Fortinet teams with NVIDIA to speed up GPU-powered AI security at scale

Fortinet said it has partnered with NVIDIA to accelerate Fortinet’s FortiAIGate offering, aiming to bring faster AI-driven security capabilities to environments that rely on high-performance GPU compute. FortiAIGate is Fortinet’s AI-focused security workflow, designed to help operationalize threat analysis and security tasks more quickly than conventional, rules-driven approaches, according to the details available in the announcement.

The collaboration is positioned around using NVIDIA’s GPU platform to increase the throughput of AI workloads that run inside security operations workflows. For security vendors, the bottleneck is often not just model accuracy, but whether AI reasoning can keep up with the volume and velocity of telemetry that organizations ingest across networks, endpoints, and cloud systems.

Fortinet did not provide specific engineering details in the announcement about which NVIDIA software stack components are involved (such as particular inference runtimes or model optimization frameworks). Nor did the company specify whether deployments would be targeted to Fortinet appliances, third-party servers, or cloud environments. That means customers will need to rely on Fortinet and NVIDIA documentation to understand where FortiAIGate will run and what hardware configurations are required.

The announcement also did not include quantitative claims such as latency improvements, throughput benchmarks, or the proportion of security processes that FortiAIGate can automate. In the absence of those figures, the most defensible takeaway is that the partnership is intended to reduce compute friction for AI security workloads by aligning Fortinet’s platform with NVIDIA’s data center GPU ecosystem.

Fortinet is not alone in pushing AI into security products, but execution tends to hinge on performance. Organizations increasingly expect security tools to flag emerging threats, prioritize alerts, and assist with incident response in near real time. GPU acceleration can help vendors run more capable AI models or run the same models more efficiently, which matters when security teams are dealing with large volumes of logs and security events.

NVIDIA, for its part, has spent years building an end-to-end AI platform for training and inference in data centers. Its involvement in security-focused product partnerships reflects a broader pattern in which infrastructure providers try to make GPU compute a default option for AI applications across industries, including enterprise cybersecurity.

It is also unclear from the initial announcement whether the partnership changes Fortinet’s go-to-market directly, such as new licensing terms, bundled hardware, or pricing tiers for GPU-enabled deployments. Fortinet did not disclose commercial terms in the announcement available here, so any timing of customer rollouts and the commercial impact will require additional confirmation.

For readers tracking the development, the next useful datapoints would be: any further detail on the deployment footprint for FortiAIGate, any released performance benchmarks tied to specific NVIDIA hardware, and any new Fortinet partner or reseller commitments that support customer migrations. Until those are clarified, the partnership should be viewed as a compute-acceleration announcement rather than a fully specified product change.