Post on 19-Jan-2017
transcript
“Fifty Years of Supercomputing: From Colliding Black Holes to Dynamic Microbiomes
to the Exascale”
Director’s Distinguished Lecture National Center for Supercomputing Applications
University of Illinois at Urbana-ChampaignSeptember 16, 2016
Dr. Larry SmarrDirector, California Institute for Telecommunications and Information Technology
Harry E. Gruber Professor, Dept. of Computer Science and Engineering
Jacobs School of Engineering, UCSDhttp://lsmarr.calit2.net
1
Abstract
For the last thirty years, NCSA has played a critical role in bringing computational science and scientific visualization to the national user community. I will embed those three decades in the 50 year period 1975 to 2025, beginning with my solving Einstein's equations for colliding black holes on the megaFLOPs CDC 6600 and ending with the Exascale supercomputer. This 50 years spans a period in which we will have seen a one trillion-fold increase in supercomputer speed. Today we see the rise of data science and a new generation of architectures that balance traditional HPC with data analytics and pattern recognition. I will illustrate that with my current research into the microbial ecology dynamics within our human body. Finally, I will describe a number of trends in the decade to come, in which brain-inspired computing will become more prevalent in both software and hardware.
Personal Reflections on a Half Century of Supercomputing
• This Lecture Will Not Attempt to Cover in Detail the History of:– Supercomputers– NCSA– NSF’s HPC Program– Black Hole Collisions– Microbiomes– Scientific Visualization– Cyberinfrastructure– March to the Exascale– Machine Learning and Artificial Intelligence
• Rather I Will Discuss Significant Trends I Have Been Involved With
Forty Years of Computing Gravitational Waves From Colliding Black Holes – One Billion Times Increase in Supercomputer Speed!
1977
L. Smarr and K. EppleyGravitational Radiation Computed
from an Axisymmetric Black Hole Collision 40 Years
2016
LIGO ConsortiumSpiral Black Hole Collision
MegaFLOPS PetaFLOPSHolst, et al. Bull. Amer. Math. Soc 53, 513-554 (1916)
Documenting The Unmet Supercomputing Needsof A Broad Range of Disciplines Led to the NCSA Proposal to NSF
1982 1983
1985
NCSA Numerical Astrophysics GroupUsed NCSA Supercomputers to Explain Cosmic Phenomena
Mike Norman, Charles Evans, Roger Ove, John Hawley, Dean Sumi, Rob Wolff, Larry Smarr
Gas Accretion Onto a Black HoleCreates “Exhaust Channels”
Cosmic JetsEmerge from
Galactic CentersCollision of Neutron Stars
Exponential Increases in Supercomputer Speed and Visualization Technology Drive Understanding and Applications
Source: Donna Cox, Robert Patterson, Bob Wilhelmson, NCSA
1987
2005
Showed Thunderstorms Arise from Solving Physics Equations
Vastly Higher Resolution Uncovers Birth of Tornadoes
From Scientific Visualization of Supercomputing Science to Movie Special Effects
http://access.ncsa.uiuc.edu/http://movies.warnerbros.com/twister
www.jurassicpark.com; www.jamescameron.orgwww.cinemenium.com/perfectstorm/
NCSA 1987
1993
1996
2000
Computer GraphicsFrom NCSA to ILM
1991
Stefen Fangmeier
The Electronic Visualization Lab’s CAVE Virtual Reality System:Fully Immersive Science and Fantasy Worlds
CAVE conceived in 1991 by Tom DeFanti and Dan Sandin (EVL co-directors) and implemented by Carolina Cruz-Neira (Ph.D. student)
Crayoland
Colliding Galaxies QUAKE II
The CAVE
• EVL Invents ‘91• Debuts SIGGRAPH ’92• National Access NCSA
‘93
Industry Has Been a Close Collaborator in NCSA Advances:Example - Caterpillar
Real Time Linked Virtual Reality and Audio-Video Between NCSA, Peoria, Houston, and Germany
www.sv.vt.edu/future/vt-cave/apps/CatDistVR/DVR.html1996
Distributed Virtual Reality for Global-Scale Collaborative Prototyping
Visual Supercomputing Goal:Make Pattern Recognition Analysis as Powerful as Simulation
Colliding Galaxies (Smithsonian IMAX)-Donna Cox, Bob Patterson, NCSA-From “Cosmic Voyage”-Nominated for Academy Award 1997
• Virtual Director in CAVE• 1000 Hour SDSC Supercomputer Run to Generate Data• Tens of Thousands of Hours of NCSA SGI Time to Render Data• Cross-Country Transfer to IMAX Film of Massive Amounts of Data
Sponsored by Motorola for the Smithsonian
NCSA’s National Supercomputing LeadershipWith Blue Waters
March 28, 2012November 5, 2008
NCSA Blue Waters Continues NCSA’s Exponential Growth in Supercomputer Power
NCSA Blue Waters
2012
10,000,000
1,000,000
Smarr 2000 Slide
Twenty-Five Years from Bleeding Edge Research to Consumer Mass Market
• 1990 Leading Edge University Research Center-NCSA– Supercomputer GigaFLOPS Cray Y-MP ($15M)– Megabit/s NSFnet Backbone
• 2015 Global Market for Smartphones– Gigaherz Smartphones ($150)– Megabit/s Wireless Cellular Internet
But, NSF Blue Waters Petascale Supercomputer Is Over 1 Million Times Faster than Cray Y-MP!
Enormous Growth in ParallelismProcessors: Y-MP 4, Blue Waters 400,000
www.ncsa.uiuc.edu/BlueWaters/system.html
Big Data Analytics is a Key Componentof The Future of Supercomputing
Global Scientific Instruments Will Produce Ultralarge Datasets Continuously Requiring Dedicated Optic Fiber and Supercomputers
Large Synoptic Survey Telescope
www.lsst.org/sites/default/files/documents/DM%20Introduction%20-%20Kantor.pdf
Tracks ~40B Objects,Creates 10M Alerts/Night
Within 1 Minute of Observing
2x40Gb/s
There are 100 billion stars in the Andromeda galaxy…
…and 100 billion galaxies in theknown universe.
It’s a microbial world…
…there are 100 million times as many bacteria on Earth as stars in the universe. Microbiology is the ultimate Big Data science!
Most of Life’s Evolutionary Time Was in the Microbial World
You Are Here
Source: Carl Woese, et al
Tree of Life Derived from 16S rRNA Sequences
Carl Woese Mentored Me
on the MicrobiomeWhile I Was at UIUC
Your Body Has 10 Times As Many Microbe Cells As DNA-Bearing Human Cells
Your Microbiome is Your “Near-Body” Environment
and its CellsContain 100x as Many DNA Genes
As Your Human Cells
Inclusion of the “Dark Matter” of the BodyWill Radically Alter Medicine
As a Model for the Precision Medicine Initiative, I Have Tracked My Internal Biomarkers To Understand My Body’s Dynamics
My Quarterly Blood DrawCalit2 64 Megapixel VROOM
Longitudinal Time Series RevealedOscillatory Behavior in an Immune Variable Which is Antibacterial
Normal Range<7.3 µg/mL
124x Upper Limit for Healthy
Lactoferrin is a Protein Shed from Neutrophils -An Antibacterial that Sequesters Iron
TypicalLactoferrin Value
for Active
Inflammatory Bowel Disease
(IBD)
To Map Out the Dynamics of Autoimmune Microbiome Ecology Couples Next Generation Genome Sequencers to Big Data Supercomputers
Source: Weizhong Li, UCSD
Our Team Used 25 CPU-yearsto Compute
Comparative Gut MicrobiomesStarting From
2.7 Trillion DNA Bases of My Samples
and Healthy and IBD Subjects
Illumina HiSeq 2000 at JCVI
SDSC Gordon Data Supercomputer
The Supercomputer Converts Tens of Billions of DNA Fragments Into Relative Abundance of Hundreds of Microbial Species
Average Over 250 Healthy PeopleFrom NIH Human Microbiome Project
Note Log Scale
Clostridium difficile
Calit2’s Qualcomm Institute Has Developed Interactive Scalable Visualization for Biological Networks
20,000 Samples60,000 OTUs
18 Million Edges
Runs Native on QI’s 64Million Pixels Tiled Wall
Software Written by Philip Weber, Calit2’s QI
Visually Analyzing Our Supercomputing Data Reveals Vast Changes in the Human Gut Microbiome in Health and Disease
IlealCrohn’s
Healthy
UlcerativeColitis
www.sandia.gov/~smartin/presentations/OpenOrd.pdf
Source: Philip Weber,
QI, UCSD
President Obama Announces National Microbiome InitiativeMay 13, 2016
Complexity of Computing First Gut Microbiome DynamicsVersus First Dynamics of Colliding Black Holes
• My 1975 PhD Dissertation– Solving Einstein’s Equations of General Relativity for Colliding Black Holes and Grav Waves
– CDC 6600 Megaflop/s– Hundred Hours of Computing
• Rob Knight & Smarr Gut Microbiome Map Using 800,000 Core-Hours on SDSC’s Comet– Mapping From Illumina Sequencing to Taxonomy and Gene Abundance Dynamics
– Comet Petaflop/s – Comet Core is 40,000x CDC6600 Speed
– ~Million Core-Hours– 10,000x Supercomputer Time
• Gut Microbiome Takes ~ ½ Billion Times the Compute Power of Early Solutions of Dynamic General Relativity
The Future Foundation of Medicine is an Exponential Scaling-Up of the Number of Deeply Quantified Humans
Source: @EricTopolTwitter 9/27/2014
President Obama Has Committed The United StatesTo Building an Exascale (1000 PetaFLOPs) Computing System by 2025
• “Accelerating delivery of a capable exascale computing system” • “Increasing coherence between the technology base used for
modeling & simulation and that used for data analytic computing”
Is It Time to Radically Expand Our Computer Architectures?
NCSA 1988
Supercomputer Architectures Remain von NeumannShared Memory CPU Plus SIMD Co-Processor
NCSA 2016
The Future of Supercomputing Will Blend Traditional HPC and Data Analytics Integrating Non-von Neumann Architectures
“High Performance Computing Will Evolve Towards a Hybrid Model,
Integrating Emerging Non-von Neumann Architectures, with Huge Potential in Pattern Recognition,
Streaming Data Analysis, and Unpredictable New Applications.”
Horst Simon, Deputy Director, U.S. Department of Energy’s
Lawrence Berkeley National Laboratory
Brain-Inspired ProcessorsAre Accelerating the non-von Neumann Architecture Era
“On the drawing board are collections of 64, 256, 1024, and 4096 chips.
‘It’s only limited by money, not imagination,’ Modha says.”Source: Dr. Dharmendra ModhaFounding Director, IBM Cognitive Computing Group
August 8, 2014
Calit2’s Qualcomm Institute Has Established a Pattern Recognition Lab For Machine Learning on non-von Neumann Processors
“On the drawing board are collections of 64, 256, 1024, and 4096 chips.
‘It’s only limited by money, not imagination,’ Modha says.”Source: Dr. Dharmendra Modha
Founding Director, IBM Cognitive Computing Group
August 8, 2014
UCSD ECE Professor Ken Kreutz-Delgado Brings the IBM TrueNorth Chip
to Start Calit2’s Qualcomm Institute Pattern Recognition Laboratory
September 16, 2015
New Brain-Inspired Non-von Neumann Processors Are Emerging:KnuEdge Has Provided Processor to Calit2’s PRL
www.tomshardware.com/news/knuedge-announces-knuverse-and-knupath,31981.html
www.calit2.net/newsroom/release.php?id=2704
“KnuEdge and Calit2 have worked together since the early days of
the KnuEdge LambdaFabric processor, when key
personnel and technology from UC San Diego
provided the genesis for the first processor design.”
www.calit2.net/newsroom/release.php?id=2726
June 6, 2016
The Rise of Brain-Inspired Computers – Driver 1:Left & Right Brain Computing: Arithmetic vs. Pattern Recognition
Adapted from D-Wave
The Rise of Brain-Inspired Computers – Driver 2: Realtime Simulation of Human Brain Possible With Exascale Supercomputer
Horst Simon, Deputy Director, Lawrence Berkeley National Laboratory’s
National Energy Research Scientific Computing Center
Fastest Supercomputer
Trend LineTianhe-2
• Exascale Power Consumption of >20–30 MW
• The Human Brain Runs on 20 W• Our Brain is a Million Times More Power Efficient!
Horst Simon, Deputy Director, Lawrence Berkeley National Laboratory’s
National Energy Research Scientific Computing Center
The Rise of Brain-Inspired Computers – Driver 3:The Exascale Power Conundrum
The Rise of Brain-Inspired Computers – Driver 4: Reverse Engineering of the Brain Is Accelerating
www.whitehouse.gov/infographics/brain-initiative
Interactively Exploring Microscope Images of Brains:40Gbps From NCMIR to Calit2 64Mpixel Wall
Data and Image Source: Mark Ellisman, NCMIR, UCSD
Artificial Intelligence (AI) is Advancing at a Amazing Pace:Deep Learning Algorithms Working on Massive Datasets
Training on 30M Moves, Then Playing Against Itself
Less Than 2 Years!
Google Released Its AI Software as Open SourceAccelerating Development
https://exponential.singularityu.org/medicine/big-data-machine-learning-with-jeremy-howard/
From Programming Computers Step by Step To Achieve a Goal
To Showing the Computer Some Examples of
What You Want It to Achieve and Then Letting the Computer
Figure It Out On Its Own--Jeremy Howard, Singularity Univ.
2015
November 9, 2015
Corporate Cloud ProvidersAre Already Using Non von Neumann Accelerators
www.microsoft.com/en-us/research/project/project-catapult/
https://cloudplatform.googleblog.com/2016/05/Google-supercharges-machine-learning-tasks-with-custom-chip.html
Microsoft Installs FPGAs into Bing Servers and 432 into TAAC for Academic Access
Deep Learning Will Provide Artificial Intelligence Personalized Assistants to Coach Us to Wellness
Where Medicine Coaching is Now
Where Wellness Coaching is Going
January 10, 2014
Can a Planetary Supercomputer with Artificial IntelligenceTransform Our Sickcare System to Healthcare?
Using this data, the planetary computer will be able to build a computational model of your body
and compare your sensor stream with millions of others. Besides providing early detection of internal changes
that could lead to disease, cloud-powered voice-recognition wellness coaches
could provide continual personalized support on lifestyle choices, potentially staving off disease
and making health care affordable for everyone.
ESSAYAn Evolution Toward a Programmable UniverseBy LARRY SMARRPublished: December 5, 2011
In Spite of the Enormous Promise,A Global Debate is Underway About the Dangers of Superintelligence
"Those disposed to dismiss an 'AI takeover' as science fiction may think again after reading this original and well-argued book." —Martin Rees, Past President, Royal Society
If our own extinction is a likely, or even
possible, outcome of our technological
development, shouldn't we proceed with great
caution? – Bill Joy
Success in creating AI would be the biggest event in human history. Unfortunately, it might also be the last, unless we learn how to avoid the risks. – Steven Hawking
Things Are About to Get Very Interesting…
Source: Hans Moravecwww.transhumanist.com/volume1/power_075.jpg
Smarr Slide from 2001