{"id":365330,"date":"2017-02-21T12:06:28","date_gmt":"2017-02-21T20:06:28","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-event&#038;p=365330"},"modified":"2025-08-06T11:58:05","modified_gmt":"2025-08-06T18:58:05","slug":"first-international-workshop-on-rack-scale-computing-wrsc-2014","status":"publish","type":"msr-event","link":"https:\/\/www.microsoft.com\/en-us\/research\/event\/first-international-workshop-on-rack-scale-computing-wrsc-2014\/","title":{"rendered":"First International Workshop on Rack-scale Computing (WRSC 2014)"},"content":{"rendered":"\n\n<p>Co-located with <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" target=\"_blank\" href=\"http:\/\/eurosys2014.vu.nl\/\">EuroSys 2014<span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/p>\n<p>Questions? Email: <a href=\"mailto:wrsc_chairs@microsoft.com\">wrsc_chairs@microsoft.com<\/a><span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<h3>Organization<\/h3>\n<ul>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/pcosta\/\">Paolo Costa<\/a>, Microsoft Research (Program Chair)<\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/dnarayan\/\">Dushyanth Narayanan<\/a>, Microsoft Research (General Chair)<\/li>\n<\/ul>\n<h3>Program Committee<\/h3>\n<ul>\n<li>Gustavo Alonso, ETH Zurich<\/li>\n<li>Edouard Bugnion, EPFL<\/li>\n<li>Luis Ceze, U. Washington<\/li>\n<li>Paolo Costa, Microsoft Research<\/li>\n<li>Leendert van Doorn, AMD<\/li>\n<li>Babak Falsafi, EPFL<\/li>\n<li>Blake Fitch, IBM Research<\/li>\n<li>Nathan Farrington, Facebook<\/li>\n<li>Tim Harris, Oracle Labs<\/li>\n<li>Michael Kaminsky, Intel Labs<\/li>\n<li>Dushyanth Narayanan, Microsoft Research<\/li>\n<li>Parthasarathy (Partha)\u00a0Ranganathan, Google<\/li>\n<li>Luigi Rizzo, U. Pisa<\/li>\n<li>Thomas Wenisch, U. Michigan<\/li>\n<li>Bernard Wong, U. Waterloo<\/li>\n<\/ul>\n<h3>Sponsored by<\/h3>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-medium wp-image-365336\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/logos2-285x300.png\" alt=\"\" width=\"285\" height=\"300\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/logos2-285x300.png 285w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/logos2.png 300w\" sizes=\"auto, (max-width: 285px) 100vw, 285px\" \/><span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<h2>Workshop Program (<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/wrsc-poster.pdf\">CFP<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>)<\/h2>\n<h3>Session 1<\/h3>\n<ul type=\"disc\">\n<li><strong>Paolo Costa, Microsoft Research<\/strong><br \/>\nTitle: Towards Rack-scale Computing: Challenges and Opportunities (<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/costa14rackscale.pdf\" target=\"_self\">slides<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>)<br \/>\nAbstract:\u00a0New hardware technology such as systems- and networks-on-chip (SOCs and NOCs), switchless network fabrics, silicon photonics, and RDMA, are redefining the landscape of data center computing, enabling interconnecting thousands of cores at high speed at the scale of today&#8217;s racks. We refer to this new class of hardware as rack-scale computers because the rack is increasingly replacing the individual server as the basic building block of modern data centers. Most of the benefits promised by these new architectures, however, can only be achieved with adequate support from the software stack. In this talk, I will describe some of the research challenges and the opportunities introduced by rack-scale computing. As a concrete example, I will provide an overview of some of the research projects related to this topic that we are pursuing in our group.<\/li>\n<li><strong>Tyler Szepesi, Bernard Wong,\u00a0Ben Cassell,\u00a0Tim Brecht, University of Waterloo<\/strong><br \/>\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/szepsei14designing.pdf\">Designing A Low-Latency Cuckoo Hash Table for Write-Intensive Workloads<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>\u00a0(<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/szepsei14designing_slides.pdf\" target=\"_self\">slides<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>)<\/li>\n<\/ul>\n<h3>Session 2<\/h3>\n<ul type=\"disc\">\n<li><strong>Martin Maas, Krste Asanovic, UC Berkeley, Tim Harris, Oracle Labs, John Kubiatowicz, UC Berkeley.<\/strong><br \/>\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/maas14case.pdf\" target=\"_self\">The Case for the Holistic Language Runtime System<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>\u00a0(<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/maas14case_slides.pdf\" target=\"_self\">slides<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>)<\/li>\n<li><strong>Jacob Nelson, Brandon Holt, Brandon Myers, Preston Briggs, Simon Kahan, Luis Ceze, Mark Oskin, University of Washington.<\/strong><br \/>\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/nelson14grappa.pdf\">Grappa: A Latency-Tolerant Runtime for Large-Scale Irregular Application<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> (<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/nelson14grappa_slides.pdf\" target=\"_self\">slides<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>)<\/li>\n<li><strong>Tim Harris, Maurice Herlihy, Yossi Lev, Oracle Labs, Yujie Liu, Lehigh University, Victor Luchangco, Virendra Marathe,\u00a0Mark Moir, Oracle Labs.<\/strong><br \/>\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/harris14towards.pdf\">Towards Whatever-Scale Abstractions for Data-Driven Parallelism<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>\u00a0(<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/harris14towards_slides.pdf\" target=\"_self\">slides<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>)<\/li>\n<\/ul>\n<h3>Session 3<\/h3>\n<ul type=\"disc\">\n<li><strong>Gustavo Alonso, ETH Zurich<\/strong><br \/>\nTitle: Rackscale &#8211; the things that matter\u00a0(<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/alonso14wrsc.pdf\" target=\"_self\">slides<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>)<br \/>\nAbstract:\u00a0Rackscale computing has become the standard for many applications running on a data center. For a variety of reasons, today it is possible to developed fully customized solution that achieve impressive performance numbers. In this talk I will argue that customization is important but needs to be sustained by general purpose techniques and components. The research agenda in the next years should focus on the latter, rather than on producing an infinity variety of high performance systems tailored for narrow use cases. Otherwise, the inevitable problems with total cost of ownership during the life cycle of a real systems (maintenance, further development, software evolution, additional functionality) will soon catch up with many existing proposals.<\/li>\n<li><strong>Tim Harris, Oracle Labs<\/strong><br \/>\nTitle: What We Talk About When We Talk About Scheduling (<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/harris14wrsc.pdf\" target=\"_self\">slides<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>)<br \/>\nAbstract:\u00a0Distributed workloads involve scheduling and resource allocation decisions at multiple levels of the stack: deciding which machines a job will run on; deciding which instances of replicated services they will use; arbitrating within those services between multiple clients; deciding which VM&#8217;s virtual CPUs get which physical cores, and which threads run on those CPUs; deciding how the instructions in those threads are scheduled in multi-threaded processors, and how they compete for resources in the cores and in the interconnect. In this talk I will illustrate the kind of interference that can occur at these different levels, and talk about some possible approaches for handling these problems.<\/li>\n<li><strong>Sanjeev Kumar, Facebook<\/strong><br \/>\nTitle: Efficiency at Scale (<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/kumar14wrsc.pdf\" target=\"_self\">slides<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>)<\/li>\n<\/ul>\n<h3><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"http:\/\/clouddp14.unine.ch\/posterSession.html\" target=\"_blank\">Joint poster session<span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/h3>\n<h3>Panel session<\/h3>\n<p>Panelists:<\/p>\n<ul>\n<li><strong>Gustavo Alonso, ETH Zurich, Steve Hand, Microsoft Research, Tim Harris, Oracle Labs<\/strong><\/li>\n<li><strong>Sanjeev Kumar, Facebook, Ant Rowstron, Microsoft Research<\/strong><\/li>\n<\/ul>\n<p><span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Co-located with EuroSys 2014 (opens in new tab) Questions? Email: wrsc_chairs@microsoft.comOpens in a new tab Organization Paolo Costa, Microsoft Research (Program Chair) Dushyanth Narayanan, Microsoft Research (General Chair) Program Committee Gustavo Alonso, ETH Zurich Edouard Bugnion, EPFL Luis Ceze, U. Washington Paolo Costa, Microsoft Research Leendert van Doorn, AMD Babak Falsafi, EPFL Blake Fitch, IBM [&hellip;]<\/p>\n","protected":false},"featured_media":0,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","msr_startdate":"2014-04-13","msr_enddate":"2014-04-13","msr_location":"","msr_expirationdate":"","msr_event_recording_link":"","msr_event_link":"","msr_event_link_redirect":false,"msr_event_time":"","msr_hide_region":false,"msr_private_event":true,"msr_hide_image_in_river":0,"footnotes":""},"research-area":[13547],"msr-region":[],"msr-event-type":[210063],"msr-video-type":[],"msr-locale":[268875],"msr-program-audience":[],"msr-post-option":[],"msr-impact-theme":[],"class_list":["post-365330","msr-event","type-msr-event","status-publish","hentry","msr-research-area-systems-and-networking","msr-event-type-workshop","msr-locale-en_us"],"msr_about":"<!-- wp:msr\/event-details {\"title\":\"First International Workshop on Rack-scale Computing (WRSC 2014)\",\"backgroundColor\":\"grey\"} \/-->\n\n<!-- wp:msr\/content-tabs --><!-- wp:msr\/content-tab {\"title\":\"Summary\"} --><!-- wp:freeform --><p>Co-located with <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" target=\"_blank\" href=\"http:\/\/eurosys2014.vu.nl\/\">EuroSys 2014<\/a><\/p>\n<p>Questions? Email: <a href=\"mailto:wrsc_chairs@microsoft.com\">wrsc_chairs@microsoft.com<\/a><span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<h3>Organization<\/h3>\n<ul>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/pcosta\/\">Paolo Costa<\/a>, Microsoft Research (Program Chair)<\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/dnarayan\/\">Dushyanth Narayanan<\/a>, Microsoft Research (General Chair)<\/li>\n<\/ul>\n<h3>Program Committee<\/h3>\n<ul>\n<li>Gustavo Alonso, ETH Zurich<\/li>\n<li>Edouard Bugnion, EPFL<\/li>\n<li>Luis Ceze, U. Washington<\/li>\n<li>Paolo Costa, Microsoft Research<\/li>\n<li>Leendert van Doorn, AMD<\/li>\n<li>Babak Falsafi, EPFL<\/li>\n<li>Blake Fitch, IBM Research<\/li>\n<li>Nathan Farrington, Facebook<\/li>\n<li>Tim Harris, Oracle Labs<\/li>\n<li>Michael Kaminsky, Intel Labs<\/li>\n<li>Dushyanth Narayanan, Microsoft Research<\/li>\n<li>Parthasarathy (Partha)\u00a0Ranganathan, Google<\/li>\n<li>Luigi Rizzo, U. Pisa<\/li>\n<li>Thomas Wenisch, U. Michigan<\/li>\n<li>Bernard Wong, U. Waterloo<\/li>\n<\/ul>\n<h3>Sponsored by<\/h3>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-medium wp-image-365336\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/logos2-285x300.png\" alt=\"\" width=\"285\" height=\"300\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/logos2-285x300.png 285w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/logos2.png 300w\" sizes=\"auto, (max-width: 285px) 100vw, 285px\" \/><span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<!-- \/wp:freeform --><!-- \/wp:msr\/content-tab --><!-- wp:msr\/content-tab {\"title\":\"Agenda\"} --><!-- wp:freeform --><h2>Workshop Program (<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/wrsc-poster.pdf\">CFP<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>)<\/h2>\n<h3>Session 1<\/h3>\n<ul type=\"disc\">\n<li><strong>Paolo Costa, Microsoft Research<\/strong><br \/>\nTitle: Towards Rack-scale Computing: Challenges and Opportunities (<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/costa14rackscale.pdf\" target=\"_self\">slides<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>)<br \/>\nAbstract:\u00a0New hardware technology such as systems- and networks-on-chip (SOCs and NOCs), switchless network fabrics, silicon photonics, and RDMA, are redefining the landscape of data center computing, enabling interconnecting thousands of cores at high speed at the scale of today&#8217;s racks. We refer to this new class of hardware as rack-scale computers because the rack is increasingly replacing the individual server as the basic building block of modern data centers. Most of the benefits promised by these new architectures, however, can only be achieved with adequate support from the software stack. In this talk, I will describe some of the research challenges and the opportunities introduced by rack-scale computing. As a concrete example, I will provide an overview of some of the research projects related to this topic that we are pursuing in our group.<\/li>\n<li><strong>Tyler Szepesi, Bernard Wong,\u00a0Ben Cassell,\u00a0Tim Brecht, University of Waterloo<\/strong><br \/>\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/szepsei14designing.pdf\">Designing A Low-Latency Cuckoo Hash Table for Write-Intensive Workloads<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>\u00a0(<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/szepsei14designing_slides.pdf\" target=\"_self\">slides<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>)<\/li>\n<\/ul>\n<h3>Session 2<\/h3>\n<ul type=\"disc\">\n<li><strong>Martin Maas, Krste Asanovic, UC Berkeley, Tim Harris, Oracle Labs, John Kubiatowicz, UC Berkeley.<\/strong><br \/>\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/maas14case.pdf\" target=\"_self\">The Case for the Holistic Language Runtime System<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>\u00a0(<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/maas14case_slides.pdf\" target=\"_self\">slides<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>)<\/li>\n<li><strong>Jacob Nelson, Brandon Holt, Brandon Myers, Preston Briggs, Simon Kahan, Luis Ceze, Mark Oskin, University of Washington.<\/strong><br \/>\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/nelson14grappa.pdf\">Grappa: A Latency-Tolerant Runtime for Large-Scale Irregular Application<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> (<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/nelson14grappa_slides.pdf\" target=\"_self\">slides<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>)<\/li>\n<li><strong>Tim Harris, Maurice Herlihy, Yossi Lev, Oracle Labs, Yujie Liu, Lehigh University, Victor Luchangco, Virendra Marathe,\u00a0Mark Moir, Oracle Labs.<\/strong><br \/>\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/harris14towards.pdf\">Towards Whatever-Scale Abstractions for Data-Driven Parallelism<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>\u00a0(<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/harris14towards_slides.pdf\" target=\"_self\">slides<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>)<\/li>\n<\/ul>\n<h3>Session 3<\/h3>\n<ul type=\"disc\">\n<li><strong>Gustavo Alonso, ETH Zurich<\/strong><br \/>\nTitle: Rackscale &#8211; the things that matter\u00a0(<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/alonso14wrsc.pdf\" target=\"_self\">slides<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>)<br \/>\nAbstract:\u00a0Rackscale computing has become the standard for many applications running on a data center. For a variety of reasons, today it is possible to developed fully customized solution that achieve impressive performance numbers. In this talk I will argue that customization is important but needs to be sustained by general purpose techniques and components. The research agenda in the next years should focus on the latter, rather than on producing an infinity variety of high performance systems tailored for narrow use cases. Otherwise, the inevitable problems with total cost of ownership during the life cycle of a real systems (maintenance, further development, software evolution, additional functionality) will soon catch up with many existing proposals.<\/li>\n<li><strong>Tim Harris, Oracle Labs<\/strong><br \/>\nTitle: What We Talk About When We Talk About Scheduling (<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/harris14wrsc.pdf\" target=\"_self\">slides<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>)<br \/>\nAbstract:\u00a0Distributed workloads involve scheduling and resource allocation decisions at multiple levels of the stack: deciding which machines a job will run on; deciding which instances of replicated services they will use; arbitrating within those services between multiple clients; deciding which VM&#8217;s virtual CPUs get which physical cores, and which threads run on those CPUs; deciding how the instructions in those threads are scheduled in multi-threaded processors, and how they compete for resources in the cores and in the interconnect. In this talk I will illustrate the kind of interference that can occur at these different levels, and talk about some possible approaches for handling these problems.<\/li>\n<li><strong>Sanjeev Kumar, Facebook<\/strong><br \/>\nTitle: Efficiency at Scale (<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/kumar14wrsc.pdf\" target=\"_self\">slides<span class=\"sr-only\"> (opens in new tab)<\/span><\/a>)<\/li>\n<\/ul>\n<h3><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"http:\/\/clouddp14.unine.ch\/posterSession.html\" target=\"_blank\">Joint poster session<span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/h3>\n<h3>Panel session<\/h3>\n<p>Panelists:<\/p>\n<ul>\n<li><strong>Gustavo Alonso, ETH Zurich, Steve Hand, Microsoft Research, Tim Harris, Oracle Labs<\/strong><\/li>\n<li><strong>Sanjeev Kumar, Facebook, Ant Rowstron, Microsoft Research<\/strong><\/li>\n<\/ul>\n<p><span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<!-- \/wp:freeform --><!-- \/wp:msr\/content-tab --><!-- \/wp:msr\/content-tabs -->","tab-content":[{"id":0,"name":"Summary","content":"<h3>Organization<\/h3>\r\n<ul>\r\n \t<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/pcosta\/\">Paolo Costa<\/a>, Microsoft Research (Program Chair)<\/li>\r\n \t<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/dnarayan\/\">Dushyanth Narayanan<\/a>, Microsoft Research (General Chair)<\/li>\r\n<\/ul>\r\n<h3>Program Committee<\/h3>\r\n<ul>\r\n \t<li>Gustavo Alonso, ETH Zurich<\/li>\r\n \t<li>Edouard Bugnion, EPFL<\/li>\r\n \t<li>Luis Ceze, U. Washington<\/li>\r\n \t<li>Paolo Costa, Microsoft Research<\/li>\r\n \t<li>Leendert van Doorn, AMD<\/li>\r\n \t<li>Babak Falsafi, EPFL<\/li>\r\n \t<li>Blake Fitch, IBM Research<\/li>\r\n \t<li>Nathan Farrington, Facebook<\/li>\r\n \t<li>Tim Harris, Oracle Labs<\/li>\r\n \t<li>Michael Kaminsky, Intel Labs<\/li>\r\n \t<li>Dushyanth Narayanan, Microsoft Research<\/li>\r\n \t<li>Parthasarathy (Partha)\u00a0Ranganathan, Google<\/li>\r\n \t<li>Luigi Rizzo, U. Pisa<\/li>\r\n \t<li>Thomas Wenisch, U. Michigan<\/li>\r\n \t<li>Bernard Wong, U. Waterloo<\/li>\r\n<\/ul>\r\n<h3>Sponsored by<\/h3>\r\n<img class=\"alignnone size-medium wp-image-365336\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/logos2-285x300.png\" alt=\"\" width=\"285\" height=\"300\" \/>"},{"id":1,"name":"Agenda","content":"<h2>Workshop Program (<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/wrsc-poster.pdf\">CFP<\/a>)<\/h2>\r\n<h3>Session 1<\/h3>\r\n<ul type=\"disc\">\r\n \t<li><strong>Paolo Costa, Microsoft Research<\/strong>\r\nTitle: Towards Rack-scale Computing: Challenges and Opportunities (<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/costa14rackscale.pdf\" target=\"_self\">slides<\/a>)\r\nAbstract:\u00a0New hardware technology such as systems- and networks-on-chip (SOCs and NOCs), switchless network fabrics, silicon photonics, and RDMA, are redefining the landscape of data center computing, enabling interconnecting thousands of cores at high speed at the scale of today's racks. We refer to this new class of hardware as rack-scale computers because the rack is increasingly replacing the individual server as the basic building block of modern data centers. Most of the benefits promised by these new architectures, however, can only be achieved with adequate support from the software stack. In this talk, I will describe some of the research challenges and the opportunities introduced by rack-scale computing. As a concrete example, I will provide an overview of some of the research projects related to this topic that we are pursuing in our group.<\/li>\r\n \t<li><strong>Tyler Szepesi, Bernard Wong,\u00a0Ben Cassell,\u00a0Tim Brecht, University of Waterloo<\/strong>\r\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/szepsei14designing.pdf\">Designing A Low-Latency Cuckoo Hash Table for Write-Intensive Workloads<\/a>\u00a0(<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/szepsei14designing_slides.pdf\" target=\"_self\">slides<\/a>)<\/li>\r\n<\/ul>\r\n<h3>Session 2<\/h3>\r\n<ul type=\"disc\">\r\n \t<li><strong>Martin Maas, Krste Asanovic, UC Berkeley, Tim Harris, Oracle Labs, John Kubiatowicz, UC Berkeley.<\/strong>\r\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/maas14case.pdf\" target=\"_self\">The Case for the Holistic Language Runtime System<\/a>\u00a0(<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/maas14case_slides.pdf\" target=\"_self\">slides<\/a>)<\/li>\r\n \t<li><strong>Jacob Nelson, Brandon Holt, Brandon Myers, Preston Briggs, Simon Kahan, Luis Ceze, Mark Oskin, University of Washington.<\/strong>\r\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/nelson14grappa.pdf\">Grappa: A Latency-Tolerant Runtime for Large-Scale Irregular Application<\/a> (<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/nelson14grappa_slides.pdf\" target=\"_self\">slides<\/a>)<\/li>\r\n \t<li><strong>Tim Harris, Maurice Herlihy, Yossi Lev, Oracle Labs, Yujie Liu, Lehigh University, Victor Luchangco, Virendra Marathe,\u00a0Mark Moir, Oracle Labs.<\/strong>\r\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/harris14towards.pdf\">Towards Whatever-Scale Abstractions for Data-Driven Parallelism<\/a>\u00a0(<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/harris14towards_slides.pdf\" target=\"_self\">slides<\/a>)<\/li>\r\n<\/ul>\r\n<h3>Session 3<\/h3>\r\n<ul type=\"disc\">\r\n \t<li><strong>Gustavo Alonso, ETH Zurich<\/strong>\r\nTitle: Rackscale - the things that matter\u00a0(<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/alonso14wrsc.pdf\" target=\"_self\">slides<\/a>)\r\nAbstract:\u00a0Rackscale computing has become the standard for many applications running on a data center. For a variety of reasons, today it is possible to developed fully customized solution that achieve impressive performance numbers. In this talk I will argue that customization is important but needs to be sustained by general purpose techniques and components. The research agenda in the next years should focus on the latter, rather than on producing an infinity variety of high performance systems tailored for narrow use cases. Otherwise, the inevitable problems with total cost of ownership during the life cycle of a real systems (maintenance, further development, software evolution, additional functionality) will soon catch up with many existing proposals.<\/li>\r\n \t<li><strong>Tim Harris, Oracle Labs<\/strong>\r\nTitle: What We Talk About When We Talk About Scheduling (<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/harris14wrsc.pdf\" target=\"_self\">slides<\/a>)\r\nAbstract:\u00a0Distributed workloads involve scheduling and resource allocation decisions at multiple levels of the stack: deciding which machines a job will run on; deciding which instances of replicated services they will use; arbitrating within those services between multiple clients; deciding which VM's virtual CPUs get which physical cores, and which threads run on those CPUs; deciding how the instructions in those threads are scheduled in multi-threaded processors, and how they compete for resources in the cores and in the interconnect. In this talk I will illustrate the kind of interference that can occur at these different levels, and talk about some possible approaches for handling these problems.<\/li>\r\n \t<li><strong>Sanjeev Kumar, Facebook<\/strong>\r\nTitle: Efficiency at Scale (<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2017\/02\/kumar14wrsc.pdf\" target=\"_self\">slides<\/a>)<\/li>\r\n<\/ul>\r\n<h3><a href=\"http:\/\/clouddp14.unine.ch\/posterSession.html\" target=\"_self\">Joint poster session<\/a><\/h3>\r\n<h3>Panel session<\/h3>\r\nPanelists:\r\n<ul>\r\n \t<li><strong>Gustavo Alonso, ETH Zurich, Steve Hand, Microsoft Research, Tim Harris, Oracle Labs<\/strong><\/li>\r\n \t<li><strong>Sanjeev Kumar, Facebook, Ant Rowstron, Microsoft Research<\/strong><\/li>\r\n<\/ul>"}],"msr_startdate":"2014-04-13","msr_enddate":"2014-04-13","msr_event_time":"","msr_location":"","msr_event_link":"","msr_event_recording_link":"","msr_startdate_formatted":"April 13, 2014","msr_register_text":"Watch now","msr_cta_link":"","msr_cta_text":"","msr_cta_bi_name":"","featured_image_thumbnail":null,"event_excerpt":"Organization Paolo Costa, Microsoft Research (Program Chair) Dushyanth Narayanan, Microsoft Research (General Chair) Program Committee Gustavo Alonso, ETH Zurich Edouard Bugnion, EPFL Luis Ceze, U. Washington Paolo Costa, Microsoft Research Leendert van Doorn, AMD Babak Falsafi, EPFL Blake Fitch, IBM Research Nathan Farrington, Facebook Tim Harris, Oracle Labs Michael Kaminsky, Intel Labs Dushyanth Narayanan, Microsoft Research Parthasarathy (Partha)\u00a0Ranganathan, Google Luigi Rizzo, U. Pisa Thomas Wenisch, U. Michigan Bernard Wong, U. Waterloo Sponsored by","msr_research_lab":[199565],"related-researchers":[{"type":"user_nicename","value":"dnarayan","display_name":"Dushyanth Narayanan","author_link":"<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/dnarayan\/\" aria-label=\"Visit the profile page for Dushyanth Narayanan\">Dushyanth Narayanan<\/a>","is_active":false,"user_id":31659,"last_first":"Narayanan, Dushyanth","people_section":0,"alias":"dnarayan"},{"type":"user_nicename","value":"pcosta","display_name":"Paolo Costa","author_link":"<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/pcosta\/\" aria-label=\"Visit the profile page for Paolo Costa\">Paolo Costa<\/a>","is_active":false,"user_id":33218,"last_first":"Costa, Paolo","people_section":0,"alias":"pcosta"}],"msr_impact_theme":[],"related-academic-programs":[],"related-groups":[],"related-projects":[],"related-opportunities":[],"related-publications":[],"related-videos":[],"related-posts":[],"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event\/365330","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-event"}],"version-history":[{"count":2,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event\/365330\/revisions"}],"predecessor-version":[{"id":1147185,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event\/365330\/revisions\/1147185"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=365330"}],"wp:term":[{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=365330"},{"taxonomy":"msr-region","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-region?post=365330"},{"taxonomy":"msr-event-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event-type?post=365330"},{"taxonomy":"msr-video-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-video-type?post=365330"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=365330"},{"taxonomy":"msr-program-audience","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-program-audience?post=365330"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=365330"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=365330"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}