CSR:Medium: A Cross-stack Approach to Reduce Memory Carbon in Cloud Data Centers

CSR：Medium：减少云数据中心内存碳的跨堆栈方法

• 批准号: 2312785
• 负责人: Xun Jian
• 金额: $ 100万
• 依托单位: Virginia Polytechnic Institute and State University
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-10-01 至 2026-09-30
• 项目状态: 未结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2312785&HistoricalAwards=false
• 关键词: CSR; Medium; Cross; stack; Approach

The scaling of memory capacity (i.e., bits per memory chip) increasingly lags behind the density scaling of other system components (e.g., cores per area in a CPU chip). As a result, a typical cloud server today contains 100s of memory chips, which may increase to 1000s under the current scaling trend; manufacturing and powering so many memory chips per server will be environmentally unsustainable. This project will explore how to co-design hardware and software to store values more densely in memory and reduce how much memory to manufacture and power to reduce the carbon footprint of future cloud data centers. While prior art has explored memory compression in hardware, they have only explored how to do so in the most rudimentary software scenarios - natively running a single program that accesses little to nothing beyond memory. The software stack in cloud is much more complex; user applications run in virtual machines, concurrently with collocated workloads, and often heavily exercise the operating system (OS) file cache and other in-memory caches. This project - CloudComp- will co-design hardware memory compression with different layers of the cloud system software (e.g., hypervisor, storage stack, in-memory databases, job scheduler) to enable practical deployment that can satisfy the diverse requirements and application scenarios in cloud. CloudComp will bring together researchers in computer architecture, cloud computing, OS, storage systems, and databases. To facilitate real world impact, this project will build and release real-system prototypes of hardware memory compression and partner closely with industry. By enabling an alternative path to scale up the effective size of the memory size for future cloud data centers, CloudComp will help reduce the impact of cloud computing on climate change over the brute-force approach of making more memory chips. Densely storing more data into the available amount of memory can also benefit climate in other ways such as enabling bigger-scale and/or finer-resolution climate modeling and simulation. CloudComp includes educational and engagement activities to broaden participation in research and attract new students, especially seeking out students from underrepresented groups. CloudComp will actively involve undergraduate students in building real-system prototypes to cultivate their curiosity for research. Lastly, the cross-layer insights gained through CloudComp will help guide the research of other complementary memory system techniques to combat the slowing physical scaling of memory. CloudComp is funded in part by the National Discovery Cloud for Climate (NDC-C) program as a core purpose of the project is to reduce the carbon emissions of cloud systems through this research.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

内存容量的扩展（即每个内存芯片的位数）越来越落后于其他系统组件的密度扩展（例如 CPU 芯片中每个区域的核心数），因此，当今典型的云服务器包含 100 个内存芯片，在当前的扩展趋势下，可能会增加到 1000 个；为每台服务器制造和供电如此多的内存芯片将在环境上不可持续。该项目将探索如何共同设计硬件和软件以更密集地存储值。虽然现有技术已经探索了硬件中的内存压缩，但他们只探索了如何在最基本的软件场景中做到这一点 - 本地运行一个。云中的软件堆栈在虚拟机中运行，与并置的工作负载同时运行，并且经常大量使用操作系统（OS）文件缓存和其他内存中的缓存。 .这个项目—— CloudComp将与云系统软件的不同层（例如管理程序、存储堆栈、内存数据库、作业调度程序）共同设计硬件内存压缩，以实现满足云中多样化需求和应用场景的实际部署。汇集计算机架构、云计算、操作系统、存储系统和数据库领域的研究人员，为了促进现实世界的影响，该项目将构建并发布硬件内存压缩的真实系统原型，并与业界密切合作。扩大有效性与未来云数据中心的内存大小相比，CloudComp 将有助于减少云计算对气候变化的影响，而不是通过暴力方式制造更多内存芯片，将更多数据密集存储到可用内存中也有利于气候变化。 CloudComp 还包括其他方式，例如实现更大规模和/或更高分辨率的气候建模和模拟，包括教育和参与活动，以扩大研究参与并吸引新学生，特别是寻找来自代表性不足群体的学生，CloudComp 将积极让本科生参与其中。构建真实系统原型最后，通过 CloudComp 获得的跨层见解将有助于指导其他互补内存系统技术的研究，以应对内存物理扩展速度放缓的问题。CloudComp 的部分资金由国家气候发现云 (National Discovery Cloud for Climate) 资助。 NDC-C）计划作为该项目的核心目的是通过这项研究减少云系统的碳排放。该奖项反映了 NSF 的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

项目成果

Towards cost-effective and resource-aware aggregation at Edge for Federated Learning

在边缘实现联邦学习的成本效益和资源感知聚合

• DOI: 10.1109/bigdata59044.2023.10386691
• 发表时间: 2023-12
• 期刊: Proceedings of IEEE International Conference on Big Data (BigData
• 作者: Khan, Ahmad Faraz;Li, Yuze;Wang, Xinran;Haroon, Sabaat;Ali, Haider;Cheng, Yue;Butt, Ali R.;Anwar, Ali
• 通讯作者: Anwar, Ali

Towards Efficient Python Interpreter for Tiered Memory Systems

面向分层内存系统的高效 Python 解释器

• 发表时间: 2024-02
• 期刊: Poster and Work-in-Progress in Proceedings of the 21st USENIX Conference on File and Storage Technologies (FAST
• 作者: Li, Yuze;Yao, Shunyu;Mobin, Jaiaid;Rafique, M. Mustafa;Nikolopoulos, Dimitrios;Sundararajah, Kirshanthan;Li, Huaicheng;Butt, Ali R.
• 通讯作者: Butt, Ali R.

Application-Attuned Memory Management for Containerized HPC Workflows

适用于容器化 HPC 工作流程的应用程序协调内存管理

• 发表时间: 2024-05
• 期刊: Proceedings of the 38th IEEE International Parallel & Distributed Processing Symposium (IPDPS
• 作者: Arif, Moiz;Maurya, Avinash;Rafique, M. Mustafa;Nikolopoulos, Dimitrios S.;Butt. Ali R.
• 通讯作者: Butt. Ali R.

FLOAT: Federated Learning Optimizations with Automated Tuning.

FLOAT：具有自动调优的联合学习优化。

• 发表时间: 2024-04
• 期刊: Nineteenth European Conference on Computer Systems (EuroSys ’24
• 作者: Khan, Ahmad Faraz;Khan, Azal Ahmad;Abdelmoniem, Ahmed M.;Fountain, Samuel;Butt, Ali R.;Anwar, Ali
• 通讯作者: Anwar, Ali