An STT-MRAM Based in Memory Architecture for Low Power Integral Computing

An STT-MRAM Based in Memory Architecture for Low Power Integral Computing
复制标题

基于内存架构的 STT-MRAM,用于低功耗集成计算

DOI:
10.1109/tc.2018.2879502
复制
发表时间:
2019-04
影响因子:
3.7
通讯作者:
Weisheng Zhao
Weisheng Zhao
中科院分区:
计算机科学2区
文献类型:
--
作者:
Yinglin Zhao;Peng Ouyang;Wang Kang;Shouyi Yin;Youguang Zhang;Shaojun Wei;Weisheng Zhao

文献摘要

参考文献

相似文献

在视觉算法中,积分直方图图像在加速特征计算方面起着重要的作用。然而,积分直方图的计算过程,称为积分计算,具有较高的计算复杂度和大量的内存访问操作,这限制了它的广泛应用。本文提出了一种基于自旋转移矩磁性随机存取存储器(STT-MRAM)的内存计算架构来解决这些问题。该体系结构可以工作在两种不同的模式,根据不同的要求:积分计算模式和内存模式。该结构在积分计算模式下可以计算出积分直方图,在存储模式下可以直接存储数据。利用STT-MRAM的非易失性、高密度和低功耗的特性,将计算单元集成到存储器阵列中,实现并行计算。减少存储单元和计算单元之间的数据传输数量有助于降低延迟和能耗。评估结果表明,与最先进的工作相比,我们的架构提供<inline-formula><tex-math notation="LaTeX">了</tex-math><alternatives><mml:math><mml:mrow><mml:mn>1.1\times\sim9\times$1</mml:mn><mml:mo>。</mml:mo><mml:mn>1</mml:mn><mml:mo></mml:mo><mml:mo>×</mml:mo><mml:mn>9</mml:mn><mml:mo>×</mml:mo></mml:mrow></mml:math><inline-graphic xlink:href="zhao-ieq1-2879502.gif"/></alternatives></inline-formula>性能改进并降低87.4 <inline-formula><tex-math notation="LaTeX">$\sim</tex-math><alternatives><mml:math><mml:mo>$97.3%</mml:mo></mml:math><inline-graphic xlink:href="zhao-ieq2-2879502.gif"/></alternatives></inline-formula>的能耗,用于<inline-formula><tex-math notation="LaTeX">$64\times 64\sim 512\times 512</tex-math><alternatives><mml:math><mml:mrow><mml:mn>$64</mml:mn><mml:mo>×</mml:mo><mml:mn>64</mml:mn><mml:mo></mml:mo><mml:mn>512</mml:mn><mml:mo>×</mml:mo><mml:mn>512</mml:mn></mml:mrow></mml:math><inline-graphic xlink:href="zhao-ieq3-2879502.gif"/></alternatives></inline-formula>大小的图像,仅需8%的面积开销。
The integral histogram image plays an important role in accelerating the feature computation in vision algorithms. However, the computational process of the integral histogram, called integral computation, has high computational complexity and numerous memory access operations, which limit its wide application. This brief proposes an in-memory computational architecture based on Spin Transfer Torque Magnetic Random Access Memory (STT-MRAM) to solve these problems. The architecture can work in two different modes depending on the requirements: the integral computation mode and the memory mode. The architecture can figure out the integral histogram when in the integral computation mode, and just store the data directly when in the memory mode. Utilizing the non-volatile, high density and low power characteristics of STT-MRAM, we integrate the computational units into the memory array to achieve parallel computation. Reduced number of data transmission between storage units and computation units contributes to cut down the latency and energy consumption. The evaluation results show that, comparing with the state-of-the-art work, our architecture provides <inline-formula><tex-math notation="LaTeX">$1.1\times \sim 9\times$</tex-math><alternatives><mml:math><mml:mrow><mml:mn>1</mml:mn><mml:mo>.</mml:mo><mml:mn>1</mml:mn><mml:mo>×</mml:mo><mml:mo>∼</mml:mo><mml:mn>9</mml:mn><mml:mo>×</mml:mo></mml:mrow></mml:math><inline-graphic xlink:href="zhao-ieq1-2879502.gif"/></alternatives></inline-formula> performance improvements and reduces 87.4 <inline-formula><tex-math notation="LaTeX">$\sim$</tex-math><alternatives><mml:math><mml:mo>∼</mml:mo></mml:math><inline-graphic xlink:href="zhao-ieq2-2879502.gif"/></alternatives></inline-formula> 97.3 percent energy consumption for <inline-formula><tex-math notation="LaTeX">$64\times 64\sim 512\times 512$</tex-math><alternatives><mml:math><mml:mrow><mml:mn>64</mml:mn><mml:mo>×</mml:mo><mml:mn>64</mml:mn><mml:mo>∼</mml:mo><mml:mn>512</mml:mn><mml:mo>×</mml:mo><mml:mn>512</mml:mn></mml:mrow></mml:math><inline-graphic xlink:href="zhao-ieq3-2879502.gif"/></alternatives></inline-formula> size images, just with a 8 percent area overhead.
DOI: 10.1109/tcsii.2014.2362651
发表时间: 2015-01
期刊: IEEE Transactions on Circuits and Systems II: Express Briefs
影响因子: --
作者:
P. Ouyang;Shouyi Yin;Yuchi Zhang;Leibo Liu;Shaojun Wei
通讯作者: P. Ouyang;Shouyi Yin;Yuchi Zhang;Leibo Liu;Shaojun Wei
DOI: 10.1109/date.2009.5090761
发表时间: 2009-04
期刊: 2009 Design, Automation & Test in Europe Conference & Exhibition
影响因子: --
作者:
Hai Helen Li;Yiran Chen
通讯作者: Hai Helen Li;Yiran Chen
DOI: 10.1109/micro.2007.30
发表时间: 2007-12
期刊: 40th Annual IEEE/ACM International Symposium on Microarchitecture (MICRO 2007)
影响因子: --
作者:
N. Muralimanohar;R. Balasubramonian;N. Jouppi
通讯作者: N. Muralimanohar;R. Balasubramonian;N. Jouppi
DOI: 10.1109/tnano.2015.2456510
发表时间: 2015-07
影响因子: 2.4
作者:
K. Kwon;Xuanyao Fong;Parami Wijesinghe;P. Panda;K. Roy
通讯作者: K. Kwon;Xuanyao Fong;Parami Wijesinghe;P. Panda;K. Roy
DOI: 10.1109/vlsic.2018.8502388
发表时间: 2018-06
期刊: 2018 IEEE Symposium on VLSI Circuits
影响因子: --
作者:
Shouyi Yin;P. Ouyang;Jianxun Yang;Tianyi Lu;Xiudong Li;Leibo Liu;Shaojun Wei
通讯作者: Shouyi Yin;P. Ouyang;Jianxun Yang;Tianyi Lu;Xiudong Li;Leibo Liu;Shaojun Wei