The rise of AI has encouraged everyone and their uncle to launch a data center business. But starting a data center is not easy.
Even if you solve the problem of securing GPUs, network switches, and storage, you still need to get it all set up, running, and able to meet diverse customer needs. Preparing a data center to provide cloud computing, AI inference, and training services can take months of work. And the longer it takes to hit the market, the higher the cost of having all those precious GPUs sitting idle.
Getting Started with Network Automation Netris He claims he can make that problem go away for neoclouds. The company provides software that runs on network switches and also offers a platform that connects to switches to help neocloud operators reduce time to go live by automating installation, configuration and operations. The platform also provides network abstraction, so hardware configurations can be changed as needed, and isolates servers and resources at the hardware layer so that neoclouds can serve multiple clients (multi-tenant).
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 operators such as Equinix, NTT, Digital Realty, Oracle, Microsoft, AWS or Google. Those companies pretty much solved network installation, configuration, and multi-tenancy themselves by hiring engineers or building the automation themselves. Small neocloud businesses rarely have those 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 used something called SDN [software-defined networking] “To do this, but SDN is falling short, because it is a software technology,” Netris CEO Alex Saroyan told TechCrunch. “For AI, software is not right, because the amount of traffic is so high that everything has to be hardware accelerated. So you need something like SDN, but completely hardware accelerated. This is what we do and this is what we have been doing for eight years.”
Saroyan said the Netris platform is vendor agnostic and compatible with networking equipment and standards used in data centers, for both Nvidia and AMD servers.
The startup’s promise has found many believers, one of whom is Nvidia. Two years ago, the chipmaker giant was so impressed with a demonstration of Netris technology that it recommended the company to several clients. Today, Netris is active on more than 35 GPU clusters around the world (about a million GPUs in total), operated by companies such as Lightning AI, Foxconn, Visionbay, Hewlett Packard Enterprise, Tensorwave, Telus, and others.
To build on that momentum, Netris has now raised $15 million in a Series A round from Andreessen Horowitz, TechCrunch has exclusively learned.
In particular, there is no AI at work here. Saryoan said the company only uses algorithms it had previously developed to run and configure automation and operations.
“We started long before AI. We understood the challenge from the beginning and started developing this algorithm from the beginning. AI is not deterministic, right? Sometimes it likes to do things on its own. It’s good for creative work, but to change thousands of switch settings, you don’t need to be creative. You need to be very persistent and repeatable.”
a16z partner Guido Appenzeller will join the company’s board of directors. Looking ahead, Netris intends to use the funds to hire more engineers and sales staff, add support for more hardware vendors, and implement more features into its algorithm.
When you purchase through links in our articles, we may earn a small commission. This does not affect our editorial independence.