Presentation is loading. Please wait.

Presentation is loading. Please wait.

NC論2 1 ネットワークコンピューティング 論Ⅱ 平成 24 年度 後期 火曜 第2時限(10:40-1 2:10) 吉永 努(UEC )

Similar presentations


Presentation on theme: "NC論2 1 ネットワークコンピューティング 論Ⅱ 平成 24 年度 後期 火曜 第2時限(10:40-1 2:10) 吉永 努(UEC )"— Presentation transcript:

1 NC論2 1 ネットワークコンピューティング 論Ⅱ 平成 24 年度 後期 火曜 第2時限(10:40-1 2:10) 吉永 努(UEC ) yosinaga@is.uec.ac.jp

2 NC論2 2 内 容内 容 分散・並列処理計算機における相互結合ネッ トワークとその上でのメッセージ・ルーティ ング技法などについて学ぶ 資料 http://comp.is.uec.ac.jp/yoshinagalab/yoshinaga/dp2.html http://ceng.usc.edu/smart/presentations/archives/Appendi xE.ppt (253 slides, 13MB) http://booksite.mkp.com/9780123838728/references/appe ndix_f.pdf (P.118, 2MB) TA: 島 圭吾君 shima@comp.is.uec.ac.jp

3 3 References T. M. Pinkston and J. Duato: Interconnection Networks, Appendix E in Computer Architecture: A Quantitative Approach, 4 th Edition, Morgan Kaufmann publishers (2006). 5 th Edition, Morgan Kaufmann publishers (2011). J. Duato, S. Yalamanchili, L. Ni: Interconnection Networks- an Engineering Approach-, 第 2 版, Morgan Kaufmann publishers (2003) 富田眞治: 並列コンピュータ、昭晃堂 ( 1996 ) W.D. Dally, B. Towles: Principles and Practices of Interconnection Networks, Morgan Kaufmann publishers (2003)

4 NC論2 4 What is an interconnection Network? It is a programmable system that transports data between terminals, such as processors and memory. It is programmable in the sense that it makes different connections at different points. It is a system because it is composed of many components: buffers, channels, switches, and controls that works together to deliver data.

5 NC論2 5 Interconnection Network (1/2) P M Interconnection Network Multicomputer P M P M

6 NC論2 6 Interconnection Network (2/2) P M Interconnection Network UMA type shared memory multiprocessor It is also called dance-hall architecture. P M P M

7 NC論2 7 Trend Its performance is increasing with processor performance at a rate of 50% per year. Communication is a limiting factor in the performance of many modern systems. Buses have been unable to keep up with the bandwidth demand, and point-to-point interconnection networks are rapidly taking over.

8 NC論2 8 Computer Classifications (%) 2012/062011/062010/06 MPP18.617.414.8 Cluster81.482.284.8 Others 0.0 0.4 http://www.top500.org/ share of the TOP500 June, 2012 – June, 2010

9 NC論2 9 Examples of MPPs processorTopology #core Rmax K computer @RIKEN Fujitsu 2011 SPARC64 VIIIfx 2GHz (16GFlops× 8 cores) 6D mesh/ 3D torus Tofu interconnect 80K-node x 8-core = 640K-core 10.51PFlops 7890KW Sequoia@LLNL IBM BlueGene/Q 2011 Power BQC 16C 1.6 GHz ( 16 cores) 5D torus SeaStar interconnect 1,572,864-core 16.32PFlops 12660KW

10 NC論2 10 Examples of clusters processors GPUInterconnect Tianhe-1A ( 天河一号 A ) China 2010 Intel EM64T Xeon X5670 2.93 GHz (11.72 GFlops) ×14,336 NVIDIA Tesla M2050 (515GFlops) ×7,168 Galaxy 160Gbps/link (proprietary) Fat tree Tsubame 2.0 Tokyo Tech. 2010 Xeon X5670 2.93GHz×1,408 + Xeon E7520 2GHz×34 Tesla M2050 (515GFlops) ×1,048×3 Infiniband QDR (40Gbps) ×2 Fat tree

11 NC論2 11 Other Networks of Supercomputers Cray XE6 (2011): 3D torus, proprietary GEMINI link) Pleiades / NASA (2011): partial 11D hypercube topology with IB QDR/DDR Red Sky/ Sandia National Lab. (2010): 3D torus (12 bristled node) with IB QDRswitches IBM Roadrunner (2009): fat-tree with IB DDR Earth Simulator2 / NEC SX-9E (2009): Fat-Tree (64GB/s/cpu, 8-CPU/node, 160 nodes) IBM Blue Gene/L (2004): 3D torus proprietary (64 x 32 x 32 = 64K nodes)

12 NC論2 12 Architecture vs. software memoryprogramming UMA (SMP)sharedOpenMP NUMA (MPP) distributed (not shared) MPI (Message Passing Interface)

13 NC論2 13 Network Design (1/3) Performance: latency and throughput (bandwidth) Scalability: #processors vs. network, memory, I/O bandwidth Incremental expandability: small to maximum size Partitionability: netwrok may be partitioned for several users

14 NC論2 14 Network Design (2/3) Simplicity: simple design, higher clock frequency, easy to use Distance span: smaller system is preferred for noise and cable delay, etc. Physical constraints: packaging (pin count), wiring(wire length), and maintenance (power consumption) should meet physical limitation.

15 NC論2 15 Network Design (3/3) Reliability: fault tolerant, reliable communication, hot swap Expected workload: robust performance over a wade range of traffic conditions. Cost: trade-offs between cost and performance.

16 NC論2 16 Classifiction of Interconnection Networks Shared-Medium Networks –Local area networks (ethernet, token ring) –Backplane bus (e.g. SUN Gigaplane) Direct Networks (router-based) –mesh, torus, hypercube, tree, … etc. Indirect Networks (switch-based) Hybrid Networks

17 NC論2 17 Shared-Medium Networks (LAN) Arbitration that determines the mastership of the shared-medium network to resolve network access is needed. The most well-known protocol is carrier-sense multiple access with collision detection (CSMA/CD). Token bus and token ring pass a token from the owner which has the right to access the bus/ring and resolve nondeterministic waiting time.

18 NC論2 18 Shared-Medium Networks (Backplane bus) It is commonly used to interconnect processor(s) and memory modules to provide SMP (Symmetrical Memory Processor) architecture. It is realized by printed lines on a circuit board by discrete wiring. Gigaplane in SUN Enterprise x000 server(1996): 2.6GB/s, 256 bits data, 42 bits address, 83.8MHz clock.

19 NC論2 19 Direct (static) Networks Consists of a set of nodes. Each node is directly connected to a subset of other nodes in the network. Examples: –2D mesh (intel Paragon), 3D mesh (MIT J-Mahine) –2D torus (Fujitsu AP3000), 3D torus (Cray T3D, T3E) –Hypercube (CM1, CM2, nCUBE)

20 NC論2 20 Mesh topology 2D 3D node

21 NC論2 21 Torus topology 2D (4-ary 2-cube) 3D (3-ary 3-cube)

22 NC論2 22 Hypercube (binary n-cube) 4D (2-ary 4-cube)

23 NC論2 23 tree Binary treefat tree x tree

24 NC論2 24 Hierarchical topology (1/2) Pyramid (Hierarchical 2D mesh) Hierarchical ring

25 NC論2 25 Hierarchical topology (2/2) Cube-connected cycles RDT (Recursive Diagonal Torus)

26 NC論2 26 Hypermesh (spaninng-bus hypercube) Single or multiple buses

27 NC論2 27 Base-m n-cube (hyper-crossbar) Base-8 3-cube (Toshiba Prodigy) 000007 070077 707 777 770 8x8 crossbar

28 NC論2 28 Diameter and degrees (1/2) 2D mesh 2D torus 3D torus binary n-cube #node N NNN = 2 n Diameter 2√N2√N √N√N √N√N log N degree 446 log N 3

29 NC論2 29 Diameter and degrees (2/2) Base-m n-cube CCCBinary tree ring #node N = m n N = n2 n NN Diameter log m N3n/22 log NN/2 degree log m N332 3


Download ppt "NC論2 1 ネットワークコンピューティング 論Ⅱ 平成 24 年度 後期 火曜 第2時限(10:40-1 2:10) 吉永 努(UEC )"

Similar presentations


Ads by Google