Geospatial Technologies and Environmental CyberInfrastructure (GeoTECI) Lab Dr. Jianting Zhang

Transcription

1 Affiliated Institutions Students: Simin You (Ph.D ), Siyu Liao (Ph.D ), Costin Vicoveanu (Undergraduate, 2014-) Bharat Rosanlall (Undergraduate, 2014), Jay Yao (MS-thesis, ), Chandrashekar Singh (MS 2013), Agniva Banerjee (MS, 2012), Roger King (MS, 2012), Wahyu Nugroho (MS, 2011), Xiao Quan Cen Feng (MS 2011), Chetram Dasrat (Undergraduate, 2008) Geospatial echnologies and Environmental CyberInfrastructure (GeoECI) Lab Dr. Jianting Zhang Department of Computer Science he City College of New York Collaborating Institutions

2 Geographical Information System Social Studies Computational Geometry Computer Graphics Spatial Databases: data modeling, indexing, query processing Scientific Data/Information Visualization Statistics/Machine learning Image Processing/Computer Vision GIS Remote Sensing Social- Economic Modeling Environmental Modeling Census/axation Urban planning ransportation Air quality Hydrology Ecology

3 Ecological Informatics Geography GIS Applications Remote Sensing Computer Science Spatial Databases Data Mining Environmental sciences Computer Science

4 Big Geospatial Data Challenges Event Locations, trajectories and O-D data E.g., axi trip records (GPS traces or O-D locations) 0.5 million in NYC (medallion taxi cab only) and 1.2 million in Beijing per day From O-D locations to trajectories to frequent patterns Satellite: e.g., from GOES to GOES-R (2015/2016) [$11B] Spectral (3X)*spatial (4X)* temporal (5X)=60X 2km*2km*5min*16bands (360*60)*(180*60)*(12*24)*16~ 1+ trillion pixels per day Derived thematic data products (vector) Species distributions E.g million occurrence records (GBIF) E.g. 717,057 polygons and 78,929,697 vertices for 4148 birds distribution data (NatureServe)

5 Cloud computing+mapreduce+hadoop GPU SIMD CPU Host (CMP) GDRAM... GDRAM PCI-E Local Cache PCI-E Ring Bus C hread Block B A Shared Cache HDD DRAM SSD MIC hreads In-Order Local Cache 16 Intel Sandy Bridge CPU cores+ 128GB RAM + 8B disk + GX IAN + Xeon Phi 3120A ~ $9,994

6 ASCI Red: 1997 First 1 eraflops (sustained) system with 9298 Intel Pentium II Xeon processors (in 72 Cabinets) Feb billion transistors (551mm²) 2,688 processors 4.5 FLOPS SP and 1.3 FLOPS DP Max bandwidth GB/s PCI-E peripheral device 250 W (17.98 GFLOPS/W -SP) Suggested retail price: $999 What can we do today using a device that is more powerful than ASCI Red 16 years ago?

7 $449,845/4yr (08/01/ /31/2017) HIGHES-DB HIgh-performance GrapHics units based Engine for Spatial-emporal data Spatial and Spatiotemporal indexing, query processing and optimization rajectory data management on GPUs Segmentation/simplification/compression/Aggregation/Warehousing Map matching with road networks Data mining (moving cluster, convoy, swarm...) when yellow cabs, green cabs and MA buses meet with multicore CPUs, GPUs and MICs in NYC

8 when GOES-R satellites, extratropical cyclones and hummingbirds meet with IAN V emporal rends High-resolution Satellite Imagery Data Assimilation In-situ Observation Sensor Data Zonal Statistics Ecological, environmental and administrative zones ROIs Global and Regional Climate Model Outputs C B High-End Computing Facility A hread Block

9 ...building a highly-configurable experimental computing environment for innovative BigData technologies CCNY Computer Science LAN GeoECI@CCNY CUNY HPCC KVM SGI Octane III Brawny GPU cluster Microway DIY Web Server/ Linux App Server Dell 5400 Windows App Server HP 8740w HP 8740w Lenovo 400s Dual Quadcore 48GB memory *2 Nvidia C2050*2 8 B storage Dual 8-core 128GB memory Nvidia GX itan Intel Xeon Phi 3120A 8 B storage Dual-core 8GB memory Nvidia GX itan 3 B storage Dual Quadcore 16GB memory Nvidia Quadro B storage Quadcore 8 GB memory Nvidia Quadro 5000m Wimmy GPU cluster Dell 7500 Dell 7500 Dell 5400 DIY Dual 6-core 24 GB memory Nvidia Quadro 6000 Dual 6-core 24 GB memory Nvidia GX 480 Dual Quadcore 16GB memory Nvidia FX3700*2 Quadcore (Haswell) 16 GB memory AMD/AI 7970