Netris has raised $15 million in a Series A round led by Andreessen Horowitz to expand its network automation platform for GPU data centres and AI cloud operators.
The company says its software helps operators configure complex networks, isolate customer workloads and bring expensive GPU clusters online faster. Andreessen Horowitz partner Guido Appenzeller led the investment and will join Netris’ board.
Platform Targets Delays in GPU Cluster Deployment
Building an AI data centre requires more than installing processors, storage and network switches. Operators must configure several network layers, connect infrastructure components and ensure that multiple customers can use the same cluster without gaining access to one another’s resources.
These tasks can require months of engineering work, leaving costly GPUs idle before the facility begins generating revenue. Smaller neocloud providers may lack the large infrastructure teams maintained by established cloud companies.
Netris’ NAAM platform, short for Network Automation, Abstraction and Multi-Tenancy, provides a central control system for configuring and operating those environments. It can automatically apply changes across switches and enforce customer isolation through the networking hardware.
The company says operators can use the platform to launch GPU clouds within weeks and provision new tenants without manually configuring every network component. It also supports the reallocation of GPU capacity as customer requirements change.
Netris Supports Multiple Networking Systems
Netris works across several technologies used in AI infrastructure, including Ethernet, Nvidia Spectrum-X, Nvidia Quantum InfiniBand, NVLink systems, BlueField data processing units and virtual networks.
Its integration with Nvidia DSX Air allows operators to simulate and validate network configurations before physical hardware arrives. The same Netris controller used in the simulated environment can later be transferred to the live cluster.
This approach can help engineers detect cabling, topology and configuration problems before the GPUs are installed. It also reduces the need to test changes on hardware already running paid customer workloads.
Chief executive Alex Saroyan said Netris relies on deterministic automation rather than generative AI to configure networks. The company considers predictable and repeatable execution more appropriate than probabilistic models when applying changes across thousands of switches.
Funding Will Support Hiring and Expansion
Netris reports more than 35 live deployments across neoclouds, sovereign AI providers and AI factories. Customers include Lightning AI, TensorWave, Telus, HPE, Visionbay, Yotta and Firmus.
The funding announcement follows 800% growth in annual recurring revenue over the previous 12 months. Netris plans to hire more engineers and sales employees, add support for additional hardware suppliers and expand its international operations and partner network.
Featured image credits: Magnific.com
For more stories like it, click the +Follow button at the top of this page to follow us.