Sponsored Session: Confidential Computing - New Capabilities for New Workloads
summary
Hardware-based Trusted Execution Environments isolate code and data from the host and other tenants, and remote attestation returns cryptographic measurements (hashes signed by chip keys) to prove which image and microcode are running; that enables provenance for Gen AI training and inference, multi-party reconciliation without mutual trust and Web3 nodes that prove they run unmodified code. GPUs and CPUs (Intel SGX/TDX, AMD SEV SNP, Arm CCA, NVIDIA) are supported, performance hits are often ~2–5% (pathological cases up to ~25%).
AI-written summary · about
chapters (18)
- 0:00Confidential Computing intro
- 1:05Privacy tech family
- 2:09Definition and components
- 4:14Limitations of virtualization
- 5:41Trusted execution environments
- 7:10Trusting cloud providers
- 8:13Remote attestation
- 13:22Cloud availability
- 14:46Primitives for new use cases
- 19:43Security assurances
- 21:27Industry use cases
- 22:34Generative AI and CC
- 26:07Interbank reconciliation
- 28:16Pharma research privacy
- 29:22UN policy collaboration
- 30:26Web 3 and decentralized
- 31:55Future trends
- 34:01Q&A performance
related talks
42mSponsored Session: Confidential Computing - New Capabilities for New Workloads
37mIt's All About Trust: Why Confidential Computing Matters
36mCTO Panel: The Impact of Confidential Computing with Intel, Azure, AMD & NVIDIA- 37mIt's All About Trust: Why Confidential Computing Matters
36mSecurity for AI and Multi-Party Collaboration with Confidential Computing (Oh, and...
11mLightning Talk: Computing Confidentially in the Clouds
16mIntroduction to confidential computing
30mK8s meets confidential computing - the different ways of scaling sensitive workloads