Segui
Lukasz Wesolowski
Lukasz Wesolowski
Research Scientist, Facebook
Email verificata su fb.com
Titolo
Citata da
Citata da
Anno
Accurate, large minibatch sgd: Training imagenet in 1 hour
P Goyal, P Dollár, R Girshick, P Noordhuis, L Wesolowski, A Kyrola, ...
arXiv preprint arXiv:1706.02677, 2017
37792017
Parallel programming with migratable objects: Charm++ in practice
B Acun, A Gupta, N Jain, A Langer, H Menon, E Mikida, X Ni, M Robson, ...
SC'14: Proceedings of the International Conference for High Performance …, 2014
2362014
Adaptive techniques for clustered N-body cosmological simulations
H Menon, L Wesolowski, G Zheng, P Jetley, L Kale, T Quinn, F Governato
Computational Astrophysics and Cosmology 2, 1-16, 2015
1612015
Scaling hierarchical N-body simulations on GPU clusters
P Jetley, L Wesolowski, F Gioachin, LV Kalé, TR Quinn
SC'10: Proceedings of the 2010 ACM/IEEE International Conference for High …, 2010
1192010
Overcoming the scalability challenges of epidemic simulations on blue waters
JS Yeom, A Bhatele, K Bisset, E Bohm, A Gupta, LV Kale, M Marathe, ...
2014 IEEE 28th International Parallel and Distributed Processing Symposium …, 2014
572014
Distributed training and prediction using elastic resources
L Wesolowski, MFM Abd El Aziz, AR Kalro, H Jia, J Parikh
US Patent 11,003,992, 2021
422021
Charm++
LV Kale, S Krishnan
421996
Charm++ for productivity and performance: A submission to the 2011 HPC class II challenge
L Kale, A Arya, A Bhatele, A Gupta, N Jain, P Jetley, J Lifflander, P Miller, ...
Parallel Programming Laboratory, Tech. Rep, 11-49, 2011
402011
Tram: Optimizing fine-grained communication with topological routing and aggregation of messages
L Wesolowski, R Venkataraman, A Gupta, JS Yeom, K Bisset, Y Sun, ...
2014 43rd International Conference on Parallel Processing, 211-220, 2014
372014
Understanding application performance via micro-benchmarks on three large supercomputers: Intrepid, Ranger and Jaguar
A Bhatelé, L Wesolowski, E Bohm, E Solomonik, LV Kalé
The International Journal of High Performance Computing Applications 24 (4 …, 2010
342010
Migratable objects+ active messages+ adaptive runtime= productivity+ performance a submission to 2012 HPC class II challenge
L Kale, A Arya, N Jain, A Langer, J Lifflander, H Menon, X Ni, Y Sun, ...
Parallel Programming Laboratory, Tech. Rep, 12-47, 2012
332012
An application programming interface for general purpose graphics processing units in an asynchronous runtime system
L Wesolowski
University of Illinois at Urbana-Champaign, 2008
302008
Architectural constraints to attain 1 exaflop/s for three scientific application classes
A Bhatele, P Jetley, H Gahvari, L Wesolowski, WD Gropp, L Kale
2011 IEEE International Parallel & Distributed Processing Symposium, 80-91, 2011
262011
2010 ACM/IEEE International Conference for High Performance Computing, Networking, Storage and Analysis
P Jetley, L Wesolowski, F Gioachin, LV Kalé, TR Quinn
IEEE Computer Society, Washington, DC, USA, 2010
232010
The Charm++ parallel programming system
L Kalé, B Acun, S Bak, A Becker, M Bhandarkar, N Bhat, A Bhatele, ...
Aug, 2019
112019
Datacenter-scale analysis and optimization of gpu machine learning workloads
L Wesolowski, B Acun, V Andrei, A Aziz, G Dankel, C Gregg, X Meng, ...
IEEE Micro 41 (5), 101-112, 2021
82021
Accelerator Support in the Charm++ Parallel Programming Model.
LV Kale, DM Kunzman, L Wesolowski
Scientific Computing with Multicore and Accelerators, 393-411, 2010
52010
Architectural constraints required to attain 1 exaflop/s for scientific applications
A Bhatele, P Jetley, H Gahvari, L Wesolowski, WD Gropp, LV Kale
Proc. Int. Parallel and Distributed Processing Symposium (IPDPS), 2011
42011
Software topological message aggregation techniques for large-scale parallel systems
L Wesolowski
University of Illinois at Urbana-Champaign, 2014
22014
Charm++ for productivity and performance
LV Kale, AAABA Gupta, N Jain, PJJ Lifflander, P Miller, Y Sun, ...
22012
Il sistema al momento non può eseguire l'operazione. Riprova più tardi.
Articoli 1–20