To address rising hardware security risks and the standardization of post-quantum cryptography (PQC), Jmem Tek introduces “ArgusNPU PQC-based Secure Edge AI Processor,” the world’s first edge AI inference platform with an integrated chip-level PQC security module. The system supports the following PQC algorithms: Dilithium, Kyber, and XMSS, with backward compatibility for EdDSA and ECDH to meet international cryptography standards. Built-in neural and vector engines accelerate AI inference. AES hardware encryption secures runtime data and resists side-channel attacks.

Key Features:
1. Sensor Data Encryption: Real-time encryption of input data to prevent interception.
2. OTA with PQC: On-chip PQC accelerators secure remote model updates against tampering and malware injection.
3. On-Chip Data Security: Encrypted DRAM storage prevents reverse engineering and data leaks, addressing security blind spots in traditional chips.
4. Encrypted AI Detection: Enables secure on-device detection of people, vehicles, and objects.

Application:
Defense UAVs, smart surveillance, V2X systems, industrial control, and AIoT. Jmem Tek is currently collaborating with domestic and international drone manufacturers for deployment and validation