arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2002.07215cs.DCcs.LG
收藏

STANNIS:使用计算存储的深度神经网络训练的低功耗加速

STANNIS: Low-Power Acceleration of Deep Neural Network Training Using Computational Storage

University of California, Irvine(加州大学欧文分校);NGD Systems Inc.(NGD系统公司)

Ali HeydariGorji, Mahdi Torabzadehkashi, Siavash Rezaei, Hossein Bobarshad, Vladimir Alves, Pai H. Chou

摘要:

本文提出了一个分布式的,在存储训练的神经网络集群的计算存储设备的框架。这些设备不仅包含硬件加速器,还消除了主机和存储之间的数据移动,从而提高了性能并节省了功耗。更重要的是,这种存储处理风格的训练确保了私有数据永远不会离开存储,同时完全控制公共数据的共享。实验结果表明,该算法的加速比高达2.7倍,能耗降低了69%,准确性没有明显损失。

英文摘要:

This paper proposes a framework for distributed, in-storage training of neural networks on clusters of computational storage devices. Such devices not only contain hardware accelerators but also eliminate data movement between the host and storage, resulting in both improved performance and power savings. More importantly, this in-storage processing style of training ensures that private data never leaves the storage while fully controlling the sharing of public data. Experimental results show up to 2.7x speedup and 69% reduction in energy consumption and no significant loss in accuracy.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑