[Colloq] PhD Thesis Defense - Felix Klock, Monday, December 20

[Rachel Kalweit]

The College of Computer and Information Science presents:

PhD Thesis Defense by: Felix Klock

When: Monday, December 20, 2010
Time: 10:00am
Where: 366 West Village H

Thesis title: Scalable Garbage Collection via Remembered Set Summarization and Refinement

Abstract:

Regional garbage collection offers a useful compromise between real-time and generational collection. Regional collectors resemble generational collectors, but are scalable. A scalable collector guarantees a positive lower bound, independent of mutator and live storage, for the theoretical worst-case minimum mutator utilization (MMU).

Standard generational collectors are not scalable. Some real-time collectors are scalable, while others assume a well-behaved mutator or provide no worst-case guarantees at all.

This dissertation presents the design of a regional collector, coupled with a theorem that the collector is scalable in the sense above. The theorem also establishes upper bounds for worst-case space usage and collection pauses.  Regional collectors cannot compete with hard real-time collectors at millisecond resolutions, but offer efficiency comparable to contemporary generational collectors combined with improved latency and MMU at resolutions on the order of hundreds of milliseconds to a few seconds.

A prototype of this regional collector shows that it performs acceptably on a wide range of benchmarks.

Thesis committee:

Will Clinger (advisor)
Gene Cooperman
Olin Shivers
Guy L Steele Jr (Oracle, external member)