Issue title: SC13 – The International Conference for High Performance Computing, Networking, Storage and Analysis
Article type: Research Article
Authors: Jung, Myoungsoo; | Wilson III, Ellis H. | Choi, Wonil; | Shalf, John; | Aktulga, Hasan Metin | Yang, Chao | Saule, Erik | Catalyurek, Umit V.; | Kandemir, Mahmut
Affiliations: Department of Electrical Engineering, The University of Texas at Dallas, Richardson, TX, USA. E-mails: {Jung, wonil.choi}@utdallas.edu | Department of Computer Science and Engineering, The Pennsylvania State University, University Park, PA, USA. E-mails: {ellis, kandemir}@cse.psu.edu | Computational Research Division, Lawrence Berkeley National Laboratory, Berkeley, CA, USA. E-mails: {jshalf, hmaktulga, cyang}@lbl.gov | National Energy Research Scientific Computing Center, Lawrence Berkeley National Laboratory, Berkeley, CA, USA | Biomedical Informatics, The Ohio State University, Columbus, OH, USA. E-mails: {esaule, umit}@bmi.osu.edu | Electrical and Computer Engineering, The Ohio State University, Columbus, OH, USA
Note: [] Corresponding author. E-mail: [email protected]
Abstract: Drawing parallels to the rise of general purpose graphical processing units (GPGPUs) as accelerators for specific high-performance computing (HPC) workloads, there is a rise in the use of non-volatile memory (NVM) as accelerators for I/O-intensive scientific applications. However, existing works have explored use of NVM within dedicated I/O nodes, which are distant from the compute nodes that actually need such acceleration. As NVM bandwidth begins to out-pace point-to-point network capacity, we argue for the need to break from the archetype of completely separated storage. Therefore, in this work we investigate co-location of NVM and compute by varying I/O interfaces, file systems, types of NVM, and both current and future SSD architectures, uncovering numerous bottlenecks implicit in these various levels in the I/O stack. We present novel hardware and software solutions, including the new Unified File System (UFS), to enable fuller utilization of the new compute-local NVM storage. Our experimental evaluation, which employs a real-world Out-of-Core (OoC) HPC application, demonstrates throughput increases in excess of an order of magnitude over current approaches.
Keywords: High performance computing, scientific application, non-volatile memory, solid state disk, out-of-core computing
DOI: 10.3233/SPR-140384
Journal: Scientific Programming, vol. 22, no. 2, pp. 125-139, 2014
