ampere

llama.cpp fork for significantly improved performance on Ampere (especially RTX 3090 / 3090 Ti): example: 110+ tok/s over a 100K-token generation at temperature 1 for qwen3.8, supporting full 262K ctx. Supports low memory (12gb) configs for 200k+ ctx

Arduino library for INA226 power sensor

A custom LLM inference engine for low-cost GPUs: NVIDIA cards from a 1 GB Quadro K420 to an RTX 3080, the AMD Radeon 780M and the CPU. Runs the Bonsai 2B to 27B models and Nemotron with an OpenAI-compatible server.

Pre-built wheels for llama-cpp-python across platforms and CUDA versions

GLM-5.3-Flash on 4× NVIDIA CMP 170HX with vLLM: up to 482.6 tok/s single-user, up to 943.7 tok/s at 8 users, 262K context, TP4 or PP4, one-command container install

Enable NVIDIA Smooth Motion on RTX 30-series / Ampere SM86 - one-click Windows utility, no per-game DLSS-FG integration required.

Ampere optimized llama.cpp

llama.cpp speculative decoding measured on one RTX 3090, Qwen3.6-35B-A3B UD-Q4_K_XL, commit 3737e4137. Published figures are re-derived from the committed data by a checker that fails on drift, a coverage probe reports how many of them it actually covers, and ERRATA.md lists this study's own retracted claims.

Cross-platform FlashAttention-2 Triton implementation for Turing+ GPUs with custom configuration mode

Setup instructions for running Valheim on Oracle Cloud Infrastructure using Arm also available at https://codeberg.org/husjon/valheim_server_oci_setup

vLLM for NVIDIA CMP 170HX (Ampere sm_80): fast, repeatable model serving on one card or several

AML's goal is to make benchmarking of various AI architectures on Ampere CPUs a pleasurable experience :)

One-command installer + terminal control panel for a modded Project Zomboid Build 42 dedicated server on ARM64 (Oracle Cloud Ampere / box64).

Arduino energy monitor, using SCT-013-030 current sensors

DACAN (Дацан): Qwen3.8-Flash-Next (125B MoE) on any NVIDIA RTX 20/30/40/50 card and any x86-64 CPU with AVX2, AVX-512 not required. One or two cards, one or two sockets, context up to 524K.

DeepSeek-V4.1-Flash on 4× NVIDIA CMP 170HX with vLLM: up to 319 tok/s for one user, up to 765 tok/s at eight, 262K context, one-command container install

Performance characterization and analysis tool for Ampere CPUs

GPTQ optimized for Ampere (RTX 3090/3060) — Marlin JIT nvcc compatibility fixes included

We are a fosshost project which is delivering ARM-based hardware into multiple, global data centers. We document and keep a diary of our project daily.

Orcle OCI Always Free Ampere A1.Flex capacity monitor and safe retry launcher for 2 OCPU / 12 GB VM deployments.