# lupinemachines/lupine

LUPINE is a GPU over IP bridge allowing GPUs on remote machines to be attached to CPU-only machines.

Repository: https://github.com/lupinemachines/lupine
Canonical: https://ross.abutalabs.com/products/lupine
Homepage: https://lupine.sh
Language: C++
License: Apache-2.0
License Family: permissive
Topics: cuda, gpu, remote-access, networking, cublas, cudnn, nvml, mlops
Last push: 2026-08-26T16:05:19+00:00

## Health v2 (maintenance only)
Score: 89/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 97, longevity 52
- inputs: {"age_days": 735, "days_push": 7, "days_rel": 22, "gap_med": 7, "n_releases_24m": 8}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2382, forks 132 (observed 2026-08-28T04:06:42.330134+00:00)

## What it is
LUPINE is a GPU-over-IP bridge that lets machines without local GPUs attach NVIDIA GPUs running on remote servers over the network. It exposes CUDA, cuBLAS, cuDNN, and NVML remotely so tools like nvidia-smi and PyTorch work transparently on CPU-only clients.

## Use cases
- use a remote GPU from a CPU-only Linux machine
- run CUDA workloads on a Mac with no NVIDIA GPU
- attach GPUs on remote servers to local containers
- run nvidia-smi against a GPU on another machine
- give CI or dev containers access to remote GPUs
- share a single GPU server across multiple client machines

## When to choose
- your client machine has no NVIDIA GPU but you need CUDA
- you want transparent remote GPU access without rewriting code
- you want to centralize GPUs on servers and share them over the network
- you develop on macOS but need to test CUDA code

## When to avoid
- you need maximum performance and a local GPU is available
- you need graphics/display workloads rather than compute
- you need non-NVIDIA GPU support
- your network latency to the GPU server is high

## Facets
- artifact type: service
- maturity: active
- function: networking, gpu-computing, sdk, developer-tools
- domain: gpu-computing, microservices, machine-learning, developer-tools, cloud-computing
- platform: cross-platform, cpp
- tags: gpu-over-ip, cuda, remote-gpu, nvidia, cublas, cudnn, nvml, mlops, network-attached-gpu, linux, macos, docker

## Member repositories
- lupinemachines/lupine (main) score 89

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:42.330134+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:34:43.108640+00:00, confidence not recorded.
  - readme: https://github.com/lupinemachines/lupine (fetched 2026-08-28T04:06:42.330134+00:00, sha 47c604aa4a78)
  - homepage: https://lupine.sh (fetched 2026-08-29T10:15:36.573129+00:00, sha c3a6d1495205)
- Data as of 2026-08-30T08:39:29.467469+00:00.