Keynote |The Microarchitectures That I Saw And The Ones That I Hope To One Day See
summary
Firsthand account of microarchitecture security shows concrete failures and fixes: proof‑of‑concept exploits and microcode reverse engineering expose ASLR‑cache bypass and TSX-based KASLR bypasses, Meltdown class faults, and a CSME code‑execution chain; running POCs across cloud providers reveals that every major cloud had IBPB/IBRS mitigation failures or kernel misconfigurations; an OEM mistakenly published a patch one week early, legal class actions follow, AMD/Intel responses include a six‑VP crisis division with no engineers, and many mitigations are quietly partial or broken.
AI-written summary · about
chapters (19)
- 0:00Introduction and background
- 1:11Disclaimers and perspectives
- 3:35My failures and research focus
- 6:02Minecraft and side channel challenges
- 8:28Security challenges: assumptions and composition
- 10:58System design and survivability
- 15:51Side channel timeline and early mistakes
- 20:05KPTI bypass and Intel’s response
- 25:18Meltdown and Spectre discovery
- 28:46Public disclosure and legal fallout
- 31:56Team dynamics and Intel’s changes
- 33:00CSME vulnerabilities and vendor responses
- 36:28AMD’s marketing and Intel’s investigations
- 39:13Vendor responses and research frustrations
- 42:03Simple bugs and overlooked mitigations
- 45:32Mitigation failures and testing gaps
- 48:19Industry collaboration and failures
- 50:25Cloud providers and kernel issues
- 51:05Future of computation and Q&A
related talks
25mThe Memory of a Meltdown
41mUsing Microarchitectural Design to Break KASLR and More
01h 01mTutorial: Side-Channel Attacks and Transient Execution Vulnerab...
24mThe Memory of a Meltdown
42mAre Microarchitectural Attacks still possible on ...
51mExploiting Hardware Glitches And Side Channels In Perfect Software
19mKeynote: MDS, Fallout, Zombieland & Linux
46mRethinking Architectures and Abstraction for a World Where Security Improvements Matter More than Performance Gains