{"id":11912,"date":"2023-03-20T16:08:22","date_gmt":"2023-03-20T15:08:22","guid":{"rendered":"https:\/\/www.betriebssysteme.org\/?page_id=11912"},"modified":"2023-04-03T14:28:35","modified_gmt":"2023-04-03T12:28:35","slug":"wsos2023lect_mutlu","status":"publish","type":"page","link":"https:\/\/www.betriebssysteme.org\/aktivitaeten\/winterschools\/wsos2023-en\/wsos2023prog\/wsos2023lect_mutlu\/","title":{"rendered":"Lecture: Memory-Centric Computing"},"content":{"rendered":"

Onur Mutlu (ETH Z\u00fcrich)<\/h3>\n
\n\t

Lecture Slides<\/a><\/p>\n

Computing is bottlenecked by data. Large amounts of application data overwhelm storage capability, communication capability, and computation capability of the modern machines we design today. As a result, many key applications' performance, efficiency, and scalability are bottlenecked by data movement. In this lecture, we describe three major shortcomings of modern architectures in terms of 1) dealing with data, 2) taking advantage of the vast amounts of data, and 3) exploiting different semantic properties of application data. We argue that an intelligent architecture should be designed to handle data well. We show that handling data well requires designing architectures based on three key principles: 1) data-centric, 2) data-driven, 3) data-aware. We give several examples for how to exploit each of these principles to design a much more efficient and high performance computing system. We especially discuss recent research that aims to fundamentally reduce memory latency and energy, and practically enable computation close to data, with at least two promising novels directions: 1) processing using memory, which exploits analog operational properties of memory chips to perform massively-parallel operations in memory, with low-cost changes, 2) processing near memory, which integrates sophisticated additional processing capability in memory controllers, the logic layer of 3D-stacked memory technologies, or memory chips to enable high memory bandwidth and low memory latency to near-memory logic. We show both types of architectures can enable orders of magnitude improvements in performance and energy consumption of many important workloads, such as graph analytics, database systems, machine learning, video processing, climate modelling, genome analysis. We discuss how to enable adoption of such fundamentally more intelligent architectures, which we believe are key to efficiency, performance, and sustainability. We conclude with some research opportunities in and guiding principles for future computing architecture and system designs.<\/p>\n

A short accompanying paper, which appeared in DATE 2021, can be found here and serves as recommended reading:<\/p>\n

https:\/\/people.inf.ethz.ch\/omutlu\/pub\/intelligent-architectures-for-intelligent-computingsystems-invited_paper_DATE21.pdf<\/a><\/p>\n

A longer overview & survey of modern memory-centric computing can be found here and also serves as recommended reading:<\/p>\n

\"A Modern Primer on Processing in Memory\" - https:\/\/arxiv.org\/abs\/2012.03112<\/a> <\/p>\n<\/div>\n<\/div><\/div><\/div>

\n
\n\t

Prof. Onur Mutlu, PhD \u00a0(ETZ Z\u00fcrich)<\/p>\n

 <\/p>\n<\/div>\n<\/div><\/div><\/div><\/div>

\n
\n\t

Onur Mutlu is a Professor of Computer Science at ETH Zurich.\u00a0He is also a faculty member at Carnegie Mellon University, where he\u00a0previously held the Strecker Early Career Professorship.\u00a0 His current\u00a0broader research interests are in computer architecture, systems,\u00a0hardware security, and bioinformatics. A variety of techniques he,\u00a0along with his group and collaborators, has invented over the years\u00a0have influenced industry and have been employed in commercial\u00a0microprocessors and\u00a0memory\/storage systems. He obtained his PhD and MS\u00a0in ECE from the University of Texas at Austin and BS degrees in\u00a0Computer Engineering and Psychology from the University of Michigan,\u00a0Ann Arbor. He started the Computer Architecture Group at Microsoft\u00a0Research (2006-2009), and held various product and research positions\u00a0at Intel Corporation, Advanced Micro Devices, VMware, and Google.\u00a0 He\u00a0received the Intel Outstanding Researcher Award, IEEE High Performance\u00a0Computer Architecture Test of Time Award, NVMW Persistent Impact Prize,\u00a0the IEEE Computer Society Edward J. McCluskey Technical Achievement\u00a0Award, ACM SIGARCH Maurice Wilkes Award, the inaugural\u00a0IEEE Computer Society Young Computer Architect Award, the inaugural\u00a0Intel Early Career Faculty Award, US National Science Foundation\u00a0CAREER Award, Carnegie Mellon University Ladd Research Award, faculty\u00a0partnership awards from various companies, and a healthy number of\u00a0best paper or \"Top Pick\" paper recognitions at various computer\u00a0systems, architecture, and security venues. He is an ACM Fellow \"for\u00a0contributions to computer architecture research, especially in\u00a0memory\u00a0systems\", IEEE Fellow for \"contributions to computer architecture\u00a0research and practice\", and an elected member of the Academy of Europe\u00a0(Academia Europaea). His computer architecture and digital logic\u00a0design course lectures and materials are freely available on YouTube\u00a0(https:\/\/www.youtube.com\/OnurMutluLectures<\/a>\u00a0), and his research group\u00a0makes a wide variety of software and hardware artifacts freely\u00a0available online (https:\/\/safari.ethz.ch\/<\/a>). For more information,\u00a0please see his webpage at\u00a0https:\/\/people.inf.ethz.ch\/omutlu\/<\/a>.<\/p>\n

 <\/p>\n<\/div>\n<\/div><\/div><\/div>

<\/div><\/div><\/div>","protected":false},"excerpt":{"rendered":"

Lecture Slides Computing is bottlenecked by data. Large amounts of application data overwhelm storage capability, communication capability, and computation capability of the modern machines we design today. As a result, many key applications‘ performance, efficiency, and scalability are bottlenecked by data movement. In this lecture, we describe three major shortcomings of modern architectures in terms […]<\/p>\n","protected":false},"author":21,"featured_media":0,"parent":11863,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"class_list":["post-11912","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/www.betriebssysteme.org\/wp-json\/wp\/v2\/pages\/11912","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.betriebssysteme.org\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/www.betriebssysteme.org\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/www.betriebssysteme.org\/wp-json\/wp\/v2\/users\/21"}],"replies":[{"embeddable":true,"href":"https:\/\/www.betriebssysteme.org\/wp-json\/wp\/v2\/comments?post=11912"}],"version-history":[{"count":8,"href":"https:\/\/www.betriebssysteme.org\/wp-json\/wp\/v2\/pages\/11912\/revisions"}],"predecessor-version":[{"id":11939,"href":"https:\/\/www.betriebssysteme.org\/wp-json\/wp\/v2\/pages\/11912\/revisions\/11939"}],"up":[{"embeddable":true,"href":"https:\/\/www.betriebssysteme.org\/wp-json\/wp\/v2\/pages\/11863"}],"wp:attachment":[{"href":"https:\/\/www.betriebssysteme.org\/wp-json\/wp\/v2\/media?parent=11912"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}