en.Wedoany.com Reported - On June 1, Israeli large-scale network solutions company DriveNets announced the completion of $410 million in Series D funding, bringing its total cumulative funding to $1 billion. The round was led by Bessemer Venture Partners and Atreides Management, with new investors AMD and Red Dot Capital joining, and existing investors such as Pitango and D1 Capital Partners continuing to participate. The funds will be used to expand Ethernet Fabric capabilities for large-scale AI deployments.

The direct backdrop for this funding is that AI infrastructure is shifting from single-vendor closed systems to more open, multi-vendor heterogeneous computing clusters. DriveNets stated that it has over $1 billion in confirmed business and has maintained positive cash flow since 2025. The new funds will primarily be used to expand inventory to support the growing pipeline of AI Fabric projects and to extend heterogeneous AI infrastructure solutions. Its Ethernet AI Fabric is built on DriveNets' existing Network Cloud engineering foundation, targeting data center clusters for foundational model labs, hyperscale cloud providers, NeoCloud providers, and large enterprises. The goal is to improve GPU utilization, shorten cluster deployment cycles, and reduce the operating cost per AI workload under a standard Ethernet framework. As AI training and inference scales continue to grow, network bottlenecks have become a critical factor affecting GPU cluster efficiency, especially in clusters composed of different accelerators, server vendors, and storage resources. Network scheduling capabilities directly impact the number of tokens generated per second, cost per token, and overall power utilization efficiency.

This funding round also brought in AMD as a new investor, indicating that the AI computing supply chain is forming new collaborative relationships around open infrastructure. The company stated that it is currently collaborating with AI suppliers such as AMD and Broadcom to strengthen integration between networking and computing, while also working with system partners like Dell and Supermicro to advance market deployment.

In terms of technical approach, DriveNets' AI Fabric solution is based on standard Ethernet, supporting horizontal scaling, vertical scaling, and cross-cluster expansion architectures, covering both the front-end network and storage connections of AI clusters. For large AI data centers, simply increasing the number of GPUs is no longer sufficient to solve all performance issues. Idle GPUs in clusters, network congestion, reliability fluctuations, and extended deployment cycles all translate into high capital expenditures and operating costs. DriveNets aims to leverage full-stack network optimization to enable different AI accelerators to collaborate within the same cluster based on training, inference, or specific task phases, thereby improving overall cluster utilization. The company also views heterogeneous AI architecture as a key direction for future AI infrastructure, meaning using AI accelerators from different vendors tailored to various task phases within the same cluster, and optimizing performance and energy consumption through coordinated networking, computing, and software.

DriveNets previously primarily provided network cloud solutions to large telecom operators, with its production networks serving tier-1 operators such as AT&T and Comcast. This Series D round further shifts the company's business focus toward the AI data center networking market, reflecting that AI infrastructure investment is extending from chips, servers, and data center facilities to the network layer that supports large-scale cluster efficiency. As enterprises and cloud providers accelerate the construction of multi-vendor AI clusters, open Ethernet Fabric, GPU utilization optimization, and heterogeneous computing scheduling are becoming key battlegrounds in AI infrastructure competition.

For the AI industry chain, the significance of DriveNets securing $410 million in funding goes beyond continued capital flow into AI infrastructure companies. It underscores that network architecture is becoming a core variable in AI economics. The cost pressures of large-scale AI systems are forcing data center operators to reassess underlying network designs. Those who can establish stable and efficient connectivity among open hardware, multi-vendor accelerators, and high-density GPU clusters are more likely to secure a key position in the next phase of AI infrastructure construction.

This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com