Table of Contents
TL;DR
Netris raised $15M from a16z for its GPU network automation platform, now live at 35-plus clusters with 800 percent annual revenue growth.
Netris, a Santa Clara startup that automates the networking layer inside GPU data centres, has raised $15 million in a Series A round led by Andreessen Horowitz. The round follows what the company says is 800 percent annual recurring revenue growth and more than 35 live deployments at GPU clusters around the world, including operations run by Lightning AI, Foxconn-backed Visionbay, Hewlett Packard Enterprise, TensorWave, and Telus.
The problem Netris addresses is not the one that gets the most attention in the AI infrastructure boom. Neoclouds have raised billions of dollars to buy GPUs and build data centres, but getting those facilities operational requires configuring the network fabric that connects thousands of servers, a process that can take months and leave expensive hardware sitting idle.
A single GPU server carries at least three north-south connections, 16 east-west connections, and four NVL72 links, according to the company. Every time a tenant is added, resized, or removed, the network must be reconfigured across every layer simultaneously, sometimes across hundreds or thousands of switches at once. One misconfiguration can take a cluster down or leak one customer’s data to another.
Netris sells software that runs on network switches and a platform that automates setup, configuration, and operations for neocloud operators. The platform also provides network abstraction, so hardware configurations can be changed as required, and it isolates servers and resources at the hardware layer to support multi-tenancy. The company calls the approach NAAM, for Network Automation, Abstraction, and Multi-Tenancy.
The 💜 of EU tech
The latest rumblings from the EU tech scene, a story from our wise ol’ founder Boris, and some questionable AI art. It’s free, every week, in your inbox. Sign up now!
CEO Alex Saroyan told TechCrunch that traditional software-defined networking falls short for AI workloads because the volume of traffic requires everything to be hardware-accelerated. “For AI, software is not okay, because the amount of traffic is so high, everything must be hardware accelerated,” he said, adding that Netris has been building hardware-accelerated network automation for eight years.
Nvidia, which has committed more than $40 billion to AI infrastructure investments this year, was an early validator. Two years ago the chipmaker was impressed enough by a demo that it began recommending Netris to its own customers. Saroyan said the platform is vendor-agnostic, compatible with networking equipment from both Nvidia and AMD.
The startup claims its platform is now live at more than 35 GPU clusters totalling roughly one million GPUs. The neocloud sector has seen a wave of funding in recent months, with companies like Runpod reaching billion-dollar valuations on the strength of the AI compute shortage. Netris occupies a different layer of the stack, selling the infrastructure software those GPU cloud operators need rather than competing with them for compute customers.
Guido Appenzeller, a partner at a16z who previously co-founded Big Switch Networks and served as CTO at VMware’s Cloud and Networking division, led the round and is joining the Netris board. “GPU clusters run across many fabrics at once, and legacy automation was never built for that,” Appenzeller said, calling Netris the platform that AI cloud operators are standardizing on.
One notable detail is that Netris does not use AI in its own product, relying instead on deterministic algorithms it developed before the AI boom. Saroyan argued that creativity is the wrong trait for a system responsible for changing thousands of switch configurations. “AI is not deterministic,” he said, adding that network configuration requires persistence and repeatability, not creativity.
Netris was founded in 2018 and operates teams across the United States, the United Kingdom, Taiwan, Australia, Armenia, and India, with Singapore opening this year. The company plans to use the funding to hire engineers and sales staff, add support for more hardware vendors, and expand the functionality of its automation platform.
If you liked the article, do not forget to share it with your friends. Follow us on Google News too, click on the star and choose us from your favorites.
If you want to read more like this article, you can visit our Technology category.
