Supercomputing Dr Virendrakumar Virendra C Bhavsar Professor Faculty

Supercomputing Dr. Virendrakumar (Virendra) C. Bhavsar Professor Faculty of Computer Science University of New Brunswick (UNB) Fredericton, Canada 1

Outline • Definitions Applications • Hardware • Software • Current Status • University of New Brunswick • Future 2

Definitions Computing Supercomputing - A supercomputer is a computer that is at the frontline of current processing capacity, particularly speed of calculation. High Performance Computing (HPC)/High Productivity Computing - supercomputing - a subset of HPC Parallel Computing - many calculations are carried out simultaneously 10**6 Million, 10**9 Billion, 10**12 Trillion 3

Human Brain 10**10 Neurons 10**4 Fan-in - Wires much slower than chips - Millions of times more volume 10**14 Inputs (Connection strngths 10**12 Connection strengths can affect processing in 5 msec Lower bound on the computational power of brain ~ 10**10 neurons, 10 spikes/sec, 10**14 connections ~10**15 operations/sec or 10**18 bits/sec 4

Connection Machine CM-2 65 K Processors, 5 CM-2 = 1. 8 x 10**13 bits/sec 10**5 times slower than brain 5

Early Computers 1950: 5, 000 operations/sec; 1970 -71: 1 Million Operations/sec

Advances in Microprocessor Technology 1974 - 1 MHz clock 1988 – 40 MHz 2002 – 2 GHz 2009 – P 4 3. 0 GHz, Quadcore 2. 66 MHz Intel Montecito chip 1. 72 Billion transistors NVidia 280 series GPU 1. 4 Billion transistors - Circuit complexity doubles every 18 months Computing power at a given cost doubles every 18 months - Processor clock rates: 40% increase/year + more instr. /cycle - DRAM Access Times: 10% increase/year caches required 7

Applications Grand Challenge Applications - cannot be solved in a reasonable amount of time with today's computers - Environment, Ecosystems, Molecular engineering, cognition, weapon design, Artificial Intelligence, (near) Real-Time Applications - Military/Defense Applications - Space -Financial Forecasting; Live data (e. g. online stock market data) 8

Applications (near) Real-Time Applications -Google - Software as a Service (Saa. S) delivery model -ATMs, online banking Data Intensive Applications -Walmart – inventory management - Data Mining 9

Applications Computational Modeling and Simulation - Science, Engineering, Social Sciences, … -Parameter sweep applications Animation and Movies 10

Applications Compute Intensive Applications Massive Data applications 11

Applications Capability Computing - Using the maximum computing power to solve a large problem in the shortest amount of time Capacity computing - Using efficient cost-effective computing power to solve - somewhat large problems - many small problems 12

Supercomputer Design Challenges Cooling Speed of Light Compute Bound Problems I/O Bound problems 13

Supercomputer Technologies Pipelining and Vector Processing Parallel and Distributed Processing Liquid Cooling Non-Uniform Memory Access Striped Disks (RAID) Parallel File System 14

Parallel and Distributed Algorithms - Intrinsic parallelism - Design of parallel algorithms - Analysis of parallel algorithms 15

Programming PVM and MPI – Loosely connected clusters Open. MP for Shared Memory Machines 16

Automatic Parallelization Compilers Limited success Application Checkpointing 17

Current Supercomputer Roadrunner applications - National Security - Planet: Earth and Environmental Sciences e. g. ground water modeling - Health: Biology, Chemistry, Life Sciences - Science: Engineering, Technology - Universe: Astronomy, Space, Astrophysics -- Modeling the decay of the US nuclear arsenal 18

Current Supercomputer Roadrunner Los Alamos National Laboratory, Los Alamos, NM, USA - >1 Petaflop (Quadrilion): million billion (10**15) floating-point operations/sec (FLOPS) -1. 71 Petaflop peak - Weight - 500, 000 pounds - Power - 4 Mega Watt - Space – 6000 square feet - Cabling 57 miles - 19

Current Supercomputer Roadrunner (Installation Year – 2008) Los Alamos National Lab, USA ~ 3, 250 compute nodes -Compute Node: Two AMD Opteron dual-core microprocessors - Each of the Opteron core: Internally attached to one of four enhanced Cell microprocessors. - Enhanced Cell: double-precision arithmetic faster and can access more memory than can the original Cell in a Play. Station 3. The entire machine will have almost 13, 000 Cells and half as many dual-core Opterons. - Interconnection Network: off-the-shelf Infiniband 20

Current Supercomputer Roadrunner (Installation Year – 2008) DOE/NNSA/LANL System Family - IBM Cluster System Model - Blade. Center QS 22 Cluster Computer - Blade. Center QS 22/LS 21 Cluster, Power. XCell 8 i 3. 2 Ghz / Opteron DC 1. 8 GHz , Voltaire Infiniband Operating System - Linux Interconnect – Infiniband Processor - Power. XCell 8 i 3200 MHz (12. 8 GFlops) 21

Hardware: Building Blocks • Building blocks – processors, memory, interconnection networks • Processors • Memory – main and secondary storage • Interconnection networks 22

Hardware: Architectures • Taxonomy: SISD, SIMD, MISD and MIMD • Shared Memory Processing versus Distributed Memory Processing Symmetric Multi-Processing (SMP) versus Non. Uniform Memory Access (NUMA) • Processors • Clusters • 23

Special Purpose Supercomputers • Specially Programmed FPGA chips • Custom VLSI Chips • Reconfigurable Computing • GPUs (Graphics Processing Units) 24

University of New Brunswick 25

High Performance Computing and Networking @ University of New Brunswick

ACEnet: Atlantic Computational Excellence Network “People, Research, Excellence” Hosting sites: Member sites:

ACEnet l Atlantic Canada is a distributed environment l $30 million initiative l Waterways make networking solutions difficult (e. g. Cabot Strait)

ACEnet l World-class HPC facilities l Behave as a single, regionally distributed “computational power grid” l Create and operate sophisticated collaboration facilities to bind together geographically dispersed research communities.

Advaced Computational Research Lab (ACRL) Infrastructure

Major Users UNB Biology Gary Saunders UNB Chemistry Scott Brownridge Larry Calhoun Ghislain Deslongchamps Friedrich Grein UNB Computer Science Eric Aubanel Virendra Bhavsar Brad Nickerson Ruth Shaw UNB Text Processing Centre Alan Burk David Gants UNB Geodesy Petr Vanícek Richard Langley UNB Mathematics Keith De’Bell Abraham Punnen UNB Mechanical Engineering Mohammad Bagher Ayani David Bonham Andrew Gerber Marwan Hassan Esam Hussein UNB Physics Dr. Eugene K Ho Dr. Zong-Chao Yan Dr. Li-Hong Xu UNB Forestry Evelyn Richards UNB Biomedical Kevin Englehart DAL Physics Andrew Rutenberg MTA Chemistry Stacey Wetmore MUN Computer Science Dwight Kuo Sick Kids Hospital, Toronto Regis Pomes Ching-Hsing Yu Len Zaifman St. FX Computer Science Laurence Yang Uof. Calgary Computer Science Peter Tieleman Justin Mac. Callum Ude. M Environmental Studies Yves Gagnon Ude. M Computer Science Jalal Almhana UPEI Physics Sheldon Opps James Polson Uof. T Computer Science Hue Sun Chan Maria Sabaye Moghaddam

ACEnet at UNB Fundy: SUN cluster, AMD Opeteron, 632 cores ACEnet: 3324 cores Internet connectivity > 2 Gbps at UNB

Collaboration Grid l Collaboration gear across Atlantic Canada l l Lecture rooms equipped so ACEnet sites can share seminars and participate remotely ACEnet cafés at each site sharing continuous video feeds Desktop level collaboration equipment for personal communication Access Grid streams tens to hundreds of Mbps across the CANARIE network ACEnet

My Research Work l l l l Special Purpose computers for Military Applications Design and development of MICRON and PLEXUS Parallel Monte Carlo Algorithms Graphics and Visualization Pa. Grid Artificial Intelligence – artificial neural networks, e. Business Bioinformatics – Canadian Potato Genome project

Future l l l l IBM Cyclops 64 – supercomputer on a chip C-DAC initiative for 2010 –petaflop machine NCSA, USA 2011 petaflop machine NASA, SGI and Intel Pleiades – 10 petaflop by 2012 1 Exaflop (10**18 flops) by 2019 Human brain neural simulations – 10 exaflop by 2025 2 -week Full Weather modeling – 1 zeta flops (10**21 flops) by 2030