The burgeoning demand for artificial intelligence has spurred a surge in new data center ventures, often referred to as ‘neoclouds.’ However, establishing these specialized facilities, particularly those geared towards AI inference and training, presents significant logistical and technical hurdles. Beyond sourcing essential hardware like GPUs, network switches, and storage, configuring and operationalizing these components to meet diverse client needs can consume months. This delay translates directly into increased costs, as valuable hardware remains idle.

Network automation startup Netris aims to streamline this complex process. The company offers a software solution that operates on network switches, complemented by a platform designed to automate the setup, configuration, and ongoing management of neocloud infrastructure. This automation significantly reduces the time-to-market for new AI data centers. Furthermore, Netris’s platform provides network abstraction, allowing for flexible hardware configuration changes, and implements hardware-level isolation to support multi-tenancy, enabling neoclouds to serve multiple clients securely and efficiently.

Historically, the intricacies of data center network setup and multi-tenancy were managed by large-scale operators like Equinix and Google, who possessed the resources to develop in-house automation or employ extensive engineering teams. Smaller neocloud businesses typically lack these advantages. Netris CEO Alex Saroyan highlighted the limitations of traditional Software-Defined Networking (SDN) for AI workloads, emphasizing the need for hardware-accelerated solutions due to the immense traffic volumes. “For AI, software is not okay, because the amount of traffic is so high, everything must be hardware accelerated. So you need something like SDN, but completely hardware accelerated. This is what we do,” Saroyan stated, noting the company’s eight years of experience in this domain.

Netris’s vendor-agnostic platform supports standard data center networking equipment and is compatible with servers from major providers like Nvidia and AMD. The company’s technology has garnered significant attention, including a notable endorsement from Nvidia two years prior, which led to Netris being recommended to several clients. Currently, Netris’s solutions are deployed in over 35 GPU clusters globally, collectively supporting approximately one million GPUs, with clients including Lightning AI, Foxconn, and Hewlett Packard Enterprise.

To fuel its growth, Netris has successfully closed a $15 million Series A funding round led by Andreessen Horowitz (a16z). Guido Appenzeller, a partner at a16z, will join Netris’s board. The company plans to utilize this investment to expand its engineering and sales teams, broaden support for additional hardware vendors, and further enhance its proprietary automation algorithms. Saroyan clarified that Netris’s success is built on its independently developed algorithms, predating the current AI boom, and that its approach prioritizes deterministic, repeatable operations over the creative, less predictable nature of AI for critical infrastructure management.

Key Takeaways

• Netris has raised $15 million in Series A funding led by Andreessen Horowitz to enhance its network automation platform for AI data centers.
• The company's software and platform automate the setup, configuration, and operation of neocloud infrastructure, significantly reducing deployment time and costs.
• Netris's hardware-accelerated approach addresses the high-traffic demands of AI workloads, offering a solution distinct from traditional software-based SDN.
• The funding will be used to expand the team, support more hardware vendors, and further develop the company's automation algorithms.

Editor’s Analysis & Impact

This funding round underscores the critical bottleneck in scaling AI infrastructure: the complex and time-consuming deployment of specialized data centers. Netris’s focus on hardware-accelerated network automation directly addresses this pain point, positioning the company as a key enabler for the rapidly growing neocloud market. The backing from Andreessen Horowitz, a prominent venture capital firm with a strong track record in technology, signals significant market confidence. As AI adoption accelerates, the demand for efficient, scalable, and rapidly deployable infrastructure will only intensify, suggesting a strong future outlook for companies like Netris that can deliver tangible solutions to these pressing challenges.

Frequently Asked Questions

Q: What problem does Netris solve for AI data centers?
A: Netris solves the problem of lengthy and complex setup times for new AI data centers (neoclouds). Its software and platform automate network configuration and operations, reducing the time it takes to get these facilities ready to serve customers and minimizing costly hardware downtime.

Q: Why is Netris's approach different from traditional SDN?
A: Netris utilizes a hardware-accelerated approach, which is crucial for the high-volume traffic generated by AI workloads. Traditional Software-Defined Networking (SDN) is software-based and can be insufficient for these demands. Netris's solution ensures that network operations are fast, reliable, and can handle the intense data flow required for AI training and inference.

Q: Who are some of Netris's existing clients?
A: Netris's technology is currently deployed in over 35 GPU clusters globally, supporting major clients such as Lightning AI, Foxconn, Visionbay, Hewlett Packard Enterprise, TensorWave, and Telus.