超级军网学霸进来看看科普一下--为啥子莫大的HPC核心数量这么少 ...
来源：百度文库 编辑：超级军网 时间：2024/04/29 07:30:28
http://www.graph500.org/june2011.html
==============
这个是啥子意思？ 有啥具体意义？
和另外一个测试有啥子区别？
--
Brief Introduction
Data intensive supercomputer applications are increasingly important for HPC workloads, but are ill-suited for platforms designed for 3D physics simulations. Current benchmarks and performance metrics do not provide useful information on the suitability of supercomputing systems for data intensive applications. A new set of benchmarks is needed in order to guide the design of hardware architectures and software systems intended to support such applications and to help procurements. Graph algorithms are a core part of many analytics workloads.
Backed by a steering committee of over 30 international HPC experts from academia, industry, and national laboratories, Graph 500 will establish a set of large-scale benchmarks for these applications. The Graph 500 steering committee is in the process of developing comprehensive benchmarks to address three application kernels: concurrent search, optimization (single source shortest path), and edge-oriented (maximal independent set). Further, we are in the process of addressing five graph-related business areas: Cybersecurity, Medical Informatics, Data Enrichment, Social Networks, and Symbolic Networks.
This is the first serious approach to complement the Top 500 with data intensive applications. Additionally, we are working with the SPEC committee to include our benchmark in their CPU benchmark suite. We anticipate the list will rotate between ISC and SC in future years.
The Graph 500 was announced at ISC2010 and the first list appeared at SC2010.
--
-
Complete Results - June 2011
Rank Machine Owner Problem Size TEPS
1 Intrepid (BG/P, 32768 nodes/ 131072 cores) ANL 38 18,508,000,000
2 Jugene (IBM, 32k nodes) Forschungszentrum Jülich 38 18,416,700,000
3 Lomonosov (MPP, 4096 nodes/ 8192 cores) Moscow State University 37 43,471,500,000
4 Hopper (XE6, 43200 cores, 1800 nodes) LBL 37 25,075,200,000
5 Franklin (XT4, 4000 nodes/ 16000 cores) LBL 36 19,955,100,000
6 Lonestar (Dell PowerEdge M610, 512 nodes/ 6144 cores) TACC 34 8,080,000,000
7 Kraken (Appro, 6128 cores/ Fusion I/O) LLNL 34 55,948,453
8 Red Sky (Sun/Red Sky, 512 nodes/ 4096 cores) SNL 33 9,470,000,000
9 Endeavor (Westmere X5670, 256 processors/ 3072 cores) Intel 33 6,860,000,000
10 SGI Altix UV 1000 (2048 cores) SGI 32 (Toy MPI Simple) 10,161,300,000
11 BlueGene/P (2048 nodes/ 8192 cores) Moscow State University 32 6,930,560,000
12 Blacklight (Altix UV 1000, 512 processors) PSC 32 4,452,270,000
13 BlueGene/Q Prototype (512 nodes) IBM Research, T.J. Watson 31 (Small) 11,323,000,000
14 DAS-4/VU (SuperMicro, 128 nodes) VU University 31 4,641,740,000
15 Matterhorn (Cray XMT, 64 nodes) CSCS 31 884,587,699
16 Kraken (Appro, 6128 cores) LLNL 31 104,644,000
17 SuperDragon-1 (Sugon, 32 nodes, 384 cores) Institute of Computing Technology, Beijing, China 30 1,454,110,000
18 Jaguar (XT5-HE, 224256 processors) ORNL 30 (Mini+) 1,010,500,000
19 Knot (HP MPI Cluster, 64 cores/ 8 processors) UCSB 30 176,627,000
20 cougarxmt (XMT, 128 nodes) PNL 29 1,223,375,650
21 graphstorm (XMT, 128 nodes) SNL 29 (Mini) 1,171,052,667
22 Matterhorn (Cray XMT, 64 nodes) CSCS 29 (Mini) 879,361,094
23 Erdos (XMT, 64-MTA) ORNL 29 701,767,082
24 Appro (7 nodes/ 84 cores) SDSC 29 (Mini) 29,994,123
25 Vortex (Convey XC-1ex, 1 node) Convey Computer Corporation 27 773,000,000
26 Westmere E7-4870 2.4GHz (quad socket 10-core, 40 cores) Intel Research 27 705,000,000
27 Minerva (IBM iDataPlex, 3096 processors) University of Warwick 26 839,444,000
28 Neumann (HPC Systems, 32 cores) UCSB 26 39,566,000
未命名.jpg (50.69 KB, 下载次数: 0)
下载附件 保存到相册
未命名.jpg (48.36 KB, 下载次数: 0)
下载附件 保存到相册
http://www.graph500.org/june2011.html
==============
这个是啥子意思？ 有啥具体意义？
和另外一个测试有啥子区别？
--
Brief Introduction
Data intensive supercomputer applications are increasingly important for HPC workloads, but are ill-suited for platforms designed for 3D physics simulations. Current benchmarks and performance metrics do not provide useful information on the suitability of supercomputing systems for data intensive applications. A new set of benchmarks is needed in order to guide the design of hardware architectures and software systems intended to support such applications and to help procurements. Graph algorithms are a core part of many analytics workloads.
Backed by a steering committee of over 30 international HPC experts from academia, industry, and national laboratories, Graph 500 will establish a set of large-scale benchmarks for these applications. The Graph 500 steering committee is in the process of developing comprehensive benchmarks to address three application kernels: concurrent search, optimization (single source shortest path), and edge-oriented (maximal independent set). Further, we are in the process of addressing five graph-related business areas: Cybersecurity, Medical Informatics, Data Enrichment, Social Networks, and Symbolic Networks.
This is the first serious approach to complement the Top 500 with data intensive applications. Additionally, we are working with the SPEC committee to include our benchmark in their CPU benchmark suite. We anticipate the list will rotate between ISC and SC in future years.
The Graph 500 was announced at ISC2010 and the first list appeared at SC2010.
--
-
Complete Results - June 2011
Rank Machine Owner Problem Size TEPS
1 Intrepid (BG/P, 32768 nodes/ 131072 cores) ANL 38 18,508,000,000
2 Jugene (IBM, 32k nodes) Forschungszentrum Jülich 38 18,416,700,000
3 Lomonosov (MPP, 4096 nodes/ 8192 cores) Moscow State University 37 43,471,500,000
4 Hopper (XE6, 43200 cores, 1800 nodes) LBL 37 25,075,200,000
5 Franklin (XT4, 4000 nodes/ 16000 cores) LBL 36 19,955,100,000
6 Lonestar (Dell PowerEdge M610, 512 nodes/ 6144 cores) TACC 34 8,080,000,000
7 Kraken (Appro, 6128 cores/ Fusion I/O) LLNL 34 55,948,453
8 Red Sky (Sun/Red Sky, 512 nodes/ 4096 cores) SNL 33 9,470,000,000
9 Endeavor (Westmere X5670, 256 processors/ 3072 cores) Intel 33 6,860,000,000
10 SGI Altix UV 1000 (2048 cores) SGI 32 (Toy MPI Simple) 10,161,300,000
11 BlueGene/P (2048 nodes/ 8192 cores) Moscow State University 32 6,930,560,000
12 Blacklight (Altix UV 1000, 512 processors) PSC 32 4,452,270,000
13 BlueGene/Q Prototype (512 nodes) IBM Research, T.J. Watson 31 (Small) 11,323,000,000
14 DAS-4/VU (SuperMicro, 128 nodes) VU University 31 4,641,740,000
15 Matterhorn (Cray XMT, 64 nodes) CSCS 31 884,587,699
16 Kraken (Appro, 6128 cores) LLNL 31 104,644,000
17 SuperDragon-1 (Sugon, 32 nodes, 384 cores) Institute of Computing Technology, Beijing, China 30 1,454,110,000
18 Jaguar (XT5-HE, 224256 processors) ORNL 30 (Mini+) 1,010,500,000
19 Knot (HP MPI Cluster, 64 cores/ 8 processors) UCSB 30 176,627,000
20 cougarxmt (XMT, 128 nodes) PNL 29 1,223,375,650
21 graphstorm (XMT, 128 nodes) SNL 29 (Mini) 1,171,052,667
22 Matterhorn (Cray XMT, 64 nodes) CSCS 29 (Mini) 879,361,094
23 Erdos (XMT, 64-MTA) ORNL 29 701,767,082
24 Appro (7 nodes/ 84 cores) SDSC 29 (Mini) 29,994,123
25 Vortex (Convey XC-1ex, 1 node) Convey Computer Corporation 27 773,000,000
26 Westmere E7-4870 2.4GHz (quad socket 10-core, 40 cores) Intel Research 27 705,000,000
27 Minerva (IBM iDataPlex, 3096 processors) University of Warwick 26 839,444,000
28 Neumann (HPC Systems, 32 cores) UCSB 26 39,566,000