FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3080451848> ?p ?o ?g. }

• W3080451848 abstract "Pollux improves scheduling performance in deep learning (DL) clusters by adaptively co-optimizing inter-dependent factors both at the per-job level and at the cluster-wide level. Most existing schedulers will assign each job a number of resources requested by the user, which can allow jobs to use those resources inefficiently. Some recent schedulers choose job resources for users, but do so without awareness of how DL training can be re-optimized to better utilize those resources. Pollux simultaneously considers both aspects. By observing each job during training, Pollux models how their goodput (system throughput combined with statistical efficiency) would change by adding or removing resources. Leveraging these models, Pollux dynamically (re-)assigns resources to maximize cluster-wide goodput, while continually optimizing each DL job to better utilize those resources. In experiments with real DL training jobs and with trace-driven simulations, Pollux reduces average job completion time by 25%-50% relative to state-of-the-art DL schedulers, even when all jobs are submitted with ideal resource and training configurations. Based on the observation that the statistical efficiency of DL training can change over time, we also show that Pollux can reduce the cost of training large models in cloud environments by 25%." @default.
• W3080451848 created "2020-09-01" @default.
• W3080451848 creator A5009547049 @default.
• W3080451848 creator A5012361506 @default.
• W3080451848 creator A5037557529 @default.
• W3080451848 creator A5048138372 @default.
• W3080451848 creator A5073310637 @default.
• W3080451848 creator A5073820927 @default.
• W3080451848 date "2020-08-27" @default.
• W3080451848 modified "2023-10-05" @default.
• W3080451848 title "Pollux: Co-adaptive Cluster Scheduling for Goodput-Optimized Deep Learning" @default.
• W3080451848 cites W157468466 @default.
• W3080451848 cites W1598866093 @default.
• W3080451848 cites W1969805974 @default.
• W3080451848 cites W1979567624 @default.
• W3080451848 cites W2005126631 @default.
• W3080451848 cites W2031489346 @default.
• W3080451848 cites W2106411961 @default.
• W3080451848 cites W2108598243 @default.
• W3080451848 cites W2131241448 @default.
• W3080451848 cites W2132737349 @default.
• W3080451848 cites W2146502635 @default.
• W3080451848 cites W2168231600 @default.
• W3080451848 cites W2182361439 @default.
• W3080451848 cites W2186615578 @default.
• W3080451848 cites W2193413348 @default.
• W3080451848 cites W2194775991 @default.
• W3080451848 cites W2219888463 @default.
• W3080451848 cites W2338908902 @default.
• W3080451848 cites W2339765813 @default.
• W3080451848 cites W2407746682 @default.
• W3080451848 cites W2427527485 @default.
• W3080451848 cites W2466406340 @default.
• W3080451848 cites W2480885269 @default.
• W3080451848 cites W2574140004 @default.
• W3080451848 cites W2605350416 @default.
• W3080451848 cites W2605643718 @default.
• W3080451848 cites W2606722458 @default.
• W3080451848 cites W2622263826 @default.
• W3080451848 cites W2732547613 @default.
• W3080451848 cites W2766164908 @default.
• W3080451848 cites W2773689216 @default.
• W3080451848 cites W2787998955 @default.
• W3080451848 cites W2798515322 @default.
• W3080451848 cites W2799042347 @default.
• W3080451848 cites W2899071864 @default.
• W3080451848 cites W2900167092 @default.
• W3080451848 cites W2902280036 @default.
• W3080451848 cites W2903697572 @default.
• W3080451848 cites W2908510526 @default.
• W3080451848 cites W2913545852 @default.
• W3080451848 cites W2914196460 @default.
• W3080451848 cites W2919594608 @default.
• W3080451848 cites W2921605821 @default.
• W3080451848 cites W2946511237 @default.
• W3080451848 cites W2954301664 @default.
• W3080451848 cites W2962725887 @default.
• W3080451848 cites W2962758826 @default.
• W3080451848 cites W2963069632 @default.
• W3080451848 cites W2963341956 @default.
• W3080451848 cites W2963403751 @default.
• W3080451848 cites W2963474950 @default.
• W3080451848 cites W2963815651 @default.
• W3080451848 cites W2964121744 @default.
• W3080451848 cites W2969388332 @default.
• W3080451848 cites W2970971581 @default.
• W3080451848 cites W2973727699 @default.
• W3080451848 cites W2991040477 @default.
• W3080451848 cites W3004822234 @default.
• W3080451848 cites W3010637564 @default.
• W3080451848 cites W3012479151 @default.
• W3080451848 cites W3021613070 @default.
• W3080451848 cites W3037519745 @default.
• W3080451848 cites W3037639655 @default.
• W3080451848 cites W3096583839 @default.
• W3080451848 cites W3096956001 @default.
• W3080451848 cites W3097108668 @default.
• W3080451848 cites W3098903812 @default.
• W3080451848 cites W3118608800 @default.
• W3080451848 cites W60686164 @default.
• W3080451848 cites W95152782 @default.
• W3080451848 hasPublicationYear "2020" @default.
• W3080451848 type Work @default.
• W3080451848 sameAs 3080451848 @default.
• W3080451848 citedByCount "2" @default.
• W3080451848 countsByYear W30804518482020 @default.
• W3080451848 countsByYear W30804518482021 @default.
• W3080451848 crossrefType "posted-content" @default.
• W3080451848 hasAuthorship W3080451848A5009547049 @default.
• W3080451848 hasAuthorship W3080451848A5012361506 @default.
• W3080451848 hasAuthorship W3080451848A5037557529 @default.
• W3080451848 hasAuthorship W3080451848A5048138372 @default.
• W3080451848 hasAuthorship W3080451848A5073310637 @default.
• W3080451848 hasAuthorship W3080451848A5073820927 @default.
• W3080451848 hasConcept C111873713 @default.
• W3080451848 hasConcept C111919701 @default.
• W3080451848 hasConcept C120314980 @default.
• W3080451848 hasConcept C127413603 @default.
• W3080451848 hasConcept C157764524 @default.
• W3080451848 hasConcept C206729178 @default.

• next