DriveNets Raises $410 Million in Series D Funding to Expand AI Ethernet Fabric Solutions

DriveNets Secures $410 Million in Series D Funding

In a significant move to bolster its capabilities, DriveNets, a leading provider of large-scale networking solutions, has completed a Series D funding round raising an impressive $410 million. This funding aims to enhance operations as the demand for Ethernet Fabric in the field of artificial intelligence (AI) continues to surge. The company has successfully amassed a total of $1 billion in capital, showcasing its robust market presence.

With a substantial order backlog exceeding $1 billion and a positive cash flow established since 2025, DriveNets plans to utilize this latest infusion of funds to scale its inventory and support its growing pipeline of AI solutions. This financing round was led by prominent investors Bessemer Venture Partners and Atreides Management, joined by existing backers Pitango and D1 Capital Partners, along with new investors AMD and Red Dot Capital.

Founded a decade ago, DriveNets has become an integral player in the networking arena, serving some of the largest telecommunications companies worldwide. DriveNets' Ethernet-based AI infrastructure is designed to meet the needs of large-scale AI developments being spearheaded by research laboratories, hyperscalers, and large enterprises.

The company is collaborating with esteemed AI providers such as AMD and Broadcom to enhance integration between network and compute components within multi-vendor AI environments, optimizing cluster performance and maximizing GPU utilization. This collaborative effort is poised to significantly contribute to improving token economics essential to AI applications.

Ido Susan, co-founder and CEO of DriveNets, commented on the funding, stating, "This financing marks a pivotal point in our journey to meet the escalating needs for large-scale AI infrastructure. Currently, one of the most valuable untapped resources globally is a GPU that is waiting on the network. We leverage our extensive expertise in high-performance networking to enable clients to achieve higher utilization rates, reduce costs per workload, and scale their AI operations efficiently across any AI accelerator they choose."

Vamsi Boppana, Senior Vice President for AI at AMD, highlighted the implications of DriveNets' advancements by stating that the AI infrastructure is entering a new era of open, integrated systems where computing power, network technology, and software must mutually scale. Their partnership facilitates efficiency in handling AI workloads using AMD Instinct accelerators and DriveNets' powerful fabric within an open infrastructure framework to foster next-generation standard-based AI data centers.

Addressing Core Challenges in AI Infrastructure

One of the most pressing issues facing AI infrastructures today is the challenge of network bottlenecks and inefficiencies, particularly concerning large GPU clusters that struggle to operate at peak potential due to these constraints. DriveNets' solutions address these challenges, offering a network fabric that alleviates such bottlenecks through end-to-end optimization across the AI stack.

This includes enhancements through shared communication libraries, transport protocols, network cards, and orchestration at the system level, developed in conjunction with leading AI accelerator manufacturers. A recently validated AMD-DriveNets reference architecture is a prime example of how system optimizations can lead to maximized GPU utilization, decreased costs per token, and swift deployment facilitates scaled environments.

Charlie Kawwas, Chair of the Semiconductor Solutions Group at Broadcom, remarked on the significance of robust networking architecture in AI, stating, "With AI systems reaching unprecedented scale, the performance of underlying network structures has become crucial to their economic viability. The collaboration between Broadcom's semiconductor and Ethernet switching solutions and DriveNets' fabric brings the scalability and efficiency needed for modern AI workloads."

Emerging Trends in AI Networking

Market projections suggest that the AI networking sector could surpass $200 billion by the end of the decade, driven largely by a shift from single-vendor stacks to multi-vendor heterogeneous AI infrastructures. DriveNets is strategically positioned to lead in this transition owing to its formidable combination of reliability and validated designs alongside an adequate inventory to serve a market undergoing supply chain challenges.

Ido Susan emphasizes, "As AI infrastructure expenditures shift from training to inference, this transition paves the way for heterogeneous architectures that optimize infrastructure costs and enhance energy efficiency. Our solutions are precisely tailored to support these environments, granting clients the flexibility needed to thrive amid evolving technological landscapes."

DriveNets remains committed to powering the next wave of AI innovation by fundamentally altering the economic landscape of large-scale infrastructures while maximizing operational performance and efficiency. For additional information about DriveNets and their offerings, visit DriveNets.