NVIDIA Confidential Computing is a technology designed to secure data and applications while they are being processed, using hardware-based trusted execution environments (TEEs). This approach ensures that sensitive data, such as personally identifiable information (PII) and intellectual property (IP), remains protected from unauthorised access and tampering during AI training and inference. NVIDIA's H100 GPU is the first GPU to support confidential computing, providing a secure environment through a combination of hardware, firmware, and software. The H100 GPU includes an on-die hardware root of trust (RoT) and supports features like secure and measured boot, secure protocols for connecting to CPU drivers, and the generation of cryptographically signed attestation reports.