arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2010.08065cs.ARcs.AI
收藏

FPRaker:一种加速神经网络训练的处理单元

FPRaker: A Processing Element For Accelerating Neural Network Training

University of Toronto(多伦多大学);Vector Institute(向量研究所)

Omar Mohamed Awad, Mostafa Mahmoud, Isak Edo, Ali Hadi Zadeh, Ciaran Bannon, Anand Jayarajan, Gennady Pekhimenko, Andreas Moshovos

摘要:

我们提出FPRaker,一个处理元件组成的训练加速器。FPRaker同时处理多个浮点乘法累加运算,并将其结果累加到更高精度的累加器中。FPRaker通过利用训练过程中自然出现的值来提高训练过程中的性能和能源效率。具体地说,它将每个乘-累加的操作数的有效数处理为一系列2的有符号幂。转换为这种形式是动态完成的。这暴露了可以跳过的无效工作:编码时的值只有很少的项,其中一些可以被丢弃,因为它们会落在浮点精度有限的累加器范围之外。我们证明,FPRaker可以用来组成一个加速器的培训,它可以提高性能和能源效率相比,使用传统的浮点单元ISO计算面积的限制。我们还证明,FPRaker提供额外的好处时,训练合并修剪和量化。最后,我们展示了FPRaker自然地放大了每层使用不同精度的训练方法的性能。

英文摘要:

We present FPRaker, a processing element for composing training accelerators. FPRaker processes several floating-point multiply-accumulation operations concurrently and accumulates their result into a higher precision accumulator. FPRaker boosts performance and energy efficiency during training by taking advantage of the values that naturally appear during training. Specifically, it processes the significand of the operands of each multiply-accumulate as a series of signed powers of two. The conversion to this form is done on-the-fly. This exposes ineffectual work that can be skipped: values when encoded have few terms and some of them can be discarded as they would fall outside the range of the accumulator given the limited precision of floating-point. We demonstrate that FPRaker can be used to compose an accelerator for training and that it can improve performance and energy efficiency compared to using conventional floating-point units under ISO-compute area constraints. We also demonstrate that FPRaker delivers additional benefits when training incorporates pruning and quantization. Finally, we show that FPRaker naturally amplifies performance with training methods that use a different precision per layer.

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。

↑