Issue link: http://osercommunicationsgroup.uberflip.com/i/412884
S u p e r C o m p u te r S h o w D a i l y Tu e s d a y, N o ve m b e r 1 8 , 2 0 1 4 6 ACM GORDON BELL PRIZE PAPERS AND AWARD PRESENTATION Hear ACM Gordon Bell Prize finalist paper "Petascale High Order Dynamic Rupture Earthquake Simulations on Heterogeneous Supercomputers" at 1:30 p.m. on Tuesday, November 18 in the New Orleans Theater. The paper's authors are Alexander Heinecke, Alexander Breuer, Sebastian Rettenberger, Michael Bader, Alice- Agnes Gabriel, Christian Pelties, Arndt Bode, William Barth, Xiang-Ke Liao, Karthikeyan Vaidyanathan, Mikhail Smelyanskiy and Pradeep Dubey. The paper will present an end-to-end optimization of the innovative Arbitrary high-order DERivative Discontinuous Galerkin (ADER-DG) software SeisSol targeting Intel ® Xeon Phi™ coprocessor platforms, achieving unprecedented earthquake model complexity through coupled simulation of full frictional slid- ing and seismic wave propagation. Following that presentation at 2 p.m., also in the New Orleans Theater, hear ACM Gordon Bell Prize finalist paper "Physics-Based Urban Earthquake Simulation Enhanced by 10.7 BlnDOF x 30 K Time-Step Unstructured FE Non- Linear Seismic Wave Simulation" by authors Tsuyoshi Ichimura, Kohei Fujita, Seizo Tanaka, Muneo Hori, Maddegedara Lalith, Yoshihisa Shizawa and Hiroshi Kobayashi. In this paper, the authors present methods for improving the reliability of urban earthquake response analyses. They posit that such analyses can change the quality of disas- ter estimations and are expected to become standard in the future. Presentations by the ACM Gordon Bell Prize finalists continue at 2:30 with "Real-Time Scalable Cortical Computing at 46 Gig Synaptic Ops/Watt with ~100x Speedup in Time to Solution and ~100,000x Reduction in Energy-to- Solution." This paper's authors are Andrew S. Cassidy, Rodrigo Alvarez-Icaza, Filipp Akopyan, Jun Sawada, John V. Arthur, Paul A. Merolla, Pallab Datta, Marc Gonzalez Tallada, Brian Taba, Alexander Andreopoulos, Arnon Amir, Steven K. Esser, Jeff Kusnitz, Rathinakumar Appuswamy, Chuck Haymes, Bernard Brezzo, Roger Moussalli, Ralph Bellofatto, Christian Baks, Michael Mastro, Kai Schleupen, Charles E. Cox, Ken Inoue, Steve Millman, Nabil Imam, Emmett McQuinn, Yutaka T. Nakamura, Ivan Vo, Chen Guo, Don Nguyen, Scott Lekuch, Sameh Assad, Daniel Friedman, Bryan L. Jackson, Myron D. Flickner, William P. Risk, Rajit Manohar and Dharmendra S. Modha. They have developed a parallel, event-driven kernel for neurosynaptic com- putation that is efficient with respect to computation, memory and communication, and the paper demonstrates TrueNorth, a co-designed silicon expression of the ker- nel that enables a broad range of cognitive applications. Talks by the ACM Gordon Bell Prize finalists continue on Wednesday, November 19 at 1:30 p.m. in the New Orleans Theater with "Anton 2: Raising the Bar for Performance and Programmability in a Special-Purpose Molecular Dynamics Supercomputer" by authors David E. Shaw, J.P. Grossman, Joseph A. Bank, Brannon Batson, J. Adam Butts, Jack C. Chao, Martin M. Deneroff, Ron O. Dror, Amos Even, Christopher H. Fenton, Anthony Forte, Joseph Gagliardo, Gennette Gill, Brian Greskamp, C. Richard Ho, Douglas J. Ierardi, Lev Iserovich, Jeffrey S. Kuskin, Richard H. Larson, Timothy Layman, Li-Siang Lee, Adam K. Lerer, Chester Li, Daniel Killebrew, Kenneth M. Mackenzie, Shark Yeuk-Hai Mok, Mark A. Moraes, Rolf Mueller, Lawrence J. Nociolo, Jon L. Peticolas, Terry Quan, Daniel Ramot, John K. Salmon, Daniele P. Scarpazza, U. Ben Schafer, Naseer Siddique, Christopher W. Snyder, Jochen Spengler, Ping Tak Peter Tang, Michael Theobald, Horia Toma, Brian Towles, Benjamin Vitale, Stanley C. Wang and Cliff Young. Anton 2 is a sec- ond-generation special-purpose supercom- puter for molecular dynamics simulations with architecture that's tailored for fine- grained event-driven operation, improving performance over Anton 1 by increasing the overlap of computation with communi- cation and allowing a wider range of algo- rithms to run efficiently. The paper's authors are all from D.E. Shaw Research and Subhash Saini from the NASA Ames Research Center is chairing the session. The final paper from the ACM Gordon Bell Prize finalists will be offered at 2 p.m. with "24.77 Pflops on a Gravitational Tree-Code to Simulate the Milky Way Galaxy with 18600 GPUs." Authors Jeroen Bédorf, Evghenii Gaburov, Michiko S. Fujii, Keigo Nitadori, Tomoaki Ishiyama and Simon Portegies Zwart reached for the stars to demonstrate that they have simulated, for the first time, the long term evolution of the Milky Way Galaxy using 51 billion particles on the Swiss Piz Daint supercomputer with their N-body gravitational tree-code Bonsai. The ACM Gordon Bell Prize for 2014 will be presented on Thursday, November 20 as part of the SC Conference Awards Presentation beginning at 12:30 p.m. in the New Orleans Theater. Other awards to be presented include the SC14 Test of Time and the IEEE Technical Committee on Scalable Computing (TCSC) Young Investigators in Scalable Computing Awards; the SC14 Best Paper, Best Student Paper, and Best Poster Awards; George Michael Memorial HPC Ph.D. Fellowship; ACM Student Research Competition; and the Student Cluster Competition awards. This year, the recipi- ent of the inaugural Best Visualization Award will also be recognized. Presenters are Barbara Chapman from the University of Houston and John Grosh from the Lawrence Livermore National Laboratory. BHYVE, THE BSD HYPERVISOR EMBRACES MODERN HARDWARE Virtualized infrastructure is here to stay, and FreeBSD 10's bhyve Hypervisor, pro- nounced "beehive," is the first open source hypervisor to rely exclusively on modern CPU virtualization features and virtualized device I/O. Unlike the Xen, KVM, Hyper-V and VMware hypervi- sors, bhyve forgoes emulation at all levels in favor of leveraging the Extended Page Table found on modern CPUs and VirtIO virtualized network and storage devices. The Extended Page Table, or EPT, replaces costly software shadow page table management with performant hard- ware assistance, resulting in a smaller, faster hypervisor. Virtualized Input/Output or VirtIO is a standard for hypervisor-native network and block stor- age devices that is supported by most modern operating systems. Combined, EPT and VirtIO allow bhyve to be small in size, yet support OpenBSD, NetBSD and GNU/Linux virtual machines at hardware- native speeds. Microsoft Windows VM support for bhyve is under development. As an integral part of FreeBSD, bhyve can also leverage advanced FreeBSD features such as the OpenZFS filesystem. OpenZFS is an advanced copy-on-write filesystem with integrated checksumming, snapshotting and redun- dancy that eliminates all of the shortcom- ings of hardware RAID devices. These shortcomings include "write holes," cache inconsistency, device portability and obfuscated disk health monitoring. OpenZFS offers the highest guarantee of data consis- tency available in any open source filesystem, and makes an excellent back- ing store for UFS, ext2 and NTFS storage devices. Unlike ZFS on Linux, FreeBSD's OpenZFS is up to date with Illumos OpenZFS and shares many core developers. Other FreeBSD features that compli- ment bhyve include the Jails lightweight system container, the Capsicum security framework, the PF network packet filter, the Clang/LLVM compiler and a unified user and developer experience. With the addition of bhyve, FreeBSD becomes a solid and flexible platform for building large-capacity virtualized environments and HPC applications. bhyve is also permis- sively-licensed under the terms of the familiar BSD two-clause copy- right. Born in a Fortune 500 com- pany, bhyve quickly transitioned to the open source community with no legal or political strings attached, and is now a native FreeBSD technology. This fact marks a historic departure from the norm in which advanced open source technologies are either "open core" or bolted on to existing systems. The result is a seamless experience from a user, developer and legal perspective. Visit booth #3461 for a demonstration of bhyve and other FreeBSD advanced tech- nologies. Learn more at www.bhyve.org or wiki.freebsd.org/bhyve.