Dev Claims Fix for NVIDIA's Intentional P2P Nerf on Consumer GPUs
QuixiAI · x · 2026-09-06
Developer QuixiAI says he identified and fixed what he claims is an intentional NVIDIA nerf of P2P support in multi-GPU consumer configurations, and submitted a patch — though he doubts it will be upstreamed. The finding suggests NVIDIA restricts consumer cards' multi-GPU direct connectivity in software.
Related event: Open-source fork unlocks P2P multi-GPU on consumer RTX cards(2 posts)→
More from Infra
- AMD exec: AI token processing could hit 120 quadrillion per month by 2030 — zephyr_z9 · 2026-09-06
- MiniMax-H3 Lip Sync on 8GB VRAM: Multishot Renders in 26 Minutes — big-boss_97 · 2026-09-06
- Oura's S-1 reveals an on-device AI stack: small models and edge compute, not cloud inference — eurie_kim · 2026-09-06
- Bosgame M5 Max With Ryzen AI Max Pro 495 and 192GB RAM Arrives October 2026 — Terminator857 · 2026-09-06
- A pragmatic guide to local agentic LLMs: compile buun's llama.cpp free on GitHub runners, Qwen3.8 27B quants span 30x — apollo_mg · 2026-09-06
- Cloud in a Bottle Launches to Make Self-Hosting Accessible to Everyone — zplizzi · 2026-09-06