Jens, the Ultimate PL Renaissance man!
summary
A concise history of concrete results: a formal featherweight X10 language models structured parallelism with async and finish and proves a progress theorem that rules out deadlock for all programs in the model; a coordination language, Concurrent Collections, yields a proved determinism theorem used to schedule CPU/GPU/FPGA kernels and report execution and energy improvements; classroom use shows the formal small‑step semantics produces multiple interleavings but often deterministic final states.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 1:28IBM and Purdue collaborations
- 2:55Featherweight X10 and structured parallelism
- 6:30Deadlock avoidance in X10
- 9:20Type systems for parallel programs
- 11:28Verification levels in parallel programming
- 13:14Featherweight X10 pedagogy
- 13:34NSF expedition and concurrent collections
- 15:20Determinism in concurrent collections
- 16:47Yens as mentor
related talks
52mThink Abstractly, Work Concretely
01h 01mSpearheading the future of programming
01h 09mSCCharts Twelve Years Later: A Reflection on Sequential Constructiveness and(…)
01h 02mProcedural Programming: It Never Went Away
01h 03mPrinciples of Persistent Programming
32mDesign of Software Representation Languages: a Historical Perspective
26mInfinite Possibilities
28mWhat You See Is What It Does: A Structural Pattern for Legible Software