatting"
2026-08-29
NIGos/dlss5-dx11-bridge
B
Lets D3D12-only DLSS 5 Neural Rendering add-ons run in DirectX 11 games
2026-08-28
tonyd2wild/GLM-5.3-Flash-NVFP4-1M-KV-4x-DGX-Spark
B
GLM-5.3-Flash (320B MoE) at TP4 across four DGX Sparks, same day as the model drop: 36 tok/s, 1.26M-token FP8 KV pool, 262K context, MTP spec decode. First TP4 glm5_next outside B200 hardware. Full patched-image recipe + the GB10 memory study.
2026-08-25
wonderfulearth/primitive-operation-painter
C
Autoregressive primitive-operation painter with GPU image-to-sequence converter
2026-09-26
mochizuki0323/DLSSNR-AMD
B
DLSS 5 neural rendering (NR) model on AMD RX 9000 (RDNA4) through Vulkan
2026-09-04
ImDreamt/MFGAdaUnlock-RenoDx
B
DLSS MFG on Ada — unlocks 3x/4x/6x frame generation on RTX 40-series cards and corrects the temporal midpoint. In-memory only.
2026-09-02
lihaoyun6/ComfyUI-H3VAE_TRT
C
Running ONNX/TRT version of the MiniMax-H3 VAE in ComfyUI, which increase speed by up to 1.7x
2026-08-28
tonyd2wild/GLM-5.3-Flash-NVFP4-2x-DGX-Spark
B
GLM-5.3-Flash (NVFP4) on 2x NVIDIA DGX Spark - vLLM TP2, 262K context, MTP. World-first deploy recipe: 7 day-0 bugs found and fixed, patched sm121 image, probes and full report.
2026-10-16
Pixelgamer4k/graden
B
Multi-layer video editor and colour grading app for Android — proxy editing, native-resolution preview, colour wheels and scopes, full-quality H.264/H.265 export. Beta.
2026-10-08
mw00/project-maya
B
Project Maya - GLM-5.3-Flash (321B MoE) on your own NVIDIA GPU(s): engine, server, dashboard, quant tools. Built on Strata.
2026-10-07
buberlo/madeira-anyps5
B
Research prototype: run AnyPS5-relinked PS5 titles on iPad via Madeira (Wine ARM64EC + FEX) with MoltenVK
2026-09-25
virtualrepublic/Gaussian-Render-Capture
B
Blender add-on: renders a model into a ready-to-train COLMAP capture for Gaussian Splatting (Postshot, LichtFeld Studio)
2026-09-23
tonyd2wild/MiMo-V2.6-Flash-DGX-Spark-Recipe
B
MiMo-V2.6-Flash-RL on 2x NVIDIA DGX Spark (GB10): vLLM TP2 + DFlash + vision/audio, fp8 KV 300K, four fixes for the stock image, measured baselines
2026-09-12
pandaligx/RTX-FG-Manager
C
RTX Frame Generation Manager by 小南瓜 · Windows binary releases and documentation.
2026-09-10
Heretek-AI/chlorine-server
B
Clean-room AGPL-3.0 reimplementation of the halogen 0.1.3 inference engine for Qwen3.8-27B on AMD Strix Halo (gfx1151) — same wire protocol and OpenAI-compatible API. Not affiliated with Peonist.
2026-08-29
tl2012tl/TE-Speed-FlashVSR
B
A very fast FlashVSR video super-resolution upscaling node for ComfyUI.
2026-11-09
mingzi206-gh/colab-t4-local-api
B
Agent-oriented Colab T4 to local OpenAI-compatible API deployment kit, with Drive cache and offline checks.
2026-10-27
challenger0303/VRSS-Tuner
B
Experimental NVIDIA VRSS2 tuner and gaze provider for VRChat; Pimax Dream Air verified
2026-10-15
HyosungSink/CANN-AscendC-optimization-Cases
B
Hands-on source code collection for Ascend C operator-level optimization on CANN 9.0+ and Ascend A2/A3
2026-10-13
michaelshimeles/terminal-rs
C
Bloomberg-style trading terminal in Rust with GPU charts and congressional trades
2026-10-07
SamG-Coder/dynamic-gaussian
B
CUDA-generated FFT ocean, refracted seabed caustics and dynamic Gaussian clouds with Three.js WebGPU
2026-10-06
wieslawsoltes/VB6
B
VB6 Studio Web — a browser-native classic Visual Basic IDE, form designer, source runtime and controls library in HTML, CSS and JavaScript, with a debugger, WebGPU drawing and single-file HTML app export.
2026-10-05
christopherowen/dgx-spark-networking
B
Switchless RoCE collectives, NCCL profiles and fabric tooling for DGX Spark inference recipes
2026-09-29
AkbarSheikh-debug/wondersnap
B
Gesture-controlled 3D models made of 200k glowing GPU particles. MediaPipe hand tracking + hand-written WebGL2, runs in the browser.
2026-09-21
athralk/TinyGPU
B
Custom RVV 64 bit vector accelerator for the CVA6 RISC-V CPU. SIMD style layout with 3 memory hierarchies inside the accelerator.
2026-09-17
r0b0tlab/qwen38-exl3-dflash2
B
Qwen3.8-27B EXL3 (4.00 bpw) + DFlash2 speculative decoding for ExLlamaV3, validated at 262k context on a 24 GB RTX 3090
2026-10-26
KohakuBlueleaf/KohakUnlock
B
Experimental NVIDIA tensor-compute unlock research: guarded TU102 initramfs port, offline tests, and Ampere reproduction notes
2026-10-26
Pixelgamer4k/VidUltra
B
RAW video camera for Android: GPU-developed RAW to 10-bit HEVC, gyro stabilization, film-like noise reduction, iPhone-style auto look and 14 log profiles. Beta.
2026-10-10
Wallawalla47/Infernix
B
C++/CUDA single-GPU inference engine for Qwen3.8 (incl. Qwen3.8-Flash-Next with offloaded experts) on the RTX 5090; grew from NInfer
2026-10-05
tsaipifong/whirl-llm
B
WHIRL (Windows HIP Inference for RDNA LLMs): native Windows C++/HIP LLM inference engine for the AMD Radeon AI PRO R9700
2026-09-19
mmastrac/djev-spark
C
DiffusionGemma NVFP4 structured decisions on a DGX Spark: container recipe
2026-09-18
EnzymeAD/ReactantNitro.jl
B
Reactant-first training for Lux models: declare the experiment, compile once, train without boilerplate.
2026-09-11
iamtechartist/ocean-simulation
B
An interactive real-time ocean simulation combining FFT/JONSWAP waves, depth-aware coastal behavior, foam, caustics, refraction and dynamic weather around a tropical island.
2026-08-31
PixelML/club-170hx
B
Community-tested recipes, diagnostics, and benchmarks for CMP 170HX AI workloads
2026-11-09
WeebLabs/Nightscape
B
A live, calming, realistic night flight over a moonlit sea of clouds, rendered in real time with WebGL 2 as a semi-interactive wallpaper.
2026-11-03
sylvieistaken/selevet-eglantine
B
Selevet Eglantine: an unofficial Eglantine (Switch 2 emulator) build tuned for Zen 3, with PGO, ThinLTO and faster GPU mapping updates
2026-10-12
isichan-ai/Strata-GLM
B
Strata fork that runs GLM-5.3-Flash (320B-A18B, 157 GB) on one PC: RTX 5090 + 128 GB RAM + NVMe, ~30 tok/s chat. Fork of Niko1221/Strata (MIT).
2026-09-27
DaVarga/bs-arm64
B
Unofficial native ARM64 Beat Saber for the Steam Frame / ARM64 Proton: no FEX, no frame drops.
2026-09-21
sudoingX/bonsai2-small-gpu
B
run ternary bonsai 2 27b well on the gpus people own: serve lines per vram tier, a 1.5x decode kernel for the prismml fork, the qwen 3.8 mtp head grafted back, sweeps by pr
2026-09-14
Speedstu/CUDA-for-AMD-Windows
C
Run CUDA-targeted Windows applications on AMD GPUs with ZLUDA + ROCm/HIP.
2026-09-11
Quentincls/localmesh-engine
B
Photographs to a textured mesh, on one NVIDIA card with 8 GB, offline. Built on TRELLIS.2, with a four view path.
2026-08-27
tonyd2wild/Qwen3.8-Flash-Next-NVFP4-DGX-Spark
B
Day-0 deployment of Qwen3.8-Flash-Next (NVFP4) on 2x DGX Spark TP2 via SGLang. Includes the SM121 QSA kernel-guard fix, launcher, benchmarks, and full report.
2026-10-15
mihawk-99/PS5_MihawkGen
B
Neural frame generation for PS5 homebrew: beats FSR 3.1 on unseen game frames, in about 2 ms on the console.
2026-10-12
griches/PS5_Dusklight
B
Dusklight for PlayStation 5: a homebrew port on RADV (Aurora, Dawn WebGPU, SDL3). Bring your own disc image.
2026-10-06
iHaiDeeZ/DolphinPS4
B
Dolphin (GameCube/Wii emulator) for jailbroken PS4: Vulkan on the PS4 GPU, PSP-style XMB menu, per-game settings
2026-10-02
Tensward/tensward
B
Profile and diagnose LLM serving on your own GPU. Measures your vLLM setup on your real workload and suggests what to change.
2026-09-27
ai-jiaqian/cannbench-operator-solutions
C
Official CANNBench best operator sources with Job provenance and synchronization
2026-09-24
ledbetterljoshua/functional-emotions-video
B
A painted music video for "Functional Emotions", made by Claude Opus 5.5: custom GPU brushstroke renderer, storyboard, and 7 parallel chapter agents.
2026-09-22
CraneBW/ninfer-ternary-bonsai-ada
B
Ternary (2.125 bit/weight) Bonsai 2 27B on the NINFER engine — Ada sm_89, Linux. Derivative work of Neroued/ninfer; see README.