big data everywhere chicago: high performance computing - contributions towards big data (hpc)
DESCRIPTION
By Sharan KalwaniTRANSCRIPT
![Page 2: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/2.jpg)
Outline
o History of Supercomputing *
o Technologies
o Modern Day HPC:
o Current State of the Art
o Peering beyond the Horizon:
o Next Set of technologies
* aka High Performance Computing
(HPC)
![Page 3: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/3.jpg)
History
Computing Demand:
driven by needs far beyond contemporary capability
Early adopters: (1970s)
LANL (Los Alamos National Lab) and
NCAR (National Center for Atmospheric Research)
Characteristics: domain specific needs
Features: High Speed Calculations: PDE, Matrices
1972: Seymour Cray (CDC, Cray Research Inc.)
1st Model Cray-1
![Page 4: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/4.jpg)
History
Cray-1 Characteristics: (1975-1976)
64 bit word length
12.5 nanosecond clock speed
80 MHz
“original” RISC
1 clock == 1 instruction
Vector instruction set, true multiplier effect, single instructions, multiple data
Matrix operations, pipelining,
included add+multiply!
memory <> processor balance
Cray-1, Cray-XMP, Cray-YMP, Cray-2, Cray 3
![Page 5: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/5.jpg)
History
Enter the domain of MPP
Massively parallel processors
Introduction of Torus architectures
Seen these days in some offerings
Cray T3 D/E/F….(1st machine to break
1,000,000,000 calculations/sec barrier)
![Page 6: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/6.jpg)
Cray T3 architecture (logical) circa 1993, looks a lot like a cluster, eh?
![Page 7: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/7.jpg)
Hardware Contributions: _Phase_ 1
Profusion of technologies: RISC inspiration (1 clock cycle → 1 instruction)
Solid State Disk – recognized the need for keeping CPU busy all the time
multi-core software – coherence + synchronization
De-coupling of I/O from compute
Massive memory
I/O technologies – HiPPI (high speed parallel interface)
Visualization driver
Chip set design, ECL -> CMOS integration
Parallel processing software foundation -> MPI
![Page 8: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/8.jpg)
Solid State Disk (grandpa USB stick)
The first CRAY X-MP system had SSD in 1982.
Designed for nearly immediate reading and writing of very large data files.
Data transfer rates of up to 1.250 GBytes / second,
Far exceeding *any* other data transfer devices in its time.
SSDs offered in sizes of 64, 128, or 256 million bytes of storage.
The hole in the cabinet was to attach a very high speed (VHISP)
data channel to an SSD.
Link referred to as the "skyway."
Via software, the SSD is logically accessed as a disk unit.
SSD driver ~ 1200 lines of C code!
![Page 9: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/9.jpg)
History Marches on….
Battle of technologies:
Silicon v. Gallium Arsenide
Vector v. Killer Micros
Accelerated Strategic Computing Initiative (mid 90s) ASCI project changed directions for everyone
![Page 10: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/10.jpg)
Speed
Everybody was focused on clock speed
Or Floating Point Operations (FLOPS/sec)
@12.5 ns clock speed 80 million Flops/sec (peak)
Leading to the famous Macho FLOPS race: USA v Japan (90s)
Megaflops Gigaflops (1000 MF)
Gigaflops Teraflops (1000 GF)
Teraflops Petaflops (1000 TF)
In 2018 the industry expects an ExaFlop machine!
![Page 11: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/11.jpg)
Speed
First GigaFlop/sec* System
Cray YMP and Cray-2
First TeraFlop/sec System
Sandia National Lab ASCI “Red” (Intel)
First PetaFlop/sec System
LANL “RoadRunner” (IBM)
First ExaFlop/sec System
???
* SUSTAINED!
![Page 12: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/12.jpg)
Is anyone keeping score?
Birth of the Top500 list
1993 – Dongarra, Strohmaier , Meuer & Simon
Linear Algebra Package (LINPAK) basis
Offshoots:
Green500 (power efficiency)
Graph500 (search oriented – little to no floating point computation)
@SC13 a new replacement metric has been proposed
![Page 13: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/13.jpg)
Is anyone keeping score? We will return to this …..
![Page 14: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/14.jpg)
Cluster Growth propelled by HPC
![Page 15: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/15.jpg)
Linux Growth propelled by HPC
![Page 16: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/16.jpg)
Track Record of Linux versions in HPC
• See also Linux Foundation report
• http://www.linuxfoundation.org/publications/linux-foundation/top500_report_2013
![Page 17: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/17.jpg)
HPC top500 - factoids • Current #1 system has 3,120,000 cores
– Located in China, called Tianhe-2 “Milky Way”
– Peak speed of 33.9 PetaFLops/second (quadrillions of calculations per second)
– Needs 17.8 MW of power
• Current #2 system @ ORNL (US Government DoE) in Tennessee
– Has 560,640 cores, called Titan
– Peak speed of 17.6 PetaFlops/second
– Needs 8.21 MW of power
![Page 18: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/18.jpg)
HPC top500 - factoids • Tianhe-2
![Page 19: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/19.jpg)
HPC top500 - factoids • Titan
(http://www.ornl.gov/info/press_releases/get_press_release.cfm?ReleaseNumber=mr20121029-00)
•
![Page 20: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/20.jpg)
HPC top500 - factoids • Titan
![Page 21: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/21.jpg)
Treemap of countries in HPC
![Page 22: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/22.jpg)
Operating Systems: History
Early HPC OS:
tiny assembled loaders
CTSS (Cray Time Sharing Systems) - LTSS
CRAY Operating Systems (COS)
CRAY UNIX ((UNICOS)
mk/Kernel – CHORUS
Beowulf cluster – Linux appears
![Page 23: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/23.jpg)
Linux Contributions: History
Linux – attack of the killer micros, 1992
![Page 24: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/24.jpg)
Linux Contributions: History
NOW – Network of workstations, 1993
![Page 25: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/25.jpg)
Linux Contributions: History 1993-1994
133 nodes – Stone Supercomputer
First Beowulf cluster
Concept pioneered at NASA/Caltech
Thomas Sterling and Donald Becker
![Page 26: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/26.jpg)
Linux Contributions: History 1993-1994
• Beowulf
![Page 27: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/27.jpg)
Linux Contributions: History 1993-1994
• Beowulf
![Page 28: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/28.jpg)
Linux Contributions: History 1993-1994
• Beowulf
• NASA
• LSU
• Indiana University
THOMAS STERLING
![Page 29: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/29.jpg)
Linux Contributions: History
• Beowulf Components:
– Parallel Virtual Machine (PVM) – U Tennesse
– Message Passing Interface (MPI) – several folks
– Jack Dongarra,Tony Hey and David Walker
– Support of NSF and ARPA
– Today we have the MPI Forum
– MPI 2 and now MPI 3
– OpenMPI, MPICH, etc
– Future Pthreads and OpenMP,
![Page 30: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/30.jpg)
HPC and Linux
• Beowulf Attributes (or cluster features):
• Open Source
• Low Cost
• Elasticity
• Equal Spread of work (seeds of cloud computing here!!)
• These days the Linux kernel can handle 64 cores! HPC pushes this limit even further….
![Page 31: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/31.jpg)
HPC and Linux pushing the boundaries
• File systems:
– Large number of high performance file systems
– Lustre, now in version 2.5
– Beats HDFS several times over!!
– You can host HDFS over many HPC filesystems for massive gains
![Page 32: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/32.jpg)
Typical Stack
Pick Distro – Linux based (usually Enterprise class)
Hardware
![Page 33: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/33.jpg)
Hardware Contributions: _Phase_ 2
Profusion of technologies:
In-memory processing, many HPC sites implemented this
1992 built special systems for the use in cryptography using these technigques
Graph traversal systems – now available as appliances by HPC vendors
Massive memory : single memory systems over several TB in size
Infiniband interconnects: hitting 100 Gbits/sec switches you can buy them now
Parallel processing software foundation -> replacements for MPI stack being worked on
![Page 34: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/34.jpg)
Modern Day HPC
• Building the ExaScale machine:
– Exascale is 1 quintillion calculations/second
– 1000x Petaflops/sec
– Also Known as 10^18 (hence 2018 projections)
–1, 000, 000, 000, 000, 000, 000 floating point calculations/second (sustained)
–How to feed this monster?
![Page 35: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/35.jpg)
Modern Day HPC
• Solutions for the ExaScale monster:
• Inevitably Big Data community should watch/support/benefit issues we are tackling now:
– Memory matters!
– Resiliency in software
– Robustness in hardware
– Co-Design critical
–Power Consumption and Cooling (estimate several megawatts w/ present day approaches)
–Utterly new architectures needed
![Page 36: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/36.jpg)
Applications: What did we use all this for?
Weather
Automotive and Aerospace Design
Traditional Sciences
Energy (Nuclear, Oil & Gas)
Bioinformatics
Cryptography
and……big data
![Page 37: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/37.jpg)
Applications: traditional HPC
Automotive & (similar) Aerospace Design
o Car Crash Analysis – prime usage, 50%
o Each physical crash test costs $0.5 million
o Virtual Prototype test - $1000 (or less)
![Page 38: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/38.jpg)
Applications: The real deal vs. HPC
• NHTSA requires physical validation
• Before total crash tests cost a total of $100 million/year
• Limited to a small suite: 12 tests
• Today we can do over 140+ different tests (for each vehicle) and with:
– Less cost (we instead increased the # of tests!)
– Faster response (5 years v 12 months)
– Many more design iterations (Hundreds v 10)
![Page 39: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/39.jpg)
HPC for weather forecasting
![Page 40: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/40.jpg)
Whither HPC and the Cloud?
![Page 41: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/41.jpg)
HPC for Crisis Assist
![Page 42: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/42.jpg)
Technology March! Or why simulation matters?
• Increasing resolving power - greater fidelity problem
• Decreasing product design turnaround times
• Increase cost-effectiveness relative to experiment & observation
• Reducing uncertainty
• Ramping up the ease of use by non-experts
• Powerful tool in resolving scientific questions, engineering designs, and policy support
• Co-execution environments for simulation, large-scale data enabled science, and scientific visualization
• Simple: Better Answers thus delivering an…..
– Attractiveness to the creative and entrepreneurial classes
– Straightforward case for national economic competitiveness!!!
![Page 43: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/43.jpg)
We need more HPC because….
![Page 44: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/44.jpg)
What about costs????
![Page 45: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/45.jpg)
What about costs????
![Page 46: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/46.jpg)
What about costs????
![Page 47: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/47.jpg)
HPC is indispensable!
• Establish Capability
• Enable Adding of Complexity
• Gain a real and better Understanding
• And do not forget all that data!
• How do we tie it in?......
![Page 48: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/48.jpg)
Approaching the eXtreme Scale
• Current paradigm: Simulation lots of equations which mimic or model actual situations “Third” Paradigm
• -------------------------------------------------------------------------------------------
• Operate without models (Big Data) “Fourth” Paradigm
![Page 49: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/49.jpg)
Operate without models (Big Data)
• BEFORE….. * NOW/FUTURE….
Models/Theory Models/Theory
DATA
![Page 50: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/50.jpg)
Best Example….. tell us what we do not know!
• Recent Success:
• Solar observations (actual data)
• Unknown Surface Perturbations or Energy
• Could not be explained by all classical models
• Resorted to automated machine learning driven alternate search
• Answer: Solar Earthquakes and Thunderclaps, classic acoustic signature!
• New profession: Solar Seismologists!
![Page 51: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/51.jpg)
Trend began a decade+ ago…. http://research.microsoft.com/en-us/collaboration/fourthparadigm/4th_paradigm_book_complete_lr.pdf
![Page 52: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/52.jpg)
Everyone is seriously interested….. http://science.energy.gov/~/media/ascr/ascac/pdf/reports/exascale_subcommittee_report.pdf.
![Page 53: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/53.jpg)
A Peek at the Future…….
• Yes…we should definitely care
• The bigger and more relevant questions are:
– What architecture? What programming model?
– Power consumption will dominate
• Currently 4 approaches:
– Stay the course ? Not!!
– All GPGPU based?
– ARM based?
– Quantum Computing ??
![Page 54: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/54.jpg)
GPGPU perspective….
G1 G2 G3 G4
8-Cores 8-Cores 16-Core Server Node
Multi-GPU Acceleration of
a 16-Core ANSYS Fluent
Simulation of External Aero
Xeon E5-2667 CPUs + Tesla K20X GPUs
2.9X Solver Speedup
CPU Configuration CPU + GPU Configuration
Click to Launch Movie
![Page 55: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/55.jpg)
A Peek at the Future……. – GPGPU
G
DD
R
GD
DR
DDR
DDR
GPU I/O Hub PCI-Express
CPU
Cache
1
2
3
![Page 56: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/56.jpg)
A Peek at the Future……. – GPGPU based?
– http://www.anl.gov/events/overview-nvidia-exascale-processor-architecture-co-design-philosophy-and-application-results
– Echelon
– DragonFly
– http://www.nvidia.com/content/PDF/sc_2010/theater/Dally_SC10.pdf
![Page 57: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/57.jpg)
A Peek at the Future…….
![Page 58: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/58.jpg)
A Peek at the Future……. – ARM or Pi based?
– http://coen.boisestate.edu/ece/files/2013/05/Creating.a.Raspberry.Pi-Based.Beowulf.Cluster_v2.pdf
![Page 59: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/59.jpg)
Quantum Computing…….
• D-WAVE systems installed at NASA Ames lab
• Uses a special chip, 512 bit “Vesuvius”
• Uses 12 KW of power
• Cooled to 0.02 Degrees K (100 times colder than outer space)
• RF shielding
![Page 60: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/60.jpg)
Quantum Computing…….
• Shor’s Integer Factorization Algorithm
• Problem: Given a composite n-bit integer, find a nontrivial factor.
– Best-known deterministic algorithm on a classical computer has time complexity exp(O( n1/3log2/3 n)).
• A quantum computer can solve this problem in O( n3 ) operations.
Peter Shor
Algorithms for Quantum Computation: Discrete Logarithms and Factoring
Proc. 35thAnnual Symposium on Foundations of Computer Science, 1994, pp. 124-134
![Page 61: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/61.jpg)
Quantum Computing…….
• Classical: number field sieve
– Time complexity: exp(O(n1/3 log2/3 n))
– Time for 512-bit number: 8400 MIPS years
– Time for 1024-bit number: 1.6 billion times longer
• Quantum: Shor’s algorithm
– Time complexity: O(n3)
– Time for 512-bit number: 3.5 hours
– Time for 1024-bit number: 31 hours
• (assuming a 1 GHz quantum machine)
See M. Oskin, F. Chong, I. Chuang
A Practical Architecture for Reliable Quantum Computers
IEEE Computer, 2002, pp. 79-87
![Page 62: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/62.jpg)
What I will be looking into…….
• Julia
• Programming Environment which combines *all* the elements of: – R (express data handling)
– Scientific and Engineering process (e.g. MATLAB like)
– Parallel processing and distributed computing functional approaches (similar to Scala, Erlang and others)
– Python and other integration packages already there
– Happy marriage of several arenas
– Now in early release
• Feel free to contact or follow up with me on this
![Page 63: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/63.jpg)
SUMMARY: Core Competencies Across HPC
Core Competencies
Extreme scale
Architecture
Compute
I/O
Memory
Storage/data management
Tera, Peta, Exabytes….
Visualization and analytics
Fast fabrics
Future architectural direction
Parallelism to extreme parallelism
Multi core
Programming models
Big Data
Models, applications, applied
analytics
Structured, unstructured data types
![Page 64: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/64.jpg)
The need for a new discipline: HPC experts + Domain Expertise ==
Simulation.Specialists Core Competencies Where would this Computational
Specialist work?
Extreme scale
Architecture
Compute
I/O
Memory
Storage/data management
Tera, Peta, Exabytes….
Visualization and analytics
Fast fabrics
Future architectural direction
Parallelism to extreme parallelism
Multi core
Programming models
Big Data
Models, applications, applied
analytics
Structured, unstructured data types
National security
Fraud detection
Grand challenge science Physics, Chemistry, Biology,
Weather/climate, energy etc.
Bio/life sciences
Healthcare
Energy/Geophysics
Financial modeling, high frequency
and algorithmic trading
Entertainment/media
Auto/aero/mfg.
Consumer
Electronics
Risk informatics: insurance, global,
financial, medical etc.
Optimization models
Discovery analytics
![Page 65: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/65.jpg)
On a lighter note…..
![Page 66: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/66.jpg)
On a lighter note…..
![Page 67: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/67.jpg)
On a lighter note…..
![Page 68: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/68.jpg)
On a lighter note…..
![Page 69: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/69.jpg)
On a lighter note…..
![Page 70: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/70.jpg)
On a lighter note…..
![Page 71: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/71.jpg)
Further reading.…..
![Page 72: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/72.jpg)
Further reading.…..
![Page 73: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/73.jpg)
Further reading.…..
Currently reviewing
![Page 74: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/74.jpg)
Further reading.…..
![Page 75: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/75.jpg)
Innovative uses of HPC (LinkedIn.com)
![Page 76: Big Data Everywhere Chicago: High Performance Computing - Contributions Towards Big Data (HPC)](https://reader034.vdocument.in/reader034/viewer/2022042816/5599e7151a28ab3f7a8b4798/html5/thumbnails/76.jpg)
Thank you…..
• Email: sharan dot kalwani at acm dot org