Some infrastructure players walk into the AI boom like it’s a gold rush. GigaIO walked in with blueprints, soldering irons, and a composable fabric that makes hyperscale hardware play nice without the corporate hangover. Now they’ve secured $21 million in the first tranche of their Series B, and if you’re in the inferencing game and not paying attention, you’re doing AI on dial-up.
Led by Impact Venture Capital, with CerraCap Ventures, G Vision Capital, Mark IV Capital, and SourceCode Cerberus jumping in, this round doesn’t just validate the tech. It confirms what the smartest players in the room already know: scaling AI isn’t just about more chips. It’s about making those chips talk to each other like they’ve known each other since Stanford dorms.
Big congrats to Alan Benjamin, Co-Founder, President & CEO, and Joey Maitra, Founder, Chief Architect & CTO. These aren’t your typical startup suits. Alan’s built global tech orgs and still carries that public accounting edge like a sharpened pencil behind the ear. Joey? He’s the architect behind FabreX, the composable infrastructure fabric that turned PCIe into a distributed network ninja.
SuperNODE is rack-scale muscle for the data center. Gryf is a suitcase-sized supercomputer for the edge that’s not playing tourist. And FabreX is the connective tissue, low-latency, high-bandwidth, PCIe/CXL magic that links GPUs, accelerators, and CPUs in memory-to-memory conversation, without the middleware migraine. The real kicker? Vendor-agnostic support across NVIDIA, AMD, and whoever dares to show up next.
And this isn’t vaporware. The San Diego Supercomputer Center tapped them for the NSF’s Prototype National Research Platform. Collaborations with SK Hynix, AMD, and NVIDIA are already in motion. When the edge, cloud, and core need to share data in real time, GigaIO isn’t just building roads, they’re laying autobahns through legacy traffic jams.
What’s next? Scaling production. Ramping up R&D. Expanding global sales and engineering teams. With ~50 employees and plans to double headcount by end of 2025, the team is laser-focused on growth markets across financial services, life sciences, telco, and autonomous systems.
Inferencing isn’t theory anymore. It’s the new operational backbone. If your AI infrastructure is built like a band-aid on a supercomputer, GigaIO’s coming for your budget, and your lunch.
