Pasithea-1: An Energy-Efficient Sequential Reconfigurable Array With CPU-Like Programmability

IF 2.4 Q2 ENGINEERING, ELECTRICAL & ELECTRONIC IEEE open journal of circuits and systems Pub Date : 2024-12-16 DOI:10.1109/OJCAS.2024.3518110
Tobias Kaiser;Esther Gottschalk;Kai Biethahn;Friedel Gerfers
{"title":"Pasithea-1: An Energy-Efficient Sequential Reconfigurable Array With CPU-Like Programmability","authors":"Tobias Kaiser;Esther Gottschalk;Kai Biethahn;Friedel Gerfers","doi":"10.1109/OJCAS.2024.3518110","DOIUrl":null,"url":null,"abstract":"This work presents Pasithea-1, a coarse-grained reconfigurable array (CGRA) that combines energy efficiency with CPU-like programmability. Its extensible instruction set uses sequential control flow in code fragments of up to 64 RISC-like instructions, which encode control and dataflow graphs in adjacency lists. Combined with dedicated, uniform processing elements, this enables fast compilation from C source code (1.4 s mean compile time). Demonstrator measurements reveal energy efficiency of up to 601 int32 MIPS/mW at 0.59V and performance of up to 148 MIPS at 0.90 V. Compared to a RISC reference system, mean energy efficiency is improved by 2.24× with 1.71× higher execution times across 12 of 14 benchmarks. Program-dependent factors underlying variations in energy efficiency are identified using dynamic program analysis. To reduce operand transfer energy, seven interconnect topologies are evaluated: a flat bus, five crossbar variants and a logarithmic network. Best results are obtained for a crossbar topology, reducing mean dynamic tile energy by 19 %. Furthermore, floating-point (FP) support is added to the instruction set and evaluated using three binary-compatible microarchitectures, presenting distinct area-performance-energy tradeoffs. The interconnect and FP microarchitecture explorations demonstrate that, unlike CGRAs utilizing low-level bitstreams, Pasithea’s instruction set hides microarchitectural details, which makes it possible to optimize hardware without severing binary compatibility.","PeriodicalId":93442,"journal":{"name":"IEEE open journal of circuits and systems","volume":"6 ","pages":"1-13"},"PeriodicalIF":2.4000,"publicationDate":"2024-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10802954","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"IEEE open journal of circuits and systems","FirstCategoryId":"1085","ListUrlMain":"https://ieeexplore.ieee.org/document/10802954/","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"ENGINEERING, ELECTRICAL & ELECTRONIC","Score":null,"Total":0}
引用次数: 0

Abstract

This work presents Pasithea-1, a coarse-grained reconfigurable array (CGRA) that combines energy efficiency with CPU-like programmability. Its extensible instruction set uses sequential control flow in code fragments of up to 64 RISC-like instructions, which encode control and dataflow graphs in adjacency lists. Combined with dedicated, uniform processing elements, this enables fast compilation from C source code (1.4 s mean compile time). Demonstrator measurements reveal energy efficiency of up to 601 int32 MIPS/mW at 0.59V and performance of up to 148 MIPS at 0.90 V. Compared to a RISC reference system, mean energy efficiency is improved by 2.24× with 1.71× higher execution times across 12 of 14 benchmarks. Program-dependent factors underlying variations in energy efficiency are identified using dynamic program analysis. To reduce operand transfer energy, seven interconnect topologies are evaluated: a flat bus, five crossbar variants and a logarithmic network. Best results are obtained for a crossbar topology, reducing mean dynamic tile energy by 19 %. Furthermore, floating-point (FP) support is added to the instruction set and evaluated using three binary-compatible microarchitectures, presenting distinct area-performance-energy tradeoffs. The interconnect and FP microarchitecture explorations demonstrate that, unlike CGRAs utilizing low-level bitstreams, Pasithea’s instruction set hides microarchitectural details, which makes it possible to optimize hardware without severing binary compatibility.
查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
一种具有类cpu可编程性的高能效顺序可重构阵列
这项工作提出了Pasithea-1,一种粗粒度可重构阵列(CGRA),它结合了能源效率和类似cpu的可编程性。它的可扩展指令集在多达64个类似risc指令的代码片段中使用顺序控制流,这些指令在邻接表中编码控制和数据流图。结合专用的、统一的处理元素,这使得从C源代码快速编译(平均编译时间1.4秒)成为可能。演示测量显示,在0.59V时,能量效率高达601 int32 MIPS/mW,在0.90 V时,性能高达148 MIPS。与RISC参考系统相比,在14个基准测试中的12个基准测试中,平均能源效率提高了2.24倍,执行时间提高了1.71倍。利用动态程序分析确定了能源效率变化的程序依赖因素。为了减少操作数传递能量,评估了7种互连拓扑:扁平总线、5种交叉杆变体和对数网络。横条拓扑的效果最好,平均动态能量降低了19%。此外,在指令集中添加了浮点(FP)支持,并使用三种二进制兼容的微体系结构进行评估,呈现出不同的面积-性能-能量权衡。互连和FP微架构的探索表明,与使用低级比特流的CGRAs不同,Pasithea的指令集隐藏了微架构细节,这使得在不切断二进制兼容性的情况下优化硬件成为可能。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 去求助
相关文献
SRSF9 mediates oncogenic RNA splicing of SLC37A4 via liquid–liquid phase separation to promote oral cancer progression
IF 10.7 1区 综合性期刊Journal of Advanced ResearchPub Date : 2025-03-08 DOI: 10.1016/j.jare.2025.03.013
Qiu Peng, Lujuan Wang, Ying Long, Hao Tian, Xuemeng Xu, Zongyao Ren, Yaqian Han, Xianjie Jiang, Zhu Wu, Shiming Tan, Wenjuan Yang, Linda Oyang, Xia Luo, Jinguan Lin, Longzheng Xia, Mingjing Peng, Nayiyuan Wu, Yanyan Tang, Qianjin Liao, Yujuan Zhou
来源期刊
自引率
0.00%
发文量
0
审稿时长
19 weeks
期刊最新文献
Nonlinear Analysis of Differential LC Oscillators and Injection Locked Frequency Dividers NLU: An Adaptive, Small-Footprint, Low-Power Neural Learning Unit for Edge and IoT Applications Energy Consumption Modeling of 2-D and 3-D Decoder Circuits 2024 Index IEEE Open Journal of Circuits and Systems Vol. 5 IEEE Circuits and Systems Society
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1