Работа с базами данных
<<  Коллекционирование 2 класс Мероприятия в очаге холеры схема  >>
Large Data Challenge: Average Throughput to End User on Shared
Large Data Challenge: Average Throughput to End User on Shared
Large Data Challenge: Average Throughput to End User on Shared
Large Data Challenge: Average Throughput to End User on Shared
Large Data Challenge: Average Throughput to End User on Shared
Large Data Challenge: Average Throughput to End User on Shared
OptIPuter Solution: Give Dedicated Optical Channels to Data-Intensive
OptIPuter Solution: Give Dedicated Optical Channels to Data-Intensive
OptIPuter Solution: Give Dedicated Optical Channels to Data-Intensive
OptIPuter Solution: Give Dedicated Optical Channels to Data-Intensive
The Global Lambda Integrated Facility-- Creating a Planetary-Scale
The Global Lambda Integrated Facility-- Creating a Planetary-Scale
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
Academic Research OptIPlanet Collaboratory: A 10Gbps “End-to-End”
The OptIPuter Project: Creating High Resolution Portals Over Dedicated
The OptIPuter Project: Creating High Resolution Portals Over Dedicated
The OptIPuter Project: Creating High Resolution Portals Over Dedicated
The OptIPuter Project: Creating High Resolution Portals Over Dedicated
The OptIPuter Project: Creating High Resolution Portals Over Dedicated
The OptIPuter Project: Creating High Resolution Portals Over Dedicated
The OptIPuter Project: Creating High Resolution Portals Over Dedicated
The OptIPuter Project: Creating High Resolution Portals Over Dedicated
The OptIPuter Project: Creating High Resolution Portals Over Dedicated
The OptIPuter Project: Creating High Resolution Portals Over Dedicated
The OptIPuter Project: Creating High Resolution Portals Over Dedicated
The OptIPuter Project: Creating High Resolution Portals Over Dedicated
The OptIPuter Project: Creating High Resolution Portals Over Dedicated
The OptIPuter Project: Creating High Resolution Portals Over Dedicated
MIT’s Ed DeLong and Darwin Project Team Using OptIPortal to Analyze
MIT’s Ed DeLong and Darwin Project Team Using OptIPortal to Analyze
MIT’s Ed DeLong and Darwin Project Team Using OptIPortal to Analyze
MIT’s Ed DeLong and Darwin Project Team Using OptIPortal to Analyze
AESOP Display built by Calit2 for KAUST-- King Abdullah University of
AESOP Display built by Calit2 for KAUST-- King Abdullah University of
The Latest OptIPuter Innovation: Quickly Deployable Nearly Seamless
The Latest OptIPuter Innovation: Quickly Deployable Nearly Seamless
The OctIPortable Being Checked Out Prior to Shipping to the
The OctIPortable Being Checked Out Prior to Shipping to the
3D Stereo Head Tracked OptIPortal: NexCAVE
3D Stereo Head Tracked OptIPortal: NexCAVE
High Definition Video Connected OptIPortals: Virtual Working Spaces
High Definition Video Connected OptIPortals: Virtual Working Spaces
High Definition Video Connected OptIPortals: Virtual Working Spaces
High Definition Video Connected OptIPortals: Virtual Working Spaces
“Blueprint for the Digital University”--Report of the UCSD Research
“Blueprint for the Digital University”--Report of the UCSD Research
Calit2 Sunlight OptIPuter Exchange Connects 60 Campus Sites Each
Calit2 Sunlight OptIPuter Exchange Connects 60 Campus Sites Each
Calit2 Sunlight OptIPuter Exchange Connects 60 Campus Sites Each
Calit2 Sunlight OptIPuter Exchange Connects 60 Campus Sites Each
UCSD Campus Investment in Fiber Enables Consolidation of Energy
UCSD Campus Investment in Fiber Enables Consolidation of Energy
UCSD Campus Investment in Fiber Enables Consolidation of Energy
UCSD Campus Investment in Fiber Enables Consolidation of Energy
UCSD Campus Investment in Fiber Enables Consolidation of Energy
UCSD Campus Investment in Fiber Enables Consolidation of Energy
UCSD Campus Investment in Fiber Enables Consolidation of Energy
UCSD Campus Investment in Fiber Enables Consolidation of Energy
UCSD Campus Investment in Fiber Enables Consolidation of Energy
UCSD Campus Investment in Fiber Enables Consolidation of Energy
UCSD Campus Investment in Fiber Enables Consolidation of Energy
UCSD Campus Investment in Fiber Enables Consolidation of Energy
UCSD Campus Investment in Fiber Enables Consolidation of Energy
UCSD Campus Investment in Fiber Enables Consolidation of Energy
UCSD Campus Investment in Fiber Enables Consolidation of Energy
UCSD Campus Investment in Fiber Enables Consolidation of Energy
UCSD Campus Investment in Fiber Enables Consolidation of Energy
UCSD Campus Investment in Fiber Enables Consolidation of Energy
Data-Intensive Supercomputer Based on SSD Flash Memory and Virtual
Data-Intensive Supercomputer Based on SSD Flash Memory and Virtual
Gordon Bests Previous Mega I/O per Second by 25x
Gordon Bests Previous Mega I/O per Second by 25x
Arista Enables SDSC’s Massive Parallel 10G Switched Data Analysis
Arista Enables SDSC’s Massive Parallel 10G Switched Data Analysis
Arista Enables SDSC’s Massive Parallel 10G Switched Data Analysis
Arista Enables SDSC’s Massive Parallel 10G Switched Data Analysis
Arista Enables SDSC’s Massive Parallel 10G Switched Data Analysis
Arista Enables SDSC’s Massive Parallel 10G Switched Data Analysis
The Next Step for Data-Intensive Science: Pioneering the HPC Cloud
The Next Step for Data-Intensive Science: Pioneering the HPC Cloud
The Next Step for Data-Intensive Science: Pioneering the HPC Cloud
The Next Step for Data-Intensive Science: Pioneering the HPC Cloud
The Next Step for Data-Intensive Science: Pioneering the HPC Cloud
The Next Step for Data-Intensive Science: Pioneering the HPC Cloud
The Next Step for Data-Intensive Science: Pioneering the HPC Cloud
The Next Step for Data-Intensive Science: Pioneering the HPC Cloud
The Next Step for Data-Intensive Science: Pioneering the HPC Cloud
The Next Step for Data-Intensive Science: Pioneering the HPC Cloud
The Next Step for Data-Intensive Science: Pioneering the HPC Cloud
The Next Step for Data-Intensive Science: Pioneering the HPC Cloud
The Next Step for Data-Intensive Science: Pioneering the HPC Cloud
The Next Step for Data-Intensive Science: Pioneering the HPC Cloud
The Next Step for Data-Intensive Science: Pioneering the HPC Cloud
The Next Step for Data-Intensive Science: Pioneering the HPC Cloud
Exploring Cosmology With Supercomputers, Supernetworks, and
Exploring Cosmology With Supercomputers, Supernetworks, and
Exploring Cosmology With Supercomputers, Supernetworks, and
Exploring Cosmology With Supercomputers, Supernetworks, and
Providing End-to-End CI for Petascale End Users
Providing End-to-End CI for Petascale End Users
Providing End-to-End CI for Petascale End Users
Providing End-to-End CI for Petascale End Users
Using Supernetworks to Couple End User’s OptIPortal to Remote
Using Supernetworks to Couple End User’s OptIPortal to Remote
Using Supernetworks to Couple End User’s OptIPortal to Remote
Using Supernetworks to Couple End User’s OptIPortal to Remote
Using Supernetworks to Couple End User’s OptIPortal to Remote
Using Supernetworks to Couple End User’s OptIPortal to Remote
Most of Evolutionary Time Was in the Microbial World
Most of Evolutionary Time Was in the Microbial World
The New Science of Microbial Metagenomics
The New Science of Microbial Metagenomics
Calit2 Microbial Metagenomics Cluster- Next Generation Optically
Calit2 Microbial Metagenomics Cluster- Next Generation Optically
Calit2 Microbial Metagenomics Cluster- Next Generation Optically
Calit2 Microbial Metagenomics Cluster- Next Generation Optically
Calit2 Microbial Metagenomics Cluster- Next Generation Optically
Calit2 Microbial Metagenomics Cluster- Next Generation Optically
Calit2 Microbial Metagenomics Cluster- Next Generation Optically
Calit2 Microbial Metagenomics Cluster- Next Generation Optically
Calit2 CAMERA: Over 4000 Registered Users From Over 80 Countries
Calit2 CAMERA: Over 4000 Registered Users From Over 80 Countries
Creating CAMERA 2.0 - Advanced Cyberinfrastructure Service Oriented
Creating CAMERA 2.0 - Advanced Cyberinfrastructure Service Oriented
Creating CAMERA 2.0 - Advanced Cyberinfrastructure Service Oriented
Creating CAMERA 2.0 - Advanced Cyberinfrastructure Service Oriented
Creating CAMERA 2.0 - Advanced Cyberinfrastructure Service Oriented
Creating CAMERA 2.0 - Advanced Cyberinfrastructure Service Oriented
Creating CAMERA 2.0 - Advanced Cyberinfrastructure Service Oriented
Creating CAMERA 2.0 - Advanced Cyberinfrastructure Service Oriented
Creating CAMERA 2.0 - Advanced Cyberinfrastructure Service Oriented
Creating CAMERA 2.0 - Advanced Cyberinfrastructure Service Oriented
The GreenLight Project: Instrumenting the Energy Cost of Computational
The GreenLight Project: Instrumenting the Energy Cost of Computational
The GreenLight Project: Instrumenting the Energy Cost of Computational
The GreenLight Project: Instrumenting the Energy Cost of Computational
The GreenLight Project: Instrumenting the Energy Cost of Computational
The GreenLight Project: Instrumenting the Energy Cost of Computational
The GreenLight Project: Instrumenting the Energy Cost of Computational
The GreenLight Project: Instrumenting the Energy Cost of Computational
The GreenLight Project: Instrumenting the Energy Cost of Computational
The GreenLight Project: Instrumenting the Energy Cost of Computational
The GreenLight Project: Instrumenting the Energy Cost of Computational
The GreenLight Project: Instrumenting the Energy Cost of Computational
The GreenLight Project: Instrumenting the Energy Cost of Computational
The GreenLight Project: Instrumenting the Energy Cost of Computational
The GreenLight Project: Instrumenting the Energy Cost of Computational
The GreenLight Project: Instrumenting the Energy Cost of Computational
The GreenLight Project: Instrumenting the Energy Cost of Computational
The GreenLight Project: Instrumenting the Energy Cost of Computational
GreenLight Project: Remote Visualization of Data Center
GreenLight Project: Remote Visualization of Data Center
GreenLight Projects Airflow dynamics
GreenLight Projects Airflow dynamics
GreenLight Project Heat Distribution
GreenLight Project Heat Distribution
Cost Per Megabase in Sequencing DNA is Falling Much Faster Than
Cost Per Megabase in Sequencing DNA is Falling Much Faster Than
BGI—The Beijing Genome Institute is the World’s Largest Genomic
BGI—The Beijing Genome Institute is the World’s Largest Genomic
BGI—The Beijing Genome Institute is the World’s Largest Genomic
BGI—The Beijing Genome Institute is the World’s Largest Genomic
BGI—The Beijing Genome Institute is the World’s Largest Genomic
BGI—The Beijing Genome Institute is the World’s Largest Genomic
From 10,000 Human Genomes Sequenced in 2011 to 1 Million by 2015 in
From 10,000 Human Genomes Sequenced in 2011 to 1 Million by 2015 in
From 10,000 Human Genomes Sequenced in 2011 to 1 Million by 2015 in
From 10,000 Human Genomes Sequenced in 2011 to 1 Million by 2015 in
Needed: Interdisciplinary Teams Made From Computer Science, Data
Needed: Interdisciplinary Teams Made From Computer Science, Data
Needed: Interdisciplinary Teams Made From Computer Science, Data
Needed: Interdisciplinary Teams Made From Computer Science, Data
Calit2 Brings Together Computer Science and Bioinformatics
Calit2 Brings Together Computer Science and Bioinformatics
Calit2 Brings Together Computer Science and Bioinformatics
Calit2 Brings Together Computer Science and Bioinformatics
Calit2 Brings Together Computer Science and Bioinformatics
Calit2 Brings Together Computer Science and Bioinformatics
Calit2 Brings Together Computer Science and Bioinformatics
Calit2 Brings Together Computer Science and Bioinformatics
Calit2 Brings Together Computer Science and Bioinformatics
Calit2 Brings Together Computer Science and Bioinformatics
Calit2 Brings Together Computer Science and Bioinformatics
Calit2 Brings Together Computer Science and Bioinformatics
Calit2 Brings Together Computer Science and Bioinformatics
Calit2 Brings Together Computer Science and Bioinformatics
GreenLight Project Allows for Testing of Novel Architectures on
GreenLight Project Allows for Testing of Novel Architectures on
GreenLight Project Allows for Testing of Novel Architectures on
GreenLight Project Allows for Testing of Novel Architectures on
Using UCSD RCI to Store and Analyze Next Gen Sequencer Datasets
Using UCSD RCI to Store and Analyze Next Gen Sequencer Datasets
NIH National Center for Microscopy & Imaging Research Integrated
NIH National Center for Microscopy & Imaging Research Integrated
NIH National Center for Microscopy & Imaging Research Integrated
NIH National Center for Microscopy & Imaging Research Integrated
UCSD Planned Optical Networked Biomedical Researchers and Instruments
UCSD Planned Optical Networked Biomedical Researchers and Instruments
Картинки из презентации «A Campus-Scale High Performance Cyberinfrastructure is Required for Data-Intensive Research» к уроку информатики на тему «Работа с базами данных»

Автор: Jerry Sheehan. Чтобы познакомиться с картинкой полного размера, нажмите на её эскиз. Чтобы можно было использовать все картинки для урока информатики, скачайте бесплатно презентацию «A Campus-Scale High Performance Cyberinfrastructure is Required for Data-Intensive Research.ppt» со всеми картинками в zip-архиве размером 10774 КБ.

A Campus-Scale High Performance Cyberinfrastructure is Required for Data-Intensive Research

содержание презентации «A Campus-Scale High Performance Cyberinfrastructure is Required for Data-Intensive Research.ppt»
Сл Текст Сл Текст
1“A Campus-Scale High Performance 208GB/s Sustained Today Phase I: > 50
Cyberinfrastructure is Required for GB/sec for Lustre (May 2011) :Phase II:
Data-Intensive Research”. Seminar >100 GB/s (Feb 2012). Source: Philip
Presentation Princeton Institute for Papadopoulos, SDSC/Calit2. 2000 TB > 50
Computational Science and Engineering GB/s. UCSD RCI. OptIPuter. Co-Lo.
(PICSciE) Princeton University Princeton, CENIC/NLR. Oasis Procurement (RFP). 5. 8.
NJ December 12, 2011. Dr. Larry Smarr 2. 32. 4. 8. 32. 2. 12. 40?128. 8. 128.
Director, California Institute for 21The Next Step for Data-Intensive
Telecommunications and Information Science: Pioneering the HPC Cloud.
Technology Harry E. Gruber Professor, 22Data Oasis – 3 Different Types of
Dept. of Computer Science and Engineering Storage.
Jacobs School of Engineering, UCSD 23Examples of Applications Built on UCSD
http://lsmarr.calit2.net. 1. RCI. DOE Remote Use of Petascale HPC Moore
2Abstract. Campuses are experiencing an Foundation Microbial Metagenomics Server
enormous increase in the quantity of data NSF GreenLight Instrumented Data Center
generated by scientific instruments and NIH Next Generation Gene Sequencers NIH
computational clusters and stored in Shared Scientific Instruments.
massive data repositories. The shared 24Exploring Cosmology With
Internet, engineered to enable interaction Supercomputers, Supernetworks, and
with megabyte-sized data objects is not Supervisualization. 40963 Particle/Cell
capable of dealing with the typical Hydrodynamic Cosmology Simulation NICS
gigabytes to terabytes of modern Kraken (XT5) 16,384 cores Output 148 TB
scientific data. Instead, a high Movie Output (0.25 TB/file) 80 TB
performance cyberinfrastructure is Diagnostic Dumps (8 TB/file).
emerging to support data-intensive Intergalactic Medium on 2 GLyr Scale.
research. Fortunately, multi-channel Science: Norman, Harkness,Paschos SDSC
optical fiber can support both the Visualization: Insley, ANL; Wagner SDSC.
traditional internet and this new data Source: Mike Norman, SDSC. ANL * Calit2 *
utility. I will give examples of early LBNL * NICS * ORNL * SDSC.
prototypes which integrate data 25Providing End-to-End CI for Petascale
generation, transmission, storage, End Users. log of gas temperature log of
analysis, visualization, curation, and gas density. Two 64K Images From a
sharing, driven by applications as diverse Cosmological Simulation of Galaxy Cluster
as genomics, ocean observatories, and Formation. Mike Norman, SDSC October 10,
cosmology. 2008.
3Large Data Challenge: Average 26Using Supernetworks to Couple End
Throughput to End User on Shared Internet User’s OptIPortal to Remote Supercomputers
is 10-100 Mbps. Transferring 1 TB: --50 and Visualization Servers. Source: Mike
Mbps = 2 Days --10 Gbps = 15 Minutes. Norman, Rick Wagner, SDSC. Real-Time
http://ensight.eos.nasa.gov/Missions/terra Interactive Volume Rendering Streamed from
index.shtml. Tested December 2011. ANL to SDSC. *ANL * Calit2 * LBNL * NICS *
4OptIPuter Solution: Give Dedicated ORNL * SDSC.
Optical Channels to Data-Intensive Users. 27Most of Evolutionary Time Was in the
10 Gbps per User ~ 100x Shared Internet Microbial World. Earth is a Microbial
Throughput. Parallel Lambdas are Driving World: For Every Human Cell There are 100
Optical Networking The Way Parallel Million Microbes. Tree of Life Derived
Processors Drove 1990s Computing. from 16S rRNA Sequences. Source: Carl
5The Global Lambda Integrated Woese, et al.
Facility-- Creating a Planetary-Scale High 28The New Science of Microbial
Bandwidth Collaboratory. Research Metagenomics. NRC Report: Metagenomic data
Innovation Labs Linked by 10G Dedicated should be made publicly available in
Lambdas. international archives as rapidly as
www.glif.is/publications/maps/GLIF_5-11_Wo possible. “The emerging field of
ld_2k.jpg. metagenomics, where the DNA of entire
6Academic Research OptIPlanet communities of microbes is studied
Collaboratory: A 10Gbps “End-to-End” simultaneously, presents the greatest
Lightpath Cloud. HD/4k Live Video. HPC. opportunity – perhaps since the invention
Local or Remote Instruments. End User of the microscope – to revolutionize
OptIPortal. National LambdaRail. Campus understanding of the microbial world.” –
Optical Switch. Data Repositories & National Research Council March 27, 2007.
Clusters. HD/4k Video Repositories. 10G 29Calit2 Microbial Metagenomics Cluster-
Lightpaths. Next Generation Optically Linked Science
7The OptIPuter Project: Creating High Data Server. Grant Announced January 17,
Resolution Portals Over Dedicated Optical 2006.
Channels to Global Science Data. 30Calit2 CAMERA: Over 4000 Registered
OptIPortal. Scalable Adaptive Graphics Users From Over 80 Countries.
Environment (SAGE). Calit2 (UCSD, UCI), http://camera.calit2.net/. Community
SDSC, and UIC Leads—Larry Smarr PI Univ. Cyberinfrastructure for Advanced Microbial
Partners: NCSA, USC, SDSU, NW, TA&M, Ecology Research and Analysis.
UvA, SARA, KISTI, AIST Industry: IBM, Sun, 31Creating CAMERA 2.0 - Advanced
Telcordia, Chiaro, Calient, Glimmerglass, Cyberinfrastructure Service Oriented
Lucent. Picture Source: Mark Ellisman, Architecture. Source: CAMERA CTO Mark
David Lee, Jason Leigh. Ellisman.
8MIT’s Ed DeLong and Darwin Project 32The GreenLight Project: Instrumenting
Team Using OptIPortal to Analyze 10km the Energy Cost of Computational Science.
Ocean Microbial Simulation. Focus on 5 Communities with At-Scale
Cross-Disciplinary Research at MIT, Computing Needs: Metagenomics Ocean
Connecting Systems Biology, Microbial Observing Microscopy Bioinformatics
Ecology, Global Biogeochemical Cycles and Digital Media Measure, Monitor, & Web
Climate. Publish Real-Time Sensor Outputs Via
9AESOP Display built by Calit2 for Service-oriented Architectures Allow
KAUST-- King Abdullah University of Researchers Anywhere To Study Computing
Science & Technology. 40-Tile 46” Energy Cost Enable Scientists To Explore
Diagonal Narrow-Bezel AESOP Display at Tactics For Maximizing Work/Watt Develop
KAUST Running CGLX. Middleware that Automates Optimal Choice
10The Latest OptIPuter Innovation: of Compute/RAM Power Strategies for
Quickly Deployable Nearly Seamless Desired Greenness Data Center for School
OptIPortables. Shipping Case. Image From of Medicine Illumina Next Gen Sequencer
the Calit2 KAUST Lab. 45 minute setup, 15 Storage and Processing. Source: Tom
minute tear-down with two people (possible DeFanti, Calit2; GreenLight PI.
with one). 33GreenLight Project: Remote
11The OctIPortable Being Checked Out Visualization of Data Center.
Prior to Shipping to the Calit2/KAUST 34GreenLight Projects Airflow dynamics.
Booth at SIGGRAPH 2011. Photo:Tom DeFanti. Live fan speeds. Airflow dynamics. 34.
123D Stereo Head Tracked OptIPortal: 35GreenLight Project Heat Distribution.
NexCAVE. Array of JVC HDTV 3D LCD Screens Combined heat + fans. Realistic
KAUST NexCAVE = 22.5MPixels. correlation.
www.calit2.net/newsroom/article.php?id=158 36Cost Per Megabase in Sequencing DNA is
. Source: Tom DeFanti, Calit2@UCSD. Falling Much Faster Than Moore’s Law.
13High Definition Video Connected www.genome.gov/sequencingcosts/.
OptIPortals: Virtual Working Spaces for 37BGI—The Beijing Genome Institute is
Data Intensive Research. 2010. NASA the World’s Largest Genomic Institute.
Supports Two Virtual Institutes. LifeSize Main Facilities in Shenzhen and Hong Kong,
HD. Calit2@UCSD 10Gbps Link to NASA Ames China Branch Facilities in Copenhagen,
Lunar Science Institute, Mountain View, Boston, UC Davis 137 Illumina HiSeq 2000
CA. Source: Falko Kuester, Kai Doerr Next Generation Sequencing Systems Each
Calit2; Michael Sims, Larry Edwards, Illumina Next Gen Sequencer Generates 25
Estelle Dodson NASA. Gigabases/Day Supported by High
14“Blueprint for the Digital Performance Computing and Storage ~160TF,
University”--Report of the UCSD Research 33TB Memory Large-Scale (12PB) Storage.
Cyberinfrastructure Design Team. No Data 38From 10,000 Human Genomes Sequenced in
Bottlenecks--Design for Gigabit/s Data 2011 to 1 Million by 2015 in Less Than
Flows. A Five Year Process Begins Pilot 5,000 sq. ft.! 4 Million Newborns / Year
Deployment This Year. April 2009. in U.S.
research.ucsd.edu/documents/rcidt/RCIDTRep 39Needed: Interdisciplinary Teams Made
rtFinal2009.pdf. From Computer Science, Data Analytics, and
15Calit2 Sunlight OptIPuter Exchange Genomics.
Connects 60 Campus Sites Each Dedicated at 40Calit2 Brings Together Computer
10Gbps. Maxine Brown, EVL, UIC OptIPuter Science and Bioinformatics.
Project Manager. 41GreenLight Project Allows for Testing
16UCSD Campus Investment in Fiber of Novel Architectures on Bioinformatics
Enables Consolidation of Energy Efficient Algorithms. “Our version of MS-Alignment
Computing & Storage. Source: Philip [a proteomics algorithm] is more than 115x
Papadopoulos, SDSC, UCSD. faster than a single core of an Intel
17Data-Intensive Supercomputer Based on Nehalem processor, is more than 15x faster
SSD Flash Memory and Virtual Shared Memory than an eight-core version, and reduces
SW Emphasizes MEM and IOPS over FLOPS the runtime for a few samples from 24
Supernode has Virtual Shared Memory: 2 TB hours to just a few hours.” —From
RAM Aggregate 8 TB SSD Aggregate Total “Computational Mass Spectrometry in a
Machine = 32 Supernodes 4 PB Disk Parallel Reconfigurable Coherent Co-processing
File System >100 GB/s I/O System Architecture,” IEEE Design & Test of
Designed to Accelerate Access to Massive Computers, Yalamarthy (ECE), Coburn (CSE),
Datasets being Generated in Many Fields of Gupta (CSE), Edwards (Convey), and Kelly
Science, Engineering, Medicine, and Social (Convey) (2011). June 23, 2009.
Science. NSF Funds a Big Data http://research.microsoft.com/en-us/um/cam
Supercomputer: SDSC’s Gordon-Dedicated ridge/events/date2011/msalignment_datepost
Dec. 5, 2011. Source: Mike Norman, Allan r_2011.pdf.
Snavely SDSC. 42Using UCSD RCI to Store and Analyze
18Gordon Bests Previous Mega I/O per Next Gen Sequencer Datasets. Stream Data
Second by 25x. from Genomics Lab to GreenLight Storage,
19Rapid Evolution of 10GbE Port Prices NFS Mount Over 10Gbps to Triton Compute
Makes Campus-Scale 10Gbps CI Affordable. Cluster. Source: Chris Misleh, SOM/Calit2
Port Pricing is Falling Density is Rising UCSD.
– Dramatically Cost of 10GbE Approaching 43NIH National Center for Microscopy
Cluster HPC Interconnects. Source: Philip & Imaging Research Integrated
Papadopoulos, SDSC/Calit2. $80K/port Infrastructure of Shared Resources. Shared
Chiaro (60 Max). $ 5K Force 10 (40 max). Infrastructure. Scientific Instruments.
~$1000 (300+ Max). $ 500 Arista 48 ports. Local SOM Infrastructure. End User
$ 400 Arista 48 ports. 2005 2007 2009 Workstations. Source: Steve Peltier, Mark
2010. Ellisman, NCMIR.
20Arista Enables SDSC’s Massive Parallel 44UCSD Planned Optical Networked
10G Switched Data Analysis Resource. Biomedical Researchers and Instruments.
Triton. Trestles 100 TF. Dash. Gordon. Connects at 10 Gbps : Microarrays Genome
10Gbps. Radical Change Enabled by Arista Sequencers Mass Spectrometry Light and
7508 10G Switch 384 10G Capable. Existing Electron Microscopes Whole Body Imagers
Commodity Storage 1/3 PB. Phase0: > Computing Storage.
A Campus-Scale High Performance Cyberinfrastructure is Required for Data-Intensive Research.ppt
http://900igr.net/kartinka/informatika/a-campus-scale-high-performance-cyberinfrastructure-is-required-for-data-intensive-research-241340.html
cсылка на страницу

A Campus-Scale High Performance Cyberinfrastructure is Required for Data-Intensive Research

другие презентации на тему «A Campus-Scale High Performance Cyberinfrastructure is Required for Data-Intensive Research»

«English for you» - You are welcome! Ты сможешь совершенствовать своё произношение. Все слова и выражения озвучены носителями языка. Узнать насколько хорошо ты усвоил материал тебе помогут: Викторины Задания Игры Игровые упражнения Ты можешь выбрать уровень сложности. «Английский для тебя» (курс английского языка с применением ИКТ).

«Data Mining» - Сокращение дерева или отсечение ветвей. Кросс-проверка - известный метод получения оценок неизвестных параметров модели. Data Mining. Метод "ближайшего соседа" или системы рассуждений на основе аналогичных случаев. Статистические методы. Задачи Data Mining. Недостатки. Остановка построения дерева.

«Управление базами данными» - Примеры настольных БД. Обзор серверных СУБД. Язык запросов. Способы создания пустой таблицы. Мастера Microsoft Access. Логические значения. Уникальная метка. Решения. Informix. Распределенные СУБД. Гибридные БД. Oracle 8. Создание базы данных. Структура реляционной таблицы. Этапы проектирования БД. Автоматическая проверка типа данных.

«Большие объекты» - Вставка. Рисунок. Операции с полями большого размера. Физическое представление объекта. Улучшенное двухуровневое разбиение. Размер хранимого поля. Алгоритм. Поля большого размера. Организация памяти. Древовидное представление. Удалить N байт. Удаление. Упражнения. Обозначения. Система ‘близнецов’ . Замечания.

«Практические работы по базам данных» - Требования к уровню подготовки. Конструктор запросов. . 1. Понятие и классификация информационных систем. 2. Способы организации данных. Типы данных. Сохранить запрос. Таблица и форма. Построение схемы данных. СУБД MS Access. 4. Формы представления баз данных. Практические работы. Построить и выполнить запрос из двух таблиц по некоторым полям.

«Запросы к данным» - Управление транзакциями. Запрос на обновление. Запрос на объединение. Файл журнала. Двоичные журналы. Creat user. Резервное копирование. Transform. Привилегии. Оператор. Управление доступом к данным. Специальные запросы. Полное резервирование. Перекрестная таблица. Запрос на добавление. Фамилия. Запрос на создание новой таблицы.

Работа с базами данных

11 презентаций о работе с базами данных
Урок

Информатика

130 тем
Картинки
900igr.net > Презентации по информатике > Работа с базами данных > A Campus-Scale High Performance Cyberinfrastructure is Required for Data-Intensive Research