Asynchronous Decentralized SGD with Quantized and Local Updates G Nadiradze, A Sabour, P Davies, S Li, D Alistarh Advances in Neural Information Processing Systems 34, 2021 | 60* | 2021 |
Distributionally linearizable data structures D Alistarh, T Brown, J Kopinsky, JZ Li, G Nadiradze Proceedings of the 30th on Symposium on Parallelism in Algorithms and …, 2018 | 33 | 2018 |
The power of choice in priority scheduling D Alistarh, J Kopinsky, J Li, G Nadiradze Proceedings of the ACM Symposium on Principles of Distributed Computing, 283-292, 2017 | 31 | 2017 |
On approximating strip packing with a better ratio than 3/2 G Nadiradze, A Wiese Proceedings of the twenty-seventh annual ACM-SIAM symposium on discrete …, 2016 | 30 | 2016 |
Elastic Consistency: A Practical Consistency Model for Distributed Stochastic Gradient Descent G Nadiradze, I Markov, B Chatterjee, V Kungurtsev, D Alistarh Proceedings of the AAAI Conference on Artificial Intelligence 35 (10), 9037-9045, 2021 | 19* | 2021 |
Relaxed schedulers can efficiently parallelize iterative algorithms D Alistarh, T Brown, J Kopinsky, G Nadiradze Proceedings of the 2018 ACM symposium on principles of distributed computing …, 2018 | 15 | 2018 |
Breaking (global) barriers in parallel stochastic optimization with wait-avoiding group averaging S Li, T Ben-Nun, G Nadiradze, S Di Girolamo, N Dryden, D Alistarh, ... IEEE Transactions on Parallel and Distributed Systems 32 (7), 1725-1739, 2020 | 13 | 2020 |
Multi-queues can be state-of-the-art priority schedulers A Postnikova, N Koval, G Nadiradze, D Alistarh Proceedings of the 27th ACM SIGPLAN Symposium on Principles and Practice of …, 2022 | 9 | 2022 |
Dynamic averaging load balancing on cycles D Alistarh, G Nadiradze, A Sabour Algorithmica 84 (4), 1007-1029, 2022 | 9 | 2022 |
Efficiency guarantees for parallel incremental algorithms under relaxed schedulers D Alistarh, G Nadiradze, N Koval The 31st ACM Symposium on Parallelism in Algorithms and Architectures, 145-154, 2019 | 9 | 2019 |
Bounded Diameter Minimum Spanning Tree G Nadiradze Budapešť, Maďarsko, 2013 | 8 | 2013 |
Quafl: Federated averaging can be both asynchronous and communication-efficient H Zakerinia, S Talaei, G Nadiradze, D Alistarh arXiv preprint arXiv:2206.10032, 2022 | 7 | 2022 |
On achieving scalability through relaxation G Nadiradze | 6 | 2021 |
Lower bounds for shared-memory leader election under bounded write contention D Alistarh, R Gelashvili, G Nadiradze arXiv preprint arXiv:2108.02802, 2021 | 2 | 2021 |
The transactional conflict problem D Alistarh, SK Haider, R Kübler, G Nadiradze Proceedings of the 30th on Symposium on Parallelism in Algorithms and …, 2018 | 2 | 2018 |
Communication-Efficient Federated Learning With Data and Client Heterogeneity H Zakerinia, S Talaei, G Nadiradze, D Alistarh International Conference on Artificial Intelligence and Statistics, 3448-3456, 2024 | 1 | 2024 |
Provably-Efficient and Internally-Deterministic Parallel Union-Find A Fedorov, D Hashemi, G Nadiradze, D Alistarh Proceedings of the 35th ACM Symposium on Parallelism in Algorithms and …, 2023 | | 2023 |
Hybrid Decentralized Optimization: First-and Zeroth-Order Optimizers Can Be Jointly Leveraged For Faster Convergence S Talaei, G Nadiradze, D Alistarh arXiv preprint arXiv:2210.07703, 2022 | | 2022 |
QuAFL: Federated Averaging Made Asynchronous and Communication-Efficient H Zakerinia, S Talaei, G Nadiradze, D Alistarh | | 2022 |
Elastic Consistency: A Consistency Criterion for Distributed Optimization D Alistarh, I Markov, G Nadiradze ACM SIGACT News 53 (2), 64-82, 2022 | | 2022 |