Every Chip Counts: Decentralized Cloud Rebellion
summary
Cloud computing is monopolized with 67% market share for the top three providers and 75% for the top six and hundreds of subprocessors listed with access to customer data, while a CrowdStrike misconfiguration recently impacts 8.5 million devices. The proposed alternative assembles a global compute mesh from mobile upcycling, leverages ARM TPUs and secure chips, runs workloads inside a trusted execution environment, separates permissionless consensus from execution, and uses reputation scoring for verifiable, confidential, decentralized compute.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 1:55Monopolization risks
- 3:41Cloud providers as geopolitical actors
- 6:29Centralization consequences
- 9:42Data ownership and security risks
- 10:47Why platforms degrade
- 12:13Need for open-source cloud
- 14:00Mobile devices as data centers
- 16:51Verifiable compute architecture
- 18:41Conclusion
related talks
40mCybersecurity for Future Extreme Computing
14mAI and Open Compute - How to Solve the World's Big Problems
43mClouds in the forecast: CrowdStrike cloud risk report
13mNext-Gen Cloud: Enabling Secure and Seamless Cross
26mWrap-Up Session Panel Discussion: Quick Takes
41mCloud Computing: The Edge talks - long road into the Cloud
23mIntegrating Confidential Computing Into Cloud Infrastructure: Challenges and Opportunities
23mCloud Security Megatrends - No Data-Driven Innovation without Security