arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1602.01616cs.ARcs.NE
收藏

基于FPGA的仅使用片上存储器的深度神经网络实现

FPGA Based Implementation of Deep Neural Networks Using On-chip Memory Only

Jinhwan Park, Wonyong Sung

摘要:

深度神经网络(DNN)需要非常大量的计算和权重存储,因此非常期望使用专用硬件的高效实现。在这项工作中,我们开发了一个基于FPGA的定点DNN系统,只使用片上存储器,不访问外部DRAM。所开发的系统的执行时间和能耗进行比较与基于GPU的实现。由于FPGA存储器容量有限,该算法仅使用3位权值,并采用基于训练的定点权值优化。使用Xilinx XC7Z045的实现进行了测试的MNIST手写数字识别基准和TIMIT语料库上的音素识别任务。所获得的速度约为基于GPU的实现的四分之一,并且比基于PC的实现好得多。全速运行时的功耗小于5瓦,与基于GPU的系统相比,效率更高。

英文摘要:

Deep neural networks (DNNs) demand a very large amount of computation and weight storage, and thus efficient implementation using special purpose hardware is highly desired. In this work, we have developed an FPGA based fixed-point DNN system using only on-chip memory not to access external DRAM. The execution time and energy consumption of the developed system is compared with a GPU based implementation. Since the capacity of memory in FPGA is limited, only 3-bit weights are used for this implementation, and training based fixed-point weight optimization is employed. The implementation using Xilinx XC7Z045 is tested for the MNIST handwritten digit recognition benchmark and a phoneme recognition task on TIMIT corpus. The obtained speed is about one quarter of a GPU based implementation and much better than that of a PC based one. The power consumption is less than 5 Watt at the full speed operation resulting in much higher efficiency compared to GPU based systems.

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。

↑