{"slug": "thunderbolt-ibverbs-we-have-infiniband-at-home", "title": "Thunderbolt-ibverbs: We have InfiniBand at home", "summary": "A developer created a Linux kernel module and userspace shim that transforms a standard USB4 connection into a low-latency InfiniBand device, enabling distributed AI inference across two 128GB Strix Halo mini PCs. The project achieved sustained bidirectional throughput of approximately 95 Gb/s and one-way latency of 7 microseconds, significantly outperforming onboard 2.5 GbE and soft-RoCE over Thunderbolt networking. The code, generated primarily by AI models, is released for research use only with no warranty.", "body_md": "# thunderbolt-ibverbs: We have InfiniBand at home\n\nI spent the past few weeks working on this project, I thought it might be interesting to write up a technical report on it, the motivation, the process, learnings, etc.\n\n**DISCLAIMER:** all of the code in this repo ([github.com/hellas-ai/thunderbolt-ibverbs](https://github.com/hellas-ai/thunderbolt-ibverbs)) is AI-generated (mostly Codex 5.5 and Opus 4.7) — while I made an effort to understand enough of it to keep it on-track, I almost certainly failed in many instances and I'm sure the code contains many false assumptions, hallucinations and plain stupidity. No warranty or guarantee offered, for research use only, not for human consumption.\n\n**TL;DR.** We write a linux kernel module and userspace shim to pretend our generic usb4 connection is a low-latency, high-performance InfiniBand device and use it to perform distributed inference across two 128GB Strix Halo mini PCs. Basic interop with Apple's native protocol is functional.\n\n**~48 Gb/s per direction (~95 Gb/s bidi total)** sustained`ib_write_bw`\n\n, 4-HCA aggregate at 1 MiB / 8 QPs with IOMMU off — vs`~2.3 Gb/s`\n\nover the onboard 2.5 GbE and`~9 Gb/s`\n\nfor soft-RoCE on top of`thunderbolt-net`\n\nat the per-rail level.**~7 µs** one-way`ib_write_lat`\n\nat 64 B, single QP — vs`~28 µs`\n\nover RXE/2.5 GbE and`~65 µs`\n\nover RXE/TBnet.", "url": "https://wpnews.pro/news/thunderbolt-ibverbs-we-have-infiniband-at-home", "canonical_source": "https://blog.hellas.ai/blog/thunderbolt-ibverbs/", "published_at": "2026-05-29 01:03:20+00:00", "updated_at": "2026-05-29 01:15:03.864289+00:00", "lang": "en", "topics": ["ai-infrastructure", "ai-research", "machine-learning", "artificial-intelligence", "large-language-models"], "entities": ["InfiniBand", "Thunderbolt", "Strix Halo", "Apple", "Codex", "Opus", "hellas-ai", "github.com/hellas-ai/thunderbolt-ibverbs"], "alternates": {"html": "https://wpnews.pro/news/thunderbolt-ibverbs-we-have-infiniband-at-home", "markdown": "https://wpnews.pro/news/thunderbolt-ibverbs-we-have-infiniband-at-home.md", "text": "https://wpnews.pro/news/thunderbolt-ibverbs-we-have-infiniband-at-home.txt", "jsonld": "https://wpnews.pro/news/thunderbolt-ibverbs-we-have-infiniband-at-home.jsonld"}}