Web-basierter Launcher für llama.cpp mit eGPU Support --LLAMA.CPP · VULKAN · AMD RX VEGA 56/64 Launcher
Converter fixes, NVFP4 findings & Strix Halo ROCmFP4 benchmarks for NVIDIA Nemotron 3.5 Lightning 30B-A3B
AI Compute Performance Analysis & Statistics Suite - RDNA4 AI Compute Toolset for Windows
A Windows desktop application for creating and applying custom fan speed curves to AMD GPUs. Built with WPF and WPF UI (Fluent)
Pilote un llama-server en router mode natif sur Strix Halo : modeles GGUF, models.ini genere, bench API, tuning MTP, service systemd
One egg. One image. Every language. A production-grade hosting platform that installs, updates, compiles and runs 50+ programming languages inside you...
High-performance shader cache and temporary app artifact cleaner built with PyQt6
Local-AI systems research: making large, sparse MoE models practical on constrained consumer hardware. Representation, MoE residency, VRAM/RAM/NVMe ti...
Explore AMD Adrenalin Hardware, FPS/Latency, and FrameTime logs. Runs entirely locally in your browser—your logs are never uploaded.
Find how a robot fails. Repair it. Prove it. Failure discovery → repair → statistical qualification for contact-rich manipulation, end to end on one A...
Automated CUDA-to-ROCm GPU kernel translation via semantic BridgeIR. 316 HIP API mappings, wavefront-aware optimization, MFMA targeting, 6 multi-targe...
LollozzOS Configurator is a utility I created (Lollozz) that includes all the tweaks and operations needed to simplify and optimize Windows 10/11. Eac...
Aegis is an AI-powered SRE tool that monitors cloud infrastructure, detects incidents, diagnoses issues, and performs automated remediation through a...
DEC (Durden’s Engine Collection) is a modular suite of source-to-memory tools for system-level programming. Originally developed as MALTOSE, DEC is ev...
[Case Study] A highly concurrent, config-driven fleet installer for managing drivers and runtimes across 20+ enterprise bare-metal AMD Instinct™ MI350...
Production ROCm llama.cpp build recipe for AMD Strix Halo (gfx1151, Ryzen AI MAX+ 395) — plus seven-model serving runbooks, KV-cache sizing against a...
Compiler for streaming time-series models on FPGAs: hardware costs predicted before synthesis, certified against machine-checked scheduling theorems (...
WhisperWriter fork with a whisper.cpp + Vulkan GPU engine for AMD GPUs/iGPUs on Windows (where faster-whisper/CUDA can't run). Dictate with Whisper la...
AMD rcraid Linux 6.14 DKMS port, NVMe RAID0 setup, mdadm/XFS tuning and fio benchmarks for high-speed workstation storage.
AMD GPU monitor for Linux and Windows with a terminal UI
Workaround for a vendor-side bug: Unsloth Studio's bundled llama.cpp prebuilt ships libhsa-runtime64.so.1 (1.21.x) which segfaults on AMD RDNA 3 GPUs....
Running LLMs locally on AMD GPUs, by people who actually do it. Cookbook + benchmark database for ROCm/HIP inference on Radeon, Instinct, and Strix Ha...
The flame graph for "why won't this model fit": live-trace local-LLM VRAM (weights vs KV cache) and predict max context before OOM. Zero-dependency Go...
AEGIS is a self-evolving WAF: when an attack slips past every static rule, an AMD GPU-hosted LLM proposes a new signature, a deterministic Rust engine...
Autonomous predictive maintenance intelligence powered by 3 AI agents on AMD MI300X GPU
Plug-and-play installer for ComfyUI on AMD GPUs (Ubuntu/Mint). ROCm setup, AllTalk/Wan2.2 workflows ready, on-demand launcher.
AMD Ryzen processor overclocking toolkit — rewritten in Go with beautiful terminal display.
Post-quantum secure enterprise AI gateway featuring client-side PQC encryption, dynamic hybrid workload routing, and isolated open-source LLM inferenc...
An end-to-end Financial Market Predictor and Retrieval-Augmented Generation (RAG) Coding Assistant powered by open-source LLMs on AMD Developer Cloud...
Kipp — A small, hand-written inference engine for a single model family on local hardware.
Manage agent configurations across projects with a single AGENTS.md file to standardize rules, skills, and permissions for any LLM-powered tool.
Windows-native AMD + DirectML lab for Qwen-style visual LoRA experiments.
Plug-and-play installer for Ollama + Open WebUI on AMD GPUs. ROCm setup, ComfyUI/AllTalk-compatible, zenity-driven UX.
"Modular Python system for recording, transcribing, and generating class notes."
An AI-powered command center that correlates AWS FinOps data with SecOps vulnerabilities using Fireworks AI and AMD Instinct MI300X.
Push-to-talk transcription using local OpenAI Whisper models with Vulkan/Metal GPU acceleration. Free, fast and accurate with LTS support for Windows,...
Ultra-fast LLM inference engine — Vulkan backend, no CUDA required, AMD/Intel/NVIDIA
Text embeddings on AMD XDNA NPU (Strix Halo) — OpenAI-compatible /v1/embeddings endpoint powered by all-MiniLM-L6-v2 running on the NPU tile array
Automatically detects your GPU and installs the correct version of PyTorch for your environment.
llama.cpp benchmarks on AMD RX 9060 XT 16GB — Vulkan vs ROCm (Linux)
Low-cost AMD/ROCm stationary-expert fabric for configurable trillion-parameter sparse MoE inference, with packed-INT4 HIP kernels, activation-only RoC...
ComfyUI for AMD Strix Halo (gfx1151) on Ubuntu 26.04 via distrobox + rootless podman
On-prem AI coding agent + LLM cost-router for regulated enterprises — 100% AMD, open-weight, air-gapped, Ed25519 tamper-evident audit. AMD ACT I 1st-p...
Operational control plane for reliable AI execution built with FastAPI, Next.js, Docker and Fireworks AI on AMD infrastructure.
Browser overlay that turns Japanese manga into invisible, selectable text for Yomitan dictionary lookups.
CScaler bridges upscaling/frame gen across GPUs. Supports DLSS2+/XeSS/FSR2+ inputs, replaces native upscalers, enables FSR-FG/XeFG on non-FG titles. S...
A generic thermal observability framework for CPU, GPU, board, and platform telemetry across vendor APIs, kernel interfaces, and runtime correlation l...
Zero-config LLM benchmarking on AMD GPUs with ROCm. Auto-detect MI300X/MI250/Radeon, CUDA and CPU fallback.
Compiler-backed personalized Python learning with real execution and graduation to standard syntax.
Windows memory exhaustion investigation and monitoring tools for Codex Desktop.