Key Takeaways
The most expensive idle asset in an AI data center is a GPU waiting on the network — and DriveNets is building the fabric to fix that.
DriveNets, an Israeli networking software firm, has raised $410 million in a Series D round led by Bessemer Venture Partners and Atreides Management, bringing its total capital raised to $1 billion. AMD and Red Dot Capital joined as new investors, alongside existing backers Pitango and D1 Capital Partners. The company, which has been cash-flow positive since 2025 and holds more than $1 billion in secured business, will use the proceeds to scale inventory for its AI fabric pipeline and expand its heterogeneous AI infrastructure solutions.
"As AI systems reach unprecedented scale, the performance of the underlying network fabric has become a primary driver of AI economics," said Charlie Kawwas, president of the semiconductor solutions group at Broadcom, a key partner. DriveNets' Ethernet-based AI fabric supports large-scale clusters built by foundation model labs, hyperscalers, and enterprises, and the company is working with AMD, Broadcom, and others to tighten integration between networking and compute in multi-vendor environments.
DriveNets' technology allows telecommunications operators and data centers to build and manage networks using standard, off-the-shelf hardware rather than costly proprietary systems — a model known as disaggregated networking. Its AI fabric supports scale-up, scale-out, and scale-across architectures, addressing two fundamental constraints: GPU clusters operating below peak efficiency due to network bottlenecks, and slow cluster bring-up time that leaves capital equipment idle. The company has published a validated reference architecture with AMD that it says maximizes GPU utilization and reduces cost-per-token.
The funding comes as AI infrastructure spending shifts from training to inference, a transition expected to drive adoption of heterogeneous architectures that use multiple AI accelerator vendors within the same cluster. DriveNets' fabric is designed to optimize performance across any combination of accelerators, a capability that positions it to capture share in a market that 650 Group analyst Alan Weckel projects will surpass $200 billion by the end of the decade.
"Every shift in compute produces a new networking giant. Cisco wired the internet. Arista wired the cloud. Nvidia wired single-vendor AI. DriveNets is wiring what comes next: heterogeneous AI," said Adam Fisher, a partner at Bessemer Venture Partners, which has backed the company since its Series A.
The broader AI infrastructure buildout is accelerating. Hyperscalers are on track to spend more than $700 billion on AI infrastructure in 2027, according to industry estimates. Dell Technologies, a DriveNets systems partner, reported AI-optimized server revenue surged 757 percent to $16.1 billion in its fiscal first quarter, with management raising its full-year AI server revenue target to $60 billion. The spending cascade — from chips to servers to networking fabric — creates a tailwind for companies like DriveNets that sit between the compute layer and the physical infrastructure.
DriveNets was founded in 2015 and counts Broadcom, Fujitsu, and Wipro among its partners, with Dell and Supermicro joining on go-to-market activities. The company's Network Cloud has become the network of record for some of the world's largest telecommunications operators, including AT&T and Comcast, providing a base of recurring revenue that supports its expansion into AI infrastructure.
The company did not disclose its valuation after the latest round.
This article is for informational purposes only and does not constitute investment advice.
