The NEC Express5800/1000 series

Introduction
HPC Architecture
  1. Shared-memory SIMD machines
  2. Distributed-memory SIMD machines
  3. Shared-memory MIMD machines
  4. Distributed-memory MIMD machines
  5. ccNUMA machines
  6. Clusters
  7. Processors
    1. AMD Opteron
    2. IBM POWER6
    3. IBM PowerPC 970
    4. IBM BlueGene processors
    5. Intel Itanium 2
    6. Intel Xeon
    7. The MIPS processor
    8. The SPARC processors
  8. Accelerators
    1. GPU accelerators
    2. General accelerators
    3. FPGA accelerators
  9. Networks
    1. Infiniband
    2. InfiniPath
    3. Myrinet
    4. QsNet
Available systems
  1. The Bull NovaScale
  2. The C-DAC PARAM Padma
  3. The Cray XT3
  4. The Cray XT4
  5. The Cray XT5h
  6. The Cray XMT
  7. The Fujitsu/Siemens M9000
  8. The Fujitsu/Siemens PRIMEQUEST
  9. The Hitachi BladeSymphony
  10. The Hitachi SR11000
  11. The HP Integrity Superdome
  12. The IBM BlueGene/L&P
  13. The IBM eServer p575
  14. The IBM System Cluster 1350
  15. The Liquid Computing LiquidIQ
  16. The NEC Express5800/1000
  17. The NEC SX-9
  18. The SGI Altix 4000
  19. The SiCortex SC series
  20. The Sun M9000
Systems disappeared from the list
Systems under development
Glossary
Acknowledgments
References

Machine type Shared-memory ccNUMA system.
Models Express 5800 1160Xf, 1320Xf.
Operating system Linux, Windows Server 2008
Connection structure Crossbar
Compilers Fortran 95, HPF, ANSI C, C++
Vendors information Web page http://www.nec.co.jp/express/products/enterprise/index.html
Year of introduction 2006.

System parameters:

Model 1160Xf 1320Xf
Clock cycle 1.6 GHz 1.6 GHz
Theor. peak performance    
Per core (64 bits) 6.4 Gflop/s 6.4 Gflop/s
Maximal 204.8 Gflop/s 409.6 Gflop/s
No. of processors 16 32
Main memory ≤ 256 GB ≤ 512 GB
Communication bandwidth    
Point-to-point — —
Cell-to-cell (see remarks) 6.4 GB/s 6.4 GB/s
Aggregate 102.4 GB/s 102.4 GB/s

Remarks:

The Express5800 series is more or less a renaming of the earlier TX7 series. The structure of the system has stayed the same but instead of the former Itanium 2 processors the new machines are offered with the Montecito processors. It is another of the Itanium 2-based servers (see, e.g., also the Bull NovaScale, the SGI Altix 4000, the HP Integrity Superdome, and the Hitachi BladeSymphony) that recently appeared on the market. The largest configuration presently offered is the 1320Xf with 32 1.6 GHz Montecito processors. NEC had already some experience with Itanium servers offering 16-processor Itanium 1 servers under the name AsuzA and the already mentioned TX7 systems. So, the Express5800 systems can be seen as a third generation.

Processors are housed in 4-processor cells that connect via a flat crossbar. The bandwidth of the crossbar links is 6.4 GB/s. Unfortunately the documentation does not mention the bandwidth of the links between processors and memory within a cell. Although NEC still calls the machines SMP systems they are in fact ccNUMA systems with a low NUMA factor. The documentation speaks about "near-uniform high speed memory access".

Unlike the other vendors that employ the Itanium processors, NEC offers its own compilers including an HPF compiler which is probably available for compatibility with the software for the NEC SX-9 because it is hardly useful on a shared-memory system like the Express5800. The software also includes MPI and OpenMP.

Like the other Japanese vendors (see Fujitsu-Siemens PRIMEQUEST, Hitachi BladeSymphony) NEC very much emphasizes the RAS features of the system, targeting mission-critical operating environments.

Measured Performances:
As yet no performance results have been published for this system.