Z Ye, W Gao, Q Hu, P Sun, X Wang, Y Luo… - ACM Computing …, 2024 - dl.acm.org
Deep learning (DL) has demonstrated its remarkable success in a wide variety of fields. The development of a DL model is a time-consuming and resource-intensive procedure. Hence …
Hardware accelerators such as GPUs are required for real-time, low-latency inference with Deep Neural Networks (DNN). However, due to the inherent limits to the parallelism they …
J Lou, Y Sun, J Zhang, H Cao, Y Zhang… - CCF Transactions on High …, 2023 - Springer
With the development of deep learning, hardware accelerators represented by GPUs have been used to accelerate the execution of deep learning applications. A key problem in GPU …
Hazelwood et al. observed that at Facebook data centers, variations in user activity (eg due to diurnal load) resulted in low utilization periods with large pools of idle resources [4]. To …
Z Chen, X Zhao, C Zhi, J Yin - IEEE Transactions on Parallel …, 2023 - ieeexplore.ieee.org
Deep learning tasks (DLT) include training and inference tasks, where training DLTs have requirements on minimizing average job completion time (JCT) and inference tasks need …
W Cui, M Wei, Q Chen, X Tang, J Leng… - 2019 IEEE 37th …, 2019 - ieeexplore.ieee.org
GPUs have been widely adopted to serve online deep learning-based services that have stringent QoS requirements. However, emerging deep learning serving systems often result …
GPUs are known to benefit structured applications with ample parallelism, such as deep learning in a datacenter. Recently, GPUs have shown promise for irregular streaming …
I Baek, M Harding, A Kanda, KR Choi… - 2020 IEEE Real …, 2020 - ieeexplore.ieee.org
Modern hardware accelerators such as GP-GPUs and DSPs are commonly being used in real-time settings such as high-performance multimedia systems and autonomous vehicles …
H Albahar, S Dongare, Y Du, N Zhao… - 2022 22nd IEEE …, 2022 - ieeexplore.ieee.org
Modern cluster management systems, such as Kubernetes, support heterogeneous workloads and resources. However, existing resource schedulers in these systems do not …