- Infos im HLRS Wiki sind nicht rechtsverbindlich und ohne Gewähr -

Difference between revisions of "NEC Cluster Hardware and Architecture (laki + laki2)"

From HLRS Platforms
m (added some links for CPUs)
Line 21: Line 21:
  
 
*'''Visualisation node'''
 
*'''Visualisation node'''
** 5 nodes each with 8 cores Intel W3540 and 24GB memory (4 for laki and 1 for laki2)
+
** 5 nodes each with 8 cores Intel [http://ark.intel.com/de/products/39719/Intel-Xeon-Processor-W3540-8M-Cache-2_93-GHz-4_80-GTs-Intel-QPI W3540] and 24GB memory (4 for laki and 1 for laki2)
 
*** Nvidia Quadro FX5800
 
*** Nvidia Quadro FX5800
  
 
* '''Node Upgrades''' (2012/2013)
 
* '''Node Upgrades''' (2012/2013)
** 328 nodes Dual Intel [[Sb|'Sandy Bridge']] E5-2670 (204 for laki and 124 for laki2)
+
** 328 nodes Dual Intel [[Sb|'Sandy Bridge']] [http://ark.intel.com/de/products/64595/Intel-Xeon-Processor-E5-2670-20M-Cache-2_60-GHz-8_00-GTs-Intel-QPI E5-2670] (204 for laki and 124 for laki2)
 
*** 2.6 Ghz, 8 Cores per processor, 16 Threads
 
*** 2.6 Ghz, 8 Cores per processor, 16 Threads
 
*** 4 memory channels per processor, DDR3 1600Mhz memory
 
*** 4 memory channels per processor, DDR3 1600Mhz memory
Line 33: Line 33:
  
 
* '''Node Upgrades''' (2014/2015)
 
* '''Node Upgrades''' (2014/2015)
** 80 nodes Dual Intel [[hsw|'Haswell']] E5-E5-2660 v3
+
** 80 nodes Dual Intel [[hsw|'Haswell']] [http://ark.intel.com/de/products/81706/Intel-Xeon-Processor-E5-2660-v3-25M-Cache-2_60-GHz E5-2660v3]
 
*** 2.6 Ghz, 10 Cores per processor, 20 Threads
 
*** 2.6 Ghz, 10 Cores per processor, 20 Threads
 
*** 4 memory channels per processor, DDR4 2133Mhz memory
 
*** 4 memory channels per processor, DDR4 2133Mhz memory
Line 41: Line 41:
  
 
* '''Node Upgrades''' (2016)
 
* '''Node Upgrades''' (2016)
** 96 nodes Dual Intel [[hsw|'Haswell']] E5-E5-2680 v3
+
** 96 nodes Dual Intel [[hsw|'Haswell']] [http://ark.intel.com/de/products/81908/Intel-Xeon-Processor-E5-2680-v3-30M-Cache-2_50-GHz E5-2680v3]
 
*** 2.5 Ghz, 12 Cores per processor, 20 Threads
 
*** 2.5 Ghz, 12 Cores per processor, 20 Threads
 
*** 4 memory channels per processor, DDR4 2133Mhz memory
 
*** 4 memory channels per processor, DDR4 2133Mhz memory
Line 48: Line 48:
  
 
* '''Additional [[Mem256gb|large memory nodes]]'''
 
* '''Additional [[Mem256gb|large memory nodes]]'''
** 10 nodes Quad Socket AMD Opteron 6238  
+
** 10 nodes Quad Socket AMD Opteron [http://products.amd.com/en-ca/search/CPU/AMD-Opteron%E2%84%A2/AMD-Opteron%E2%84%A2-6200-Series-Processor/6238/32 6238]
 
** 2.6 Ghz, 12 cores per processor
 
** 2.6 Ghz, 12 cores per processor
 
** 4 memory channels per processor, DDR3 1600Mhz memory
 
** 4 memory channels per processor, DDR3 1600Mhz memory
Line 93: Line 93:
 
    
 
    
 
'''Features'''
 
'''Features'''
* Operating System: ScientificLinux 6.4 ''(internal test on Windows HPC Server 2008)''
+
* Operating System: ScientificLinux 6.4 ''(internal test was done on Windows HPC Server 2008)''
 
* Batchsystem: Torque/Moab
 
* Batchsystem: Torque/Moab
 
* node-node interconnect: Infiniband + GigE
 
* node-node interconnect: Infiniband + GigE
 
* Global Disk 240 TB (lustre) for laki + 110TB (lustre) for laki2
 
* Global Disk 240 TB (lustre) for laki + 110TB (lustre) for laki2
 
* Many Software Packages for Development
 
* Many Software Packages for Development

Revision as of 16:08, 14 June 2016

Hardware

  • ~150 compute are of type NEC HPC-144 Rb-1 Server (see NEC Products)
    • dual CPU compute nodes: 2x Intel Xeon X5560 "Gainestown" (5000 Sequence specifications)
      • 4 cores, 8 threads
      • 2.80 GHz (3.20 Ghz max. Turbo frequency)
      • 8MB L3 Cache
      • 1333 MHz Memory Interface, 6.4 GT/s QPI
      • TDP 95W, 45nm technology
      • "Nehalem" microarchitecture
    • compute node RAM: triple-channel memory
      • standard: 12 GB RAM
      • 20 nodes upgraded to 24GB, 48GB or 144GB RAM
        • 2 nodes of 144GB Memory nodes have additional a 6TB local scratch disk installed
        • 1 nodes of 144GB Memory nodes have additional a 2TB local scratch disk installed
    • 16 compute nodes have additional Nvidia Tesla S1070 GPU's installed.
  • Pre- & Postprocessing node (smp node)
    • 8x Intel Xeon X7542 6-core CPUs with 2.67GHz (8*6=48 Cores)
    • 1TB RAM
    • shared access
  • Visualisation node
    • 5 nodes each with 8 cores Intel W3540 and 24GB memory (4 for laki and 1 for laki2)
      • Nvidia Quadro FX5800
  • Node Upgrades (2012/2013)
    • 328 nodes Dual Intel 'Sandy Bridge' E5-2670 (204 for laki and 124 for laki2)
      • 2.6 Ghz, 8 Cores per processor, 16 Threads
      • 4 memory channels per processor, DDR3 1600Mhz memory
      • 36 nodes with 64GB RAM
      • 292 nodes with 32GB RAM
      • QDR Mellanox ConnectX-3 IB HCAs (40gbit)
  • Node Upgrades (2014/2015)
    • 80 nodes Dual Intel 'Haswell' E5-2660v3
      • 2.6 Ghz, 10 Cores per processor, 20 Threads
      • 4 memory channels per processor, DDR4 2133Mhz memory
      • 76 nodes with 128GB RAM (hsw128gb)
      • 4 nodes with 256GB RAM (hsw256gb)
      • QDR Mellanox ConnectX-3 IB HCAs (40gbit)
  • Node Upgrades (2016)
    • 96 nodes Dual Intel 'Haswell' E5-2680v3
      • 2.5 Ghz, 12 Cores per processor, 20 Threads
      • 4 memory channels per processor, DDR4 2133Mhz memory
      • 96 nodes with 128GB RAM (hsw128gb12c)
      • QDR Mellanox ConnectX-3 IB HCAs (40gbit)
  • Additional large memory nodes
    • 10 nodes Quad Socket AMD Opteron 6238
    • 2.6 Ghz, 12 cores per processor
    • 4 memory channels per processor, DDR3 1600Mhz memory
    • 256GB RAM
    • QDR Mellanox ConnectX-2 IB HCAs (40gbit)
    • 4 nodes have additional a 4TB local scratch disk



Architecture

The NEC Cluster platform (laki and laki2) consists of several frontend nodes for interactive access (for access details see Access) and several compute nodes of different types for execution of parallel programs.


Compute node types installed:

  • Intel Xeon 5560 (nehalem)
  • Intel Xeon E5-2670 (Sandy Bridge)
  • AMD Opteron 6238 (interlagos)
  • Nvidia Tesla S1070 (consisting of C1060 devices)
  • Large Memory nodes (144GB, 256GB)
  • Pre-Postprocessing node with very large memory (1TB)
  • Visualisation nodes with Nvidia Quadro FX5800
  • Different memory nodes (12GB, 24GB, 32GB, 48GB, 64GB)


Features

  • Operating System: ScientificLinux 6.4 (internal test was done on Windows HPC Server 2008)
  • Batchsystem: Torque/Moab
  • node-node interconnect: Infiniband + GigE
  • Global Disk 240 TB (lustre) for laki + 110TB (lustre) for laki2
  • Many Software Packages for Development