- The a16z team behind Netris includes Nicira's founder, a former VMware CEO and the co-founder of Big Switch — SDN's founding generation
- Netris applies SDN's original vision to GPU clouds: a single control plane across multivendor network fabrics, without relying on AI
- Netris posted 800% ARR growth in 12 months with 35+ live deployments across neoclouds, sovereign AI providers, and AI factories globally
Netris, an emerging Silicon Valley network automation provider, raised $15 million in a Series A led by a team at Andreessen Horowitz that helped pioneer software-defined networking nearly 20 years ago.
"A16Z and Netris get the SDN band back together," said R. Scott Raynovich, Futuriom founder and principal analyst. "The a16z investment team is a 'Who's Who' of experts from the early days of SDN. It's the original team that took startup Nicira and integrated it into VMware to build one of the industry's most successful virtualized networking platforms, NSX," Raynovich wrote.
A16z general partner Guido Appenzeller led the round and is joining the Netris board. Appenzeller previously co-founded Big Switch Networks — later acquired by Arista — and served as CTO of VMware's Cloud and Networking Division.
Also backing the deal are A16z general partners Martin Casado, who founded Nicira before VMware paid $1.26 billion for it in 2012, and former VMware CEO Raghu Raghuram.
"Netris is doing what SDN was trying to do," Appenzeller wrote on LinkedIn, "but now switches are open enough that you are no longer limited to soft switches and whitebox switches running your own software, but can control the entire stack."
The networking bottleneck GPU clouds can't escape
The funding targets a specific problem for neoclouds: network configuration. "Network automation is where operators get stuck," Netris said.
A single GPU server runs multiple networks. These are at least three north/south, 16 east/west and four Nvidia NVL72 connections. 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. One misconfiguration can take down an entire cluster or, worse, expose one tenant's data to another.
Netris' NAAM platform — Network Automation, Abstraction, and Multi-Tenancy — automates those operations and enforces hard multi-tenancy in networking hardware rather than software. It gives operators a single control plane across every fabric: Ethernet, InfiniBand, the NVL72 scale-up fabric, and virtual and edge networking. Raynovich writes that Netris "may have solved one of the key bottlenecks in AI infrastructure." The timing is opportune: neocloud revenue topped $25 billion in 2025 and analysts project it growing at a 69% annual clip through 2030.
Netris claims 800% ARR growth in the past 12 months, with 35-plus live deployments spanning neoclouds Lightning AI, STN and TensorWave; sovereign AI operators Telus, DCAI and YOTTA; and AI factories including Foxconn-backed Visionbay.ai, which operates Taiwan's largest GPU cluster, and Firmus, which operates Australia's largest renewable-powered sovereign AI factory. HPE also uses the platform to deliver full-stack AI solutions for research, education and government customers.
No AI required
Surprisingly, the Netris platform does not use AI. CEO Artur Saroyan told TechCrunch the company runs on deterministic algorithms developed long before the current AI wave. "AI is not deterministic. Sometimes it likes to do things on its own. It's good for creative work, but for changing many thousands of switch configurations, you don't need to be creative. You need to be very persistent and repeatable."
That's a contrast to the competition. Cisco launched Cloud Control at its annual customer conference in Las Vegas this month — a unified platform that harnesses AI agents to manage and defend infrastructure across Cisco's product lines as well as other vendors' equipment. But Cloud Control's value compounds with the amount of Cisco gear a customer runs; how well it handles multivendor GPU fabrics, the situation most neoclouds actually face, is still an open question.
Arista, HPE, and Nvidia each ship their own orchestration and management tools as well, but like Cisco's, those are built around their own hardware. Netris positions itself above all of them — abstracting whatever mix of vendors' switches is running underneath and providing a single control plane regardless of the hardware stack. Its ecosystem includes Nvidia, Mirantis, Rafay, Red Hat, Spectro Cloud, vCluster and HPE.
The question for Netris is whether its head start can survive the inevitable push from incumbents with deeper pockets and broader customer relationships. With $15 million, 800% ARR growth, and a team of SDN veterans behind it, Netris is betting the answer is yes.