Getting Your Research Adopted
summary
Advice focuses on concrete tactics: embed yourself via internships or sabbaticals, publish code and data on GitHub with a usable README, and build real systems that run on real workloads rather than toy benchmarks (SPEC and GitHub apps were emphasized). Case studies: Automan (Scala DSL for crowdsourced correctness) fails adoption; Stabilizer exposes that -O2 vs -O3 differences are often indistinguishable; Horde allocator succeeds and gets cited in Apple code. Also recommends go to the mountain — give industry talks and engage users.
AI-written summary · about
chapters (16)
- 0:00Start
- 2:08Step One Embed Yourself
- 5:11Building Real Systems
- 6:02Stop Playing with Toys
- 6:24Spec Benchmark Suite
- 12:04Your Talk Should Tell a Story
- 13:56Engage with the Community
- 18:13Case Studies
- 18:32Automan
- 22:57Memory Layout Drastically Affects Performance
- 23:47Statistical Hypothesis Testing
- 30:35Patent Infringement
- 33:52Research Adoption within the Academic Community
- 36:55Graduate Students
- 40:02Do You Always Need To Build Real Systems
- 53:45Code Names
