99.83% TAR
NIST FRTE 1:1 evaluation at a FAR of 1 in 1 million

Optimized for Intel® Core™ Ultra processors, FaceMe uses integrated graphics (iGPU) and OpenVINO™ toolkit to accelerate facial recognition without a discrete GPU. Run recognition locally on compact edge devices with simpler hardware configurations and a smaller footprint.
FaceMe combines NIST-evaluated recognition accuracy with Intel iGPU and OpenVINO™ optimization for faster processing and flexible deployment.
99.83% TAR
NIST FRTE 1:1 evaluation at a FAR of 1 in 1 million

Intel iGPU Acceleration
More Processing Power. Same Hardware.
Core Ultra X7 358H: iGPU vs. CPU-Only

Edge AI
No discrete GPU required. Less hardware to deploy.

FaceMe leverages the integrated graphics (iGPU) in Intel® Core™ Ultra processors and the OpenVINO™ toolkit to accelerate facial recognition at the edge. By eliminating the need for a discrete GPU, organizations can simplify hardware configurations, reduce space requirements and help lower overall deployment costs.

FaceMe uses Intel integrated graphics (iGPU) and OpenVINO™ to leverage the processor's built-in computing resources and accelerate facial recognition.
Deploy in space-constrained edge environments, with the flexibility to choose hardware based on application requirements, people traffic and camera count.
Powered by Intel platforms, FaceMe Security combines facial recognition, video surveillance and access control with employee and visitor identity verification for office buildings, factories and retail stores.
FaceMe SDK leverages Intel integrated graphics (iGPU) and the OpenVINO™ toolkit to accelerate facial recognition on Intel® Core™ Ultra Series 3 platforms. The benchmarks below demonstrate the performance gains from iGPU acceleration on the same processor, as well as the combined benefits of a processor upgrade and iGPU acceleration.
Same Processor: iGPU vs. CPU-Only
FaceMe SDK · Intel Core Ultra X7 358H
20×
Facial Recognition Speed
Frames Processed per Second (FPS)
11.1 FPS
217.9FPS
CPU-only modeCore Ultra X7 358H
Intel iGPU modeCore Ultra X7 358H
Processing 1080p video with the Ultra High Accuracy (UH10) model, FaceMe SDK achieved 217.9 frames per second (FPS) using iGPU acceleration on the Intel® Core™ Ultra X7 358H, compared with 11.1 FPS in CPU-only mode on the same processor. That's approximately 20× the processing speed.
Process more video frames in the same amount of time by using the processor’s integrated graphics, without adding a discrete GPU.
CyberLink Internal Test Configuration (September 24, 2026):
FaceMe SDK combines Intel® Core™ Ultra Series 3 processors with iGPU acceleration to handle larger facial recognition workloads on compact edge hardware. Using the UH10 model to process 1080p video, the Intel® Core™ Ultra X7 358H reached 217.9 FPS in iGPU mode—about 37× the processing speed of the Intel® Core™ i5-1235U in CPU-only mode (5.9 FPS).
Process more video frames in the same amount of time without adding a discrete graphics card, giving applications more compute capacity to grow while keeping hardware requirements streamlined.
CyberLink Internal Test Configuration (September 24, 2026):
Platform Upgrade + iGPU Acceleration
FaceMe SDK
37×
Facial Recognition Speed
Frames Processed per Second (FPS)
5.9 FPS
217.9FPS
12th Gen Intel Corei5-1235U, CPU mode
Intel Core Ultra Series 3X7 358H, iGPU mode
FaceMe Security is an all-in-one solution for facial recognition, video surveillance and access control, with employee and visitor identity verification. It helps system integrators and enterprises plan multi-camera security and access control deployments in office buildings, factories and retail stores.

Built on the Intel Core Ultra-optimized FaceMe SDK, FaceMe Security runs facial recognition, blocklist alerts and access control on a single mini PC. Intel Arc B390 integrated graphics deliver accelerated performance without a discrete GPU, helping reduce hardware requirements and simplify deployment.
Up to 59 Cameras
Blocklist Alert Scenario
Up to 49 Cameras
Access Control Scenario
CyberLink Internal Test Configuration (September 24, 2026):
Run FaceMe facial recognition locally on Intel® Core™ Ultra platforms, keeping organizations in control of their data. Local processing supports data protection and flexible deployment across security, access control and AIoT applications.

Video Surveillance and Security
FaceMe Security combines multi-camera video with facial recognition to identify individuals on a blocklist and trigger real-time alerts, helping strengthen security at entrances, exits and critical areas.

Smart Offices and Workforce Management
Verify employee and visitor identities with facial recognition, and integrate access control with attendance management to streamline entry and exit at office buildings and corporate campuses.

Retail and Hospitality
Integrate FaceMe SDK into point-of-sale (POS) systems, self-service kiosks and check-in devices to recognize members and VIPs, enabling more convenient and personalized customer experiences.
FaceMe® SDK
Integrate FaceMe SDK's facial recognition engine and development tools into your applications and smart devices. Leverage Intel iGPU acceleration to develop solutions tailored to your requirements.
Explore FaceMe SDK >
FaceMe® Security
Combine facial recognition, video surveillance, access control and blocklist alerts in one solution. Verify employee and visitor identities while simplifying security deployments for enterprises and system integrators.
Explore FaceMe Security >FaceMe SDK supports Intel iGPU acceleration on Windows and Ubuntu x64. Check the FaceMe SDK system requirements for supported operating system versions and hardware.
Yes. FaceMe Security Central, Workstation and the database can run on a single mini PC powered by an Intel® Core™ Ultra processor, such as the ASUS NUC 16 Pro. The right configuration depends on camera count, people traffic and enabled features. Contact the FaceMe team for guidance.
FaceMe facial recognition processing speed and camera capacity depend on the use case, Intel processor model, system configuration, video resolution, frame rate, people traffic per camera and enabled features. The performance figures shown are from CyberLink internal tests on specified configurations. Contact the FaceMe team for hardware recommendations based on your application and deployment requirements.