Fortinet partners with Arista Networks on Secure AI data center solution

Fortinet and Arista Networks announced a Secure Artificial Intelligence (AI) Data Center solution that was deployed at Monolithic Power Systems and provided a validated, scalable Zero Trust Architecture (ZTA) for AI data centers.

The collaboration expanded Fortinet’s Secure AI Data Center framework through multivendor integration and described a reference architecture intended to address infrastructure cost, vendor lock-in, performance bottlenecks, and security risks, noting that architectural complexity and skill gaps contributed to 95% of AI deployment failures.

The joint design combined Arista ultra-low-latency networking with Fortinet capabilities, including ASIC-accelerated firewalls, AI-aware threat protection, zero-trust segmentation, Encrypted Traffic Inspection (ETI), and automated response; the release also described offloading HTTPS/TLS to an Application-Specific Integrated Circuit (ASIC) to increase throughput and reduce latency and jitter.

As part of the effort, the partners co-developed and validated the reference architecture with Monolithic Power Systems; the release said Model Partitioning Strategy (MPS) leveraged the Fortinet security ecosystem to design and operate AI data center infrastructure aligned with deployment and operational requirements.

“AI infrastructure is advancing faster than most organizations can secure or operationalize it,” said Nirav Shah, Senior Vice President of Products and Solutions at Fortinet. “Arista Networks is a leader in data-driven AI networking solutions. Our collaboration with Fortinet brings together best-in-class cybersecurity with Arista’s ultra-low-latency, high-speed networking, which is essential for AI applications,” said Ed Chapman, Vice President of Business Development at Arista Networks.

The companies described plans to support next-generation AI accelerators, enable deployments up to 80% faster via Zero-Touch Provisioning (ZTP), and use ASIC-based Transport Layer Security (TLS) offload to increase HTTPS/TLS performance up to 33x.