vgpu.rs is the fractional GPU & vgpu-hypervisor implementation written in Rust

ai nvidia gpu-utilization vgpu ai-infra fractional-gpu vgpu-hypervisor virtual-gpu

• Updated Oct 15, 2026
• Rust

Distributed Virtual GPU — aggregate multi-device GPU/CPU into one virtual GPU. Native dark dashboard. Zero config. Windows + Android. MIT.

android c go ai cross-platform gpu opencl distributed-computing grpc mdns gpu-cluster distributed-gpu llm-inference virtual-gpu

• Updated Jul 30, 2026
• Go

reims-vgpu on Windows 11: paravirtual Metal→Vulkan GPU for macOS guests under QEMU + WHPX. 5–7 → ~93 fps median on a Safari CSS animation (RTX 4060, 1080p macOS 13 guest). Experimental.

windows macos metal gpu vulkan hypervisor qemu whpx virtual-gpu paravirtualization

• Updated Nov 3, 2026
• Rust

AXON-vCXL-vGPU: Infinite VRAM & Zero-OOM AI Accelerator for Windows (Run 70B LLMs on 8GB/12GB/16GB GPUs)

windows inference vram vgpu llm deepseek llama3 ai-accelerator virtual-gpu zero-oom vcxl

• Updated Sep 28, 2026
• Python

DisplayLink — independent third-party profile of a public API surface, by API Evangelist. DisplayLink, a brand of Synaptics Incorporated, develops USB-based virtual graphics technology that pairs host driver software with a device-side decoder SoC to act as a virtual graphics card.

company hardware synaptics semiconductors displays docking-stations apis-json virtual-gpu usb-graphics

• Updated Oct 29, 2026