Akamai integrates Guardicore security into NVIDIA's AI data center hardware

This move cements Akamai’s role as an infrastructure provider for the AI era, moving beyond simple content delivery. By embedding security directly into NVIDIA’s AI silicon, Akamai solves the latency penalty typically associated with host-based security tools in high-performance computing. For the streaming and media ecosystem, this shift signals that the 'edge' is no longer just for caching video—it is becoming a secure processing layer for agentic AI architectures. Watch for Akamai's non-CDN revenue growth and margin expansion in 2027 as these deep silicon integrations move from pilot to production across global data centers.

The collaboration follows several years of strategic repositioning for Akamai. Since acquiring Linode in 2022 for approximately $900 million, the company has aggressively expanded its compute capabilities to compete with hyperscalers. Per Akamai's 2025 year-end financial results reported in February 2026, its Security segment generated $2.24 billion in revenue, up 10% year-over-year, while the legacy Delivery segment fell 5% to $1.25 billion. Specifically, its Guardicore and API Security products saw 43% year-over-year growth in 2025, reaching $293 million in combined revenue.

NVIDIA’s Vera BlueField-4 STX, unveiled at GTC 2026 in March, is specialized for "agentic AI" tasks that require long-context reasoning. According to NVIDIA's June 2026 announcements, the architecture aims to eliminate GPU underutilization caused by traditional CPU-based storage paths. By routing data through a dedicated accelerated storage layer, the BlueField-4 STX can process tokens up to five times faster than existing models. Akamai’s role in this stack is to provide the security logic that regulates how these autonomous agents interact with context memory without introducing the 'speed bumps' that current software-based firewalls create in high-speed AI factories.