CNCC2024 Confidential Computing Forum – Schedule, Speakers, and Research Abstracts

Forum Overview: The second Confidential Computing Forum, titled “How Will Confidential Computing Evolve?”, takes place on October 26, 13:30‑17:30 at Summer Garden – French Pavilion, Hall 2. It aims to explore frontier TEE technologies, analyze industry trends, and promote autonomous, secure computing solutions for China’s data infrastructure.

Context: In the digital economy, data is a strategic resource; confidential computing, especially based on Trusted Execution Environments (TEE), has emerged as a key method to ensure data security. Major CPU vendors now provide hardware‑supported TEE, laying the foundation for further development.

Key Challenges: Enhancing TEE security, detecting and repairing TEE‑targeted attacks, and designing robust cryptographic algorithms for TEE are critical research focuses.

Forum Highlights: The agenda includes a keynote speech, six technical talks, and a panel discussion.

Schedule and Speakers:

1. Opening Remarks – Feng Dengguo (Chinese Academy of Sciences, CCF Fellow)

2. "Security Analysis of Confidential Computing from the NbSP Paradigm" – Wei Tao (Ant Group)

3. "From Application Platform to Trusted Cloud Foundations: The Jeddak Data Security Sandbox" – Zhang Yao (ByteDance)

4. "Ciphertext Side‑Channel Vulnerabilities: Exploitation, Detection, and Repair" – Zhang Yinqian (Southern University of Science and Technology)

5. "Collaborative Trust Technology for Confidential Computing based on TEE and TPM" – Qin Yu (Institute of Software, Chinese Academy of Sciences)

6. "Exploring Cryptographic Applications for Trusted Execution Environments" – Zhu Hui (Xidian University)

Panel – moderated by Zhang Fengwei with participants from academia and industry.

Technical Abstracts:

Wei Tao: Introduces the NbSP zero‑over‑paradigm for holistic security measurement of confidential computing, identifying current weaknesses and future improvement directions.

Zhang Yao: Presents the Jeddak Sandbox, a cloud‑native confidential container platform supporting AI large‑model workloads, detailing transparent verification, data controllability, and heterogeneous trusted computing.

Zhang Yinqian: Discusses ciphertext side‑channel attacks on deterministic memory‑encryption in TEEs (e.g., AMD SEV‑SNP), covering discovery, exploitation, automated detection, and LLVM‑based remediation.

Qin Yu: Explores multi‑root trust architectures (CPU‑TEE, heterogeneous TEE, TPM) and software‑hardware co‑trust mechanisms for secure data sharing in confidential computing platforms.

Zhu Hui: Proposes cryptographic schemes for multi‑party collaboration under TEE, including homomorphic encryption, federated learning, and neural‑network inference, with experimental validation.

About CNCC2024: The conference runs from October 24‑26 in Hengdian, Zhejiang, featuring 18 invited talks, 3 plenary forums, 138 specialized forums, and over 100 exhibitions. More than 800 speakers, including Turing Award winners and academicians, are expected to attract over ten thousand participants.

For full details, visit the official CNCC2024 website.