000221890 001__ 221890
000221890 005__ 20180913063917.0
000221890 0247_ $$2doi$$a10.1145/2872362.2872385
000221890 022__ $$a0362-1340
000221890 02470 $$2ISI$$a000379415100055
000221890 02470 $$2ISI$$a000385493900055
000221890 037__ $$aCONF
000221890 245__ $$aProteusTM: Abstraction Meets Performance in Transactional Memory
000221890 260__ $$aNew York$$bAssoc Computing Machinery$$c2016
000221890 269__ $$a2016
000221890 300__ $$a15
000221890 336__ $$aConference Papers
000221890 520__ $$aThe Transactional Memory (TM) paradigm promises to greatly simplify the development of concurrent applications. This led, over the years, to the creation of a plethora of TM implementations delivering wide ranges of performance across workloads. Yet, no universal implementation fits each and every workload. In fact, the best TM in a given workload can reveal to be disastrous for another one. This forces developers to face the complex task of tuning TM implementations, which significantly hampers their wide adoption. In this paper, we address the challenge of automatically identifying the best TM implementation for a given workload. Our proposed system, ProteusTM, hides behind the TM interface a large library of implementations. Underneath, it leverages a novel multi-dimensional online optimization scheme, combining two popular learning techniques: Collaborative Filtering and Bayesian Optimization. We integrated ProteusTM in GCC and demonstrate its ability to switch between TMs and adapt several configuration parameters (e.g., number of threads). We extensively evaluated ProteusTM, obtaining average performance < 3 % from optimal, and gains up to 100x over static alternatives.
000221890 6531_ $$aTransactional Memory
000221890 6531_ $$aRecommender Systems
000221890 6531_ $$aPerformance Tuning
000221890 6531_ $$aAdaptive System
000221890 700__ $$aDidona, Diego$$uUniv Lisbon, INESC ID, Inst Super Tecn, Lisbon, Portugal
000221890 700__ $$aDiegues, Nuno$$uUniv Lisbon, INESC ID, Inst Super Tecn, Lisbon, Portugal
000221890 700__ $$aKermarrec, Anne-Marie$$uINRIA, Rocquencourt, France
000221890 700__ $$0240335$$aGuerraoui, Rachid$$g105326
000221890 7112_ $$a21st International Conference on Architectural Support for Programming Languages and Operating Systems$$cAtlanta, GA$$dAPR 02-06, 2016
000221890 773__ $$j51$$k4$$q757-771$$tAcm Sigplan Notices
000221890 909C0 $$0252114$$pDCL$$xU10407
000221890 909CO $$ooai:infoscience.tind.io:221890$$pconf$$pIC
000221890 917Z8 $$x105326
000221890 937__ $$aEPFL-CONF-221890
000221890 973__ $$aEPFL$$rREVIEWED$$sPUBLISHED
000221890 980__ $$aCONF