Renu Raman, Terizza | SC23
summary
Claims that large language models become model as unit of compute and may replace databases as systems of record, driving a shift to declarative, English-language programming and new performance tradeoffs. Argues hardware and stack must be refactored — GPUs, different memory types, optics and cooling — and previews a memory-centric operating system being developed by Terizza (still stealth, possible Q1 reveal). Discusses white‑box supply chains and specialty clouds as emerging infrastructure paths.
AI-written summary · about
chapters (16)
- 0:00Introduction to Theresa
- 0:42Theresa's origin and vision
- 3:10Major tech inflection points
- 4:11AI as the new abstraction layer
- 6:00Consumerization of AI interfaces
- 7:03New infrastructure categories emerging
- 8:28Specialty clouds and operating models
- 9:31Market evolution and Theresa's role
- 10:55Black swan event in AI
- 11:37Enterprise AI readiness
- 12:40Future of the tech stack
- 14:02Theresa's name and differentiation
- 15:04Hardware middleware app paradigm shift
- 16:28Theresa's stealth mode updates
- 16:49Systems of the future
- 17:52Systems thinking as the new trend
