With Intel’s NPU-enabled processor, Acronis integrates its cybersecurity platform, enabling local AI-based threat detection on enterprise devices.
Acronis has integrated its Cyber Protect Cloud platform with Intel’s Core Ultra processors, enabling endpoint devices to run AI-based threat detection locally using the chip’s Neural Processing Unit (NPU).
The shift targets performance bottlenecks in traditional endpoint security architectures, where detection tasks are typically CPU-intensive and reliant on cloud connectivity. By moving behavioural analysis and anomaly scoring to the NPU, the integration reduces strain on central processing and network infrastructure.
Internal testing by Acronis indicates that offloading these workloads results in up to 92% lower CPU usage. For enterprise IT teams and managed service providers, this means security processes can run in real time without degrading device performance—particularly important in remote or hybrid deployments with constrained resources.
“This allows managed service providers and enterprise IT teams to apply real-time detection without degrading system performance,” said Gaidar Magdanurov, President at Acronis.
Intel’s OpenVINO toolkit is used to deploy machine learning models directly onto the NPU. By keeping threat detection local, the system reduces dependence on cloud-based inference engines and limits data exposure over networks.
“By running complex security tasks on the AI PC itself, we reduce the need to send data off the device,” said Carla Rodríguez, Vice President and General Manager of Client Software Enabling at Intel. “That has implications not just for performance, but also for privacy.”
Cyber Protect Cloud remains a multi-tenant SaaS platform offering integrated detection and response (EDR/XDR), data loss prevention, and backup. With security functions distributed to endpoint hardware, the integration aligns with the NIST framework’s detection, response, and recovery principles—executed closer to where attacks occur.
