An Energy Efficient Floating Point Computing Infrastructure Embedding Ferroelectric Field Effect Transistor Based Ternary Content Addressable Memories
-
摘要: 随着数据密集型应用的日益增多，内存墙问题已成为制约计算效率的瓶颈。该文提出一种新型的浮点数(FP)运算结构，该结构嵌入了基于铁电场效应晶体管(FeFET)的三元内容寻址存储器(TCAM)以实现高效的计算。通过特定规则设计的超高密度TCAM结构，可以用能效更高的TCAM搜索操作代替部分传统浮点运算，从而节约整体能耗。仿真实验证明，该文所提结构和运算执行流程，与常规浮点运算单元(FPU)相比，可以降低多达33%的能耗。Abstract: With the growing abundance of data-intensive applications, memory wall has become a bottleneck to computing efficiency. A novel Floating Point (FP) computing infrastructure that embeds Ferroelectric Field Effect Transistor (FeFET) based Ternary Content Addressable Memories (TCAMs) for energy efficient computing is proposed. With an ultra-dense TCAM implementation following the designed guidelines, the infrastructure can replace unnecessary Float Point Unit (FPU) executions with more efficient TCAM searching, thereby saving the overall energy consumption. Thanks to the proposed execution flow, the infrastructure can achieve up to 33% energy saving compared to regular FPUs.
-
[1] WONG H S P and SALAHUDDIN S. Memory leads the way to better computing[J]. Nature Nanotech, 2015, 10(3): 191–194. doi: 10.1038/nnano.2015.29 [2] KARAM R, PURI R, GHOSH S, et al. Emerging trends in design and applications of memory-based computing and content-addressable memories[J]. Proceedings of the IEEE, 2015, 103(8): 1311–1330. doi: 10.1109/JPROC.2015.2434888 [3] LI Jing, MONTOYE R K, ISHII M, et al. 1 Mb 0.41 µm2 2T-2R cell nonvolatile TCAM with two-bit encoding and clocked self-referenced sensing[J]. IEEE Journal of Solid-State Circuits, 2014, 49(4): 896–907. doi: 10.1109/JSSC.2013.2292055 [4] GHOFRANI A, RAHIMI A, LASTRAS-MONTAÑO M A, et al. Associative memristive memory for approximate computing in GPUs[J]. IEEE Journal on Emerging and Selected Topics in Circuits and Systems, 2016, 6(2): 222–234. doi: 10.1109/JETCAS.2016.2538618 [5] IMANI M, RAHIMI A, and ROSING T S. Resistive configurable associative memory for approximate computing[C]. 2016 Design, Automation & Test in Europe Conference & Exhibition, Dresden, Germany, 2016: 1327–1332. [6] YIN Xunzhao, NIEMIER M, and HU X S. Design and benchmarking of ferroelectric FET based TCAM[C]. Design, Automation & Test in Europe Conference & Exhibition, Lausanne, Switzerland, 2017: 1444–1449. [7] NI Kai, YIN Xunzhao, LAGUNA A F, et al. Ferroelectric ternary content-addressable memory for one-shot learning[J]. Nature Electronics, 2019, 2(11): 521–529. doi: 10.1038/s41928-019-0321-3 [8] YIN Xunzhao, LI Chao, HUANG Qingrong, et al. FeCAM: A universal compact digital and analog content addressable memory using ferroelectric[J]. IEEE Transactions on Electron Devices, 2020, 67(7): 2785–2792. doi: 10.1109/TED.2020.2994896 [9] REIS D, NI Kai, CHAKRABORTY W, et al. Design and analysis of an ultra-dense, low-leakage, and fast FeFET-based random access memory array[J]. IEEE Journal on Exploratory Solid-State Computational Devices and Circuits, 2019, 5(2): 103–112. doi: 10.1109/JXCDC.2019.2930284 [10] PAGIAMTZIS K and SHEIKHOLESLAMI A. Content-addressable memory (CAM) circuits and architectures: A tutorial and survey[J]. IEEE Journal of Solid-State Circuits, 2006, 41(3): 712–727. doi: 10.1109/JSSC.2005.864128 [11] NI Kai, JERRY M, SMITH J A, et al. A circuit compatible accurate compact model for ferroelectric-FETs[C]. 2018 IEEE Symposium on VLSI Technology, Honolulu, USA, 2018: 131–132. [12] SONG T K. Landau-Khalatnikov simulations for ferroelectric switching in ferroelectric random access memory application[J]. Journal of the Korean Physical Society, 2005, 46(1): 5–9. [13] JASON P, HESTNESS J, ORR M S, et al. gem5-gpu: A Heterogeneous CPU-GPU Simulator[J]. IEEE Computer Architecture Letters, 2015, 14(1): 34–36. doi: 10.1109/LCA.2014.2299539 [14] FloPoCo[EB/OL]. http://flopoco.gforge.inria.fr. [15] Caltech 101[EB/OL]. http://www.vision.caltech.edu/Image_Datasets/Caltech101. -
下载:
图(7) / 表(1)
计量
- 文章访问数: 887
- HTML全文浏览量: 373
- PDF下载量: 98
- 被引次数: 0
