Netris raises $15M Series A from a16z to help new AI-powered clouds run faster

It has encouraged the boom in artificial intelligence Everyone and his uncle To launch the data center business. But setting up a data center is not easy.

Even if you solve it Insurance problem GPUs, network switches, storage, you still have to get everything up and running and be able to meet diverse customer needs. Setting up a data center ready to deliver cloud computing, AI inference and training services can take months of work. The longer it takes to get to market, the more expensive it will be for all those precious GPUs to sit idle.

Start network automation Nitris It claims it can make this problem go away for new fasteners. The company provides software that runs on network switches and a platform that connects to the switches to help neocloud operators reduce go-live time by automating setup, configuration and operations. The platform also provides network abstraction, so hardware configurations can be changed as needed, and it isolates servers and resources at the hardware layer so new clouds can serve multiple (multi-tenant) clients.

If this sounds like a solution to an obvious problem, you’re not wrong. Until recently, data centers were largely the domain of large infrastructure players like Equinix, NTT, Digital Realty, Oracle, Microsoft, AWS, or Google. These companies have largely solved the process of network setup, configuration, and multi-tenancy for themselves by hiring ranks of engineers or building the automation system themselves. New cloud companies rarely have such resources at their disposal.

“As a GPU cluster operator, you need to make configuration changes on every link, every day. In traditional data centers, they were using what’s called SDN (software-defined networking) to do that, but SDN is not enough, because it’s a software technology,” Netris CEO Alex Saroyan told TechCrunch. “For AI, software is not so good, because the traffic volume is very high, and everything has to be hardware-accelerated. So you need something like SDN, but with full hardware acceleration. That’s what we’re doing, and that’s what we’ve been doing for eight years.”

The Netris platform is vendor-agnostic and is compatible with networking equipment and standards used in data centers, for both Nvidia and AMD servers, Saroyan said.

The startup’s promise has found many believers, including Nvidia. Two years ago, the chipmaking giant was so impressed by a demonstration of Netris’ technology that it recommended the company to several customers. Today, Netris runs on more than 35 GPU clusters around the world (about 1 million GPUs in total), operated by companies like Lightning AI, Foxconn, Visionbay, Hewlett Packard Enterprise, Tensorwave, Telus, and others.

Building on this momentum, Netris has now raised $15 million in a Series A round from Andreessen Horowitz, TechCrunch has learned exclusively.

Notably, there is no artificial intelligence at work here. Sariwan said the company only uses algorithms it previously developed to run and configure automation and processes.

“We started before AI. We understood the challenge early, and we started developing this algorithm early. AI is not deterministic, right? Sometimes it likes to do things on its own. It’s good for creative work, but to change several thousand transformer configurations, you don’t need to be creative. You have to be very persistent and repeatable.”

a16z Partner Guido Appenzeller joins the company’s board of directors. Looking to the future, Netris aims to use the funding to hire more engineers and sales staff, add support for more hardware vendors, and implement more functionality into its algorithm.