Founding Phase — Turing Registry Now Open
MATR·IX — Massive Access Turing Registry IX — is a purpose-built internet exchange for NeoClouds, IX Factories, and organizations moving AI workloads at scale. 400G minimum ports. Members only. Built by the people who built the IX industry.
The industry pays an estimated $70–80 billion a year in cloud egress fees. No existing internet exchange was built to solve that problem for AI-scale workloads.
MATR·IX exists to change that math.
No legacy ports. No 10G. No 100G. MATR·IX starts where AI-scale data movement begins. Purpose-built on an 800G-native 51.2Tbps OSFP fabric. If you're not thinking in terabits, MATR·IX isn't for you — and that's by design.
Access to the MATR·IX fabric requires Turing Registry membership. Every participant is vetted. Every port is intentional. This isn't a general-purpose exchange — it's a curated AI interconnection community.
Built from the ground up for NeoClouds, IX Factories, and the organizations serving AI workloads. No legacy ISP politics. No CDN traffic shaping. Just the interconnection your AI infrastructure actually needs.
Founding members don't just use the fabric — they shape it. Voting rights, governance participation, and co-founder upside for the organizations that help build MATR·IX from day one.
Built around where AI infrastructure actually operates today. If your colos are here, MATR·IX is coming to you.
The founding cohort is capped at 6 organizations. These are the companies that define AI-scale interconnection in the US — and lock their rates before the rest of the market catches up.
MATR·IX is being built by a founding team with deep roots in internet exchange operations, network engineering, and AI infrastructure — people who have seen firsthand what existing exchanges can and cannot do at scale.
Our founding expertise includes co-founding IX-Denver — the largest internet exchange in the Rocky Mountain West — chairing OIX, the global interconnection standards body, over 20 years of network engineering, 8 patents in network technology, two books, an IETF RFC, and a track record of building neutral, community-driven infrastructure that lasts.
MATR·IX wasn't dreamed up in a boardroom. It came from conversations with the NeoClouds and operators who are living this problem today — and who told us plainly that nothing on the market is built for what they actually need.
Help Found the FutureSubmit your application. We review your network, your use case, and your fit. MATR·IX is intentionally selective — only organizations with genuine AI-scale interconnection needs qualify for membership.
Our founding team reviews every application personally. We're building a community, not a customer list. Expect a direct conversation with our team — no sales calls, no pitch decks.
Once approved, you'll work with our operations team to provision your 400G (or larger) port. US-first, with global expansion on the roadmap. Founding members get priority port allocation as the fabric scales.
You're on the fabric. Peer with other MATR·IX members directly. Move training datasets, model checkpoints, fine-tuning corpora, and batch inference outputs between NeoClouds at the speed and scale the work actually demands — without the egress tax that makes every petabyte a budget conversation.