Berth: a lightweight open-source GPU resource manager with multi-GPU scheduling and VRAM parking
Distinct-Pie2389 · reddit · 2026-11-19
A developer announced berth, a lightweight GPU resource manager built to stop local LLMs, image generation, and voice cloning from fighting over VRAM.
Features include parking engines in system RAM, parking via CUDA checkpoints, priority mapping, scheduling, lending part of a GPU's VRAM pool, and leasing whole GPUs via VFIO passthrough. Even with a single dedicated engine, berth keeps squatters (nodes, stale or orphaned processes) out of its VRAM. The author aims to solve VRAM issues at the driver level while staying lightweight and durable.
More from Infra
- AMD Loom boosts Qwen3.5-4B prefill up to 2.6x on Radeon 890M laptops — jfowers_amd · 2026-11-19
- Dev submits PR adding periodic checkpoints to GKE sandboxes so agents don't lose progress — yoimnotkesku · 2026-11-19
- AWS Lambda Adds GPU Support via Managed Instances for Inference and Media Workloads — DavidWells · 2026-11-19
- Nvidia Allegedly Made GPUs Artificially Scarce to Keep Prices Sky-High — GaryMarcus · 2026-11-19
- AWS adds private CA support to Bedrock AgentCore Identity for connecting agents to self-hosted IdPs — AWS ML Blog · 2026-11-19
- Leak Hints at Classified 1M-GPU Training Cluster in China That Skips HBM — teortaxesTex · 2026-11-19