BEGIN:VCALENDAR VERSION:2.0 PRODID:Linklings LLC BEGIN:VTIMEZONE TZID:America/New_York X-LIC-LOCATION:America/New_York BEGIN:DAYLIGHT TZOFFSETFROM:-0500 TZOFFSETTO:-0400 TZNAME:EDT DTSTART:19700308T020000 RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU END:DAYLIGHT BEGIN:STANDARD TZOFFSETFROM:-0400 TZOFFSETTO:-0500 TZNAME:EST DTSTART:19701101T020000 RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20210402T160556Z LOCATION:Track 5 DTSTART;TZID=America/New_York:20201112T100000 DTEND;TZID=America/New_York:20201112T164000 UID:submissions.supercomputing.org_SC20_sess211@linklings.com SUMMARY:Fifth International Parallel Data Systems Workshop DESCRIPTION:Workshop\n\nI/O Traces of HPC Applications\n\nWang, Mohror, Sn ir\n\n---------------------\nFifth International Parallel Data Systems Wor kshop – Introduction\n\nCarns, Ibrahim, Sato, Lofstead\n\nEfficient data s torage and data management are crucial to scientific productivity in both traditional simulation-oriented HPC environments and Big Data analysis env ironments. This issue is further exacerbated by the growing volume of expe rimental and observational data, the widening gap between the ...\n\n----- ----------------\nFifth International Parallel Data Systems Workshop – Bre ak\n\n\n\n---------------------\nScalable Communication and Data Persisten ce Layer for NVM-Based Storage Systems\n\nOhtsuji, Yoshida, Okamoto, Hayas hi\n\n---------------------\nFifth International Parallel Data Systems Wor kshop – Closing Remarks\n\n\n\n---------------------\nFifth International Parallel Data Systems Workshop – Break\n\n\n\n---------------------\nPDSW 2020 - Keynote: "Sink or Swim: How Not to Drown in Colossal Streams of Dat a?"\n\nAgrawal\n\n---------------------\nFifth International Parallel Data Systems Workshop – Break\n\n\n\n---------------------\nDeriving Storage I nsights from the IO500\n\nLogan, Lofstead, Kougkas, Sun\n\n--------------- ------\nQ&A with Work-in-Progress Speakers\n\n\n\n---------------------\nK eeping It Real: Why HPC Data Services Don't Achieve I/O Microbenchmark Per formance\n\nCarns, Harms, Settlemyer, Atkinson, Ross\n\nHPC storage softwa re developers rely on benchmarks as reference points for performance evalu ation. Low-level synthetic microbenchmarks are particularly valuable for i solating performance bottlenecks in complex systems and identifying optimi zation opportunities.\n\nThe use of low-level microbenchmarks ...\n\n----- ----------------\nToward On-Demand I/O Forwarding in HPC Platforms\n\nBez, Zanon Boito, Miranda, Nou, Cortes...\n\nI/O forwarding is an established and widely-adopted technique in HPC to reduce contention and improve I/O p erformance in the access to shared storage infrastructure. On such machine s, this layer is often physically deployed on dedicated nodes, and their c onnection to the clients is static. Furthermo...\n\n---------------------\ nGPU Direct I/O with HDF5\n\nRavi, Byna, Koziol\n\nExascale HPC systems ar e being designed with accelerators, such as GPUs, to accelerate parts o f applications. In machine learning workloads as well as large-scale simu lations that use GPUs as accelerators, the CPU (or host) memory is current ly used as a buffer for data transfers between GPU (or ...\n\n------------ ---------\nFingerprinting the Checker Policies of Parallel File Systems\n\ nHan, Zhang, Zheng\n\nParallel file systems (PFSes) play an essential role in high performance computing. To ensure the integrity, many PFSes are de signed with a checker component, which serves as the last line of defense to bring a corrupted PFS back to a healthy state. Motivated by real-world incidents of PFS corruptio...\n\n---------------------\nFifth Internationa l Parallel Data Systems Workshop – Break\n\n\n\n---------------------\nPan geo Benchmarking Analysis: Object Storage vs. POSIX File System\n\nXu, Pau l, Banihirwe\n\nPangeo is a community of scientists and software developer s collaborating to enable Big Data Geoscience analysis interactively in th e public cloud and on high-performance computing (HPC) systems. At the co re of the Pangeo software stack is (1) Xarray, which adds labels to metada ta such as dimension...\n\n---------------------\nFractional-Overlap Declu stered Parity: Evaluating Reliability for Storage Systems\n\nKe, Gunawi, M anno, Bonnie, Settlemyer\n\nIn this paper, we propose a flexible an d practical data protection scheme, fractional-overlap declustered p arity( FODP), to explore the trade-offs between fault tolerance and rebuil d performance. Our experiments show that FODP is able to bring forth up to 99% less probability of...\n\n---------------------\nGauge: An In teractive Data-Driven Visualization Tool for HPC Application I/O Performan ce Analysis\n\ndel Rosario, Currier, Isakov, Madireddy, Balaprakash...\n\n Understanding and alleviating I/O bottlenecks in HPC system workloads is d ifficult due to the complex, multi-layered nature of HPC I/O subsystems. E ven with full visibility into the jobs executed on the system, the lack of tooling makes debugging I/O problems difficult. In this work, we introduc e Ga...\n\n---------------------\nEmulating I/O Behavior in Scientific Wor kflows on High Performance Computing Systems\n\nChowdhury, Zhu, Di Natale, Moody, Gonsiorowski...\n\nScientific application workflows leverage the c apabilities of cutting-edge high-performance computing (HPC) facilities to enable complex applications for academia, research, and industry communit ies. Data transfer and I/O dependency among different modules of modern HP C workflows can increase the co...\n\n\nTag: Big Data, Data Analytics, Com pression, and Management, Data Movement, File Systems and I/O, Storage\n\n Registration Category: Workshop Reg Pass END:VEVENT END:VCALENDAR