BEGIN:VCALENDAR VERSION:2.0 PRODID:Linklings LLC BEGIN:VTIMEZONE TZID:America/Chicago X-LIC-LOCATION:America/Chicago BEGIN:DAYLIGHT TZOFFSETFROM:-0600 TZOFFSETTO:-0500 TZNAME:CDT DTSTART:19700308T020000 RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU END:DAYLIGHT BEGIN:STANDARD TZOFFSETFROM:-0500 TZOFFSETTO:-0600 TZNAME:CST DTSTART:19701101T020000 RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20211207T055339Z LOCATION:Online DTSTART;TZID=America/Chicago:20211114T090000 DTEND;TZID=America/Chicago:20211114T173000 UID:submissions.supercomputing.org_SC21_sess428@linklings.com SUMMARY:FTXS: Workshop on Fault-Tolerance for HPC at Extreme Scale DESCRIPTION:Workshop\n\nCharacterizing Per-Node Memory Failures Using Benf ord’s Law\n\nFerreira, Levy\n\nFault tolerance is a key challenge as high performance computing systems continue to increase component counts, indiv idual component reliability decreases, and hardware and software complexit y increases. To better understand the potential impacts of failures on nex t-generation systems, significant e...\n\n---------------------\nAccelerat ing Checkpoint/Restart with Lossy Methods\n\nIldes, Kastoras, Keller, Baut ista Gomez\n\nApproximate computing targets applications with the ability to tolerate losses of accuracy in the computational results. The essence o f approximate computing is to use a data representation, that allows to re duce the data size or speed up computations at the cost of data accuracy. \n\nData reduction i...\n\n---------------------\nFTXS 2021 : Oren et al. Q&A\n\nFridman\n\nQ&A following " Assessing the Use Cases of Persistent Me mory in High-Performance Scientific Computing" Oren, Fridman.\n\n--------- ------------\nDoubt and Redundancy Kill Soft Errors—Toward Detection and C orrection of Silent Data Corruption in Task-Based Numerical Software\n\nSa mfass, Weinzierl, Reinarz, Bader\n\nResilient algorithms in high-performan ce computing are subject to rigorous non-functional constraints. Resilienc y must not increase the runtime, memory footprint or I/O demands too signi ficantly. We propose a task-based soft error detection scheme that relies on error criteria per task outcome. They...\n\n---------------------\nInco rporating Fault-Tolerance Awareness into System-Level Modeling and Simulat ion\n\nJohnson, Lam\n\nAs the design space for supercomputers grows, model ing and simulation (MODSIM) becomes more important to facilitate system-le vel design space exploration (DSE). Furthermore, extreme-scale systems and newer technologies can lead to higher fault rates, which negatively affec ts system performance. Ther...\n\n---------------------\nFTXS Lunch Break (12:30-2:00)\n\n\n\n---------------------\nFTXS 2021 : Ildes et al. Q&A\n\ nIldes, Kastoras\n\nQ&A following "Accelerating checkpoint/restart with lo ssy methods", Ildes, Kastoras, Keller, Bautista Gomez.\n\n---------------- -----\nFTXS Afternoon Break (3:00-3:30pm)\n\n\n\n---------------------\nFT XS Morning Break (10-10:30)\n\n\n\n---------------------\nFTXS 2021 : Miao Q&A\n\nMiao\n\nQ&A following "Relaxed Replication for Energy Efficient an d Resilient GPU Computing", Miao.\n\n---------------------\nAssessing the Use Cases of Persistent Memory in High-Performance Scientific Computing\n\ nOren, Fridman\n\nAs the High Performance Computing world moves towards th e Exa-Scale era, huge amounts of data should be analyzed, manipulated and stored. In the traditional storage/memory hierarchy, whenever the DRAM's c apacity becomes insufficient for storing data in a node, the computation s hould either be distri...\n\n---------------------\nFTXS 2021 : Johnson et al. Q&A\n\nJohnson\n\nQ&A following "Incorporating Fault-Tolerance Awaren ess into System-Level Modeling and Simulation", Johnson, Lam.\n\n--------- ------------\nRelaxed Replication for Energy Efficient and Resilient GPU C omputing\n\nMiao, Calhoun, Ge\n\nPower and reliability are two intertwine d challenges in GPU-accelerated large-scale computing. Aggressive power re duction pushes hardware to its operating limit and increases the failure r ate. Resilience allows programs to progress when subjected to faults and i s an integral component of large-scal...\n\n---------------------\nFTXS: W orkshop on Fault-Tolerance for HPC at Extreme Scale\n\nLevy, Teranishi, Da ly\n\nIncreases in the number, variety and complexity of components requir ed to compose next-generation extreme-scale systems mean that systems will experience significant increases in aggregate fault rates, fault diversit y and fault complexity. Additionally, the widespread availability of new s torage dev...\n\n---------------------\nFTXS 2021 : Featured Speaker - Dr. Catherine Schuman (Fault Tolerance and Resilience in Neuromorphic Systems )\n\nSchuman\n\nBrain-inspired neuromorphic computing systems are promisin g for the future of computing beyond the looming end of Moore's law. Thou gh robustness and resilience are commonly quoted as features of neuromorph ic computing systems, the expected performance of neuromorphic systems in the face of hardware...\n\n---------------------\nFTXS 2021 : Featured Spe aker Q&A\n\nSchuman\n\nQ&A following the featured speaker.\n\n------------ ---------\nFTXS 2021 : Opening Remarks\n\nLevy\n\nWelcome to FTXS 2021!\n\ n---------------------\nFTXS 2021 : Ferreira et al. Q&A\n\nFerreira\n\nQ&A following "Characterizing Per-node Memory Failures Using Benford’s Law", Ferreira, Levy.\n\n---------------------\nStatistical Framework for Two-Pa rty Acceptance Testing of HPC Systems for Reliability\n\nDeBardeleben, Bur r, Penton, Walker, Loncaric...\n\nHPC clusters and supercomputers are capi tal investments and undergo great scrutiny to be sure that the system meet s agreed upon metrics of performance, reliability, and usability. As such , careful evaluation of a system occurs once delivered to evaluate the agr eed upon specifications have been met....\n\n---------------------\nFTXS 2 021 : Closing Remarks\n\nLevy\n\n---------------------\nFTXS 2021 : Samfas s et al. Q&A\n\nSamfass\n\nQ&A following "Doubt and Redundancy Kill Soft E rrors—Towards Detection and Correction of Silent Data Corruption in Task-b ased Numerical Software", Samfass, Weinzierl, Reinarz, Bader.\n\n--------- ------------\nFTXS 2021 : Debardeleben et al. Q&A\n\nDebardeleben\n\nQ&A f ollowing "Statistical Framework for Two-Party Acceptance Testing of HPC Sy stems for Reliability", DeBardeleben, Burr, Penton, Walker, Loncaric, Jone s.\n\n---------------------\nFTXS 2021: Invited Speaker - Devesh Tiwari (M aking Erroneous Executions on Quantum Computers Meaningful)\n\nTiwari\n\nQ uantum computing has moved from theoretical promise to practical realizati on -- at a very rapid pace in the last decade. But, prohibitively high err or rates on existing Near-term Intermediate-Scale Quantum (NISQ) computers limit their usability even for quantum-advantage-proven algorithms (that is,...\n\n\nTag: Online Only, Extreme Scale Computing, Reliability and Res iliency\n\nRegistration Category: Workshop Reg Pass END:VEVENT END:VCALENDAR