synchopate/llama-modchip
Resource
Can I trust it?
Worth a closer look before you use it.
Warnings Check
Very new: it was created 7 days ago. Wait a little before installing it.
Stars from real accounts No data
We haven't checked its stars for fakes yet.
Active OK
Last change today.
Licence OK
Apache-2.0: you can use it for free, also in a commercial product. Keep the licence notice.
Quick checks from public data. They help, but they are no guarantee.
What it is
Describes a research write-up and engineering approach for running a 30B hybrid Mamba-2 MoE model (Nemotron) on a Ryzen 9 9955HX CPU with a 256K context window, using llama.cpp with custom AVX-512 kernels, cache tuning, and a draftless speculative decoding engine. The README…
Who it's for
Developers and researchers interested in CPU-only LLM inference, llama.cpp internals, and AVX-512 optimization
How to use it
Ask your AI to set it up
Paste this into Claude, ChatGPT or your coding assistant.
Help me install and set up synchopate/llama-modchip in my project. Follow its README: https://github.com/synchopate/llama-modchip. Explain each step and check it is safe before running anything.
Star curve
Alternatives
Similar projects named in its profile. Each link searches the directory.
Get alerts for repos like this
We'll email you when something in Models & inference starts rising. Free, once a week, only what matches.
What arrives: one email on Mondays with the repos that match, never more. See last week's. Privacy
For maintainers
Add this badge to your README. It shows the repo's place on the radar, updates every hour and links back to this page.
[](https://repoalerts.com/repo/synchopate/llama-modchip?ref=badge)Similar repos
- freriadoc/Avalon
A fork of llama.cpp tuned for running the Qwen3.8-Flash-Next 125B MoE model on a single consumer AMD GPU with experts offloaded to system RAM. It uses an expert cache in VRAM, CPU computation of missi
- prabanta-dev/janas
A C inference engine for running large mixture-of-experts language models on ordinary computers, with or without a GPU. Weights stay on disk, and the experts each token needs are streamed from an NVMe
- wici-ai/inferssd
Runs large mixture-of-experts GGUF models that exceed available RAM or VRAM by streaming expert weights from SSD. Built on llama.cpp, it provides a terminal chat, a browser UI and an OpenAI-compatible
- MaxDevv/lean-qwenflash
Inference runtime for the Qwen3.8-Flash-Next 125B MoE model that streams uncached expert weights from NVMe via O_DIRECT into a bounded GPU cache, targeting full 262k context on 16 GB VRAM and 8 GB sys
- Edge0-AI/Edge0
A streaming MoE inference framework that runs large sparse-MoE models on consumer hardware by streaming expert weights from SSD on demand, so peak memory depends on the active experts. A trained prero
- eagle-hq/EAGLE-4
Official inference code for EAGLE-4, a speculative decoding system whose drafter adapts its speculation to batch size, hardware and workload. It serves EAGLE-4 drafts on SGLang, vLLM and TokenSpeed th