T
Tatiana Shpeisman
Researcher at Intel
Publications - 86
Citations - 3152
Tatiana Shpeisman is an academic researcher from Intel. The author has contributed to research in topics: Transactional memory & Compiler. The author has an hindex of 31, co-authored 84 publications receiving 2998 citations. Previous affiliations of Tatiana Shpeisman include University of Maryland, College Park & PARC.
Papers
More filters
Proceedings ArticleDOI
Affinity-aware work-stealing for integrated CPU-GPU processors
TL;DR: A preliminary implementation of the work-stealing scheduler, Libra, is described, which includes techniques to deal with architectural differences in integrated CPU-GPU processors, and Libra's affinity-aware techniques achieve significant performance gains over classically-implemented work-Stealing.
Posted Content
HPAT: High Performance Analytics with Scripting Ease-of-Use
TL;DR: HPAT as mentioned in this paper is an auto-parallelizing compiler approach that exploits the characteristics of the data analytics domain such as the map/reduce parallel pattern and is robust, unlike previous autoparallelization methods.
Patent
Transactional memory management techniques
TL;DR: In this paper, techniques for improved transactional memory management are described, where a processor element, an execution component for execution by the processor element to concurrently execute a software transaction and a hardware transaction according to a transactional processing process, and a finalization component to abort the hardware transaction when the global lock is active when execution of the software transaction completes.
Patent
Forward-looking machine learning for decision systems
TL;DR: In this paper, a machine learning decision system includes an online decision system and an offline decision system, which produces a first-time slice-specific decision output corresponding to a first time slice based on one or more situational inputs received in the first- time slice.
Patent
Method for providing garbage collection support
Tatiana Shpeisman,Guei-Yuan Lueh +1 more
TL;DR: In this paper, live reference data to support garbage collection is stored in null operation instructions (NOPs) of an instruction set within the native code, which can be retrieved in constant time.