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 POWER7
    3. IBM BlueGene/Q processor
    4. Intel Xeon
    5. The SPARC processors
  8. Accelerators
    1. GPU accelerators
      1. ATI/AMD
      2. nVIDIA
    2. General computational accelerators
      1. Intel Xeon Phi
    3. FPGA accelerators
      1. Convey
      2. Kuberre
      3. SRC
  9. Interconnects
    1. Infiniband
Available systems
  • The Bull bullx system
  • The Cray XC30
  • The Cray XE6
  • The Cray XK7
  • The Eurotech Aurora
  • The Fujitsu FX10
  • The Hitachi SR16000
  • The IBM BlueGene/Q
  • The IBM eServer p775
  • The NEC SX-9
  • The SGI Altix UV series
  • Systems disappeared from the list
    Systems under development

    Machine type Distributed-memory multi-vector processor
    Models SX-9B, SX-9A, SX-9xMy, y = 2,…512
    Operating system Super-UX (Unix variant based on BSD V.4.3 Unix)
    Connection structure Multi-stage crossbar (see Remarks)
    Compilers Fortran 90, HPF, ANSI C, C++
    Vendors information Web page
    Year of introduction 2007

    System parameters:

    Model SX-9B SX-9A SX-9xMy
    Clock cycle 3.2 GHz 3.2 GHz 3.2 GHz
    Theor. peak performance      
    Per Proc. (64 bits) 102.4 Gflop/s 102.4 Gflop/s 102.4 Gflop/s
    Single frame: 819.2 Gflop/s 1.6 Tflop/s Gflop/s
    Multi frame: 838.9 Tflop/s
    Main memory, DDR2-SDRAM 256–512 GB 512–1024 GB ≤ 512 TB
    Main memory, FCRAM 128–256 GB 256–512 GB ≤ 256 TB
    No. of processors 4–8 8–16 32–8192


    The NEC SX-9 is a technology shrunken version of its predecessor the SX-8 (see Systems disappeared from the list). As a result the clock cycle has increased from 2.0 to 3.2 GHz, the density of processors/frame has doubled and the power consumption almost halved. The structure of the CPUs, however, has stayed the same. The SX-9 series is basically offered in three models as displayed in the table above. All models are based on the same processor, an 8-way replicated vector processor where each set of vector pipes contains a logical, mask, add/shift, multiply, and division pipe (see section \ref{s:sm-simd} for an explanation of these components). As multiplication and addition can be chained (but not division) and two of each are present, the peak performance of a pipe set at 3.2 GHz is 12.8 Gflop/s. Because of the 8-way replication a single CPU can deliver a peak performance of 102.4 Gflop/s. The official NEC documentation quotes higher peak performances because the peak performance of the scalar processor (rated at 6.4 Gflop/s, see below) is added to the peak performance of the vector processor to which it belongs. We do not follow this practice as a full utilisation of the scalar processor along with the vector processor in reality will be next to non-existent. The scalar processor that is 2-way super scalar and at 3.2 GHz has a theoretical peak of 6.4 Gflop/s. The peak bandwidth per CPU is 160 B/cycle. This is sufficient to ship 20 8-byte operands back or forth, enough to feed 5 operands every 2 cycles to each of the replicated pipe sets.

    Unlike from what one would expect from the naming the SX-9B is the simpler configuration of the two single-frame systems: it can be had with 4–8 processors but is in virtually all other respects equal to the larger SX-9A that can house 8–16 processors. There is one difference connected to the maximal amout of memory per frame: NEC now offers the interesting choice between the usual DDR2-SDRAM or FCRAM (Fast Cycle Memory. The latter type of memory can a factor of 2–3 faster than the former type of memory. However, because of the more complex structure of the memory, the density is about two times lower. Hence that in the system parameters table, the entries for FCRAM are about two times lower than for SDRAM. The lower bound for SDRAM in the SX-8A and SX-8B systems are the same: 32 GB. For the very memory-hungry applications that are usually run on vector-type systems, the availability of FCRAM can be beneficial for quite some of these applications.

    In a single frame of the SX-9A models fit up to 16 CPUs. Internally the CPUs in the frame are connected by a 1-stage crossbar with the same bandwidth as that of a single CPU system: 512 GB/s/port. The fully configured frame can therefore attain a peak speed of 1.6 Tflop/s.

    In addition, there are multi-frame models (SX-9xMy) where x = 8,...,8192 is the total number of CPUs and y = 2,...,512 is the number of frames coupling the single-frame systems into a larger system. There are two ways to couple the SX-9 frames in a multi-frame configuration: NEC provides a full crossbar, the so-called IXS crossbar to connect the various frames together at a speed of 128 GB/s for point-to-point unidirectional out-of-frame communication. When connected by the IXS crossbar, the total multi-frame system is globally addressable, turning the system into a NUMA system. However, for performance reasons it is advised to use the system in distributed memory mode with MPI.

    For distributed computing there is an HPF compiler and for message passing an optimised MPI (MPI/SX) is available. In addition for shared memory parallelism, OpenMP is available.

    Measured Performances:
    In [39] the SX-9/E based Earth Simulator shows a performance of 122.4 Tflop/s for a linear system of order N = 1,556,480, attaining an efficiency of 93%.