Seizing the Bandwidth Scaling of On-Package Interconnect in a Post-Moore's Law World

Seizing the Bandwidth Scaling of On-Package Interconnect in a Post-Moore's Law World
复制标题

在后摩尔定律世界中抓住封装互连的带宽扩展

DOI:
10.1145/3577193.3593702
复制
发表时间:
2023
期刊:
Proceedings of the 2023 International Conference on Supercomputing
影响因子:
--
通讯作者:
Wentzlaff, David
Wentzlaff, David
中科院分区:
--
文献类型:
--
作者:
Chirkov, Grigory;Wentzlaff, David

文献摘要

参考文献

相似文献

摩尔定律的缓慢和可预见的终结迫使设计者将目光投向了简单地增加晶体管,鼓励他们利用其他未使用的资源来提高芯片性能。与此同时,近年来,芯片间互连技术取得了巨大的飞跃,极大地增加了可用带宽。虽然摩尔定律的终结将不可避免地减缓单芯片设置的性能进步,但互连技术可能会继续扩大规模。我们展望了未来,设计者必须创造方法来利用互连利用率来获得更好的系统性能。作为将互连利用率转换为性能的功能的一个例子,我们提出了Meduza-一种用于未来芯片系统的写-更新一致性协议。Meduza将以前的写更新协议扩展到具有多级缓存层次结构的系统。与基于芯片的系统上的MESIF一致性协议相比,Meduza将我们基准测试套件的执行速度提高了19%。此外,Meduza承诺在未来的系统中拥有更多优势。这项工作表明,通过利用多余的互连带宽,在现代和未来的小芯片系统中具有显著的额外性能潜力。
The slowing and forecasted end of Moore's Law have forced designers to look beyond simply adding transistors, encouraging them to employ other unused resources as a manner to increase chip performance. At the same time, in recent years, inter-die interconnect technologies made a huge leap forward, dramatically increasing the available bandwidth. While the end of Moore's Law will inevitably slow down the performance advances of single-die setups, interconnect technologies will likely continue to scale. We envision a future where designers must create ways to exploit interconnect utilization for better system performance.As an example of a feature that converts interconnect utilization into performance, we present Meduza - a write-update coherence protocol for future chiplet systems. Meduza extends previous write-update protocols to systems with multi-level cache hierarchies. Meduza improves execution speed in our benchmark suite by 19% when compared to the MESIF coherence protocol on a chiplet-based system. Moreover, Meduza promises even more advantages in future systems. This work shows that by exploiting excess interconnect bandwidth, there is significant potential for additional performance in modern and future chiplet systems.
DOI: 10.1145/2818950.2818951
发表时间: 2015-10
期刊: Proceedings of the 2015 International Symposium on Memory Systems
影响因子: --
作者:
G. Loh;Natalie D. Enright Jerger;Ajaykumar Kannan;Yasuko Eckert
通讯作者: G. Loh;Natalie D. Enright Jerger;Ajaykumar Kannan;Yasuko Eckert
使用开源硬件组件进行敏捷芯片开发的可扩展方法:(特邀论文)
DOI: 10.1145/3508352.3561102
发表时间: 2022
期刊: 2022 IEEE/ACM International Conference On Computer Aided Design (ICCAD)
影响因子: --
作者:
Maico Cassel dos Santos;Tianyu Jia;M. Cochet;Karthik Swaminathan;Joseph Zuckerman;Paolo Mantovani;Davide Giri;J. Zhang;Erik Jens Loscalzo;Gabriele Tombesi;Kevin Tien;Nandhini Chandramoorthy;J. Wellman;David Brooks;Gu;K. Shepard;L. Carloni;P. Bose
通讯作者: P. Bose
适用于企业工作负载的实时 AI:IBM Telum 处理器
DOI: 10.1109/hcs52781.2021.9567422
发表时间: 2021
期刊: 2021 IEEE Hot Chips 33 Symposium (HCS)
影响因子: --
作者:
C. Jacobi
通讯作者: C. Jacobi
CryoCore:用于低温计算的快速而密集的处理器架构
DOI: 10.1109/isca45697.2020.00037
发表时间: 2020
期刊: 2020 ACM/IEEE 47th Annual International Symposium on Computer Architecture (ISCA)
影响因子: --
作者:
Ilkwon Byun;Dongmoon Min;Gyu;Seongmin Na;Jangwoo Kim
通讯作者: Jangwoo Kim
多处理器缓存一致性方案综述
DOI: 10.1109/2.55497
发表时间: 1990
期刊: Computer
影响因子: 2.2
作者:
P. Stenström
通讯作者: P. Stenström