| Machine type | RISC-based ccNUMA system. |
|---|---|
| Models | HP 9000 SuperDome. |
| Operating system | HP-UX (HP's usual Unix flavour) |
| Connection structure | Crossbar |
| Compilers | Fortran 77, Fortran 90, Parallel Fortran, HPF, C, C++ |
| Vendors information Web page | http://www.hp.com/products1/servers/integrity/superdome_high_end/ |
| Year of introduction | 2000. |
System parameters:
| Model | Integrity SuperDome |
|---|---|
| Clock cycle | 1.5 GHz |
| Theor. peak performance | |
| Per proc. (64-bits) | 6 Gflop/s |
| Maximal (64-bits) | 768 Gflop/s |
| Main memory | |
| Memory/node | ≤ 128 GB |
| Memory/maximal | 1 TB |
| No. of processors | ≤ 128 |
| Communication bandwidth | |
| aggregate (global) | 64 GB/s |
| (cell—backplane) | 8 GB/s |
| (within cell, see below) | 16 GB/s |
Remarks:
The Integrity Superdome is HP's investment in the future for high-end servers.
Within a timespan of a few years it should replace the PA-RISC-based HP 9000
Superdome. HP has anticipated on this by giving it exactly the same macro
structure: cells are connected to a backplane crossbar that enables the
communication between cells. For the backplane it is immaterial whether a cell
contains PA-RISC or Itanium processors. For a discussion of the macro
architecture see the Remarks part of the HP 9000
Superdome. Also within a cell the structure of the system is very similar:
an identical Cell Controller ASIC as in the HP 9000 Superdome is the central
hub that controls local memory access, I/O requests, and communication to the
other cells via the backplane.
The difference lies in the processor, the 1.5 GHz Itanium 2. HP managed to make
a special processor socket that allows 2 Itaniums to be placed in one socket.
Therefore 8 Itaniums in total can be housed in the four sockets on the cell
board. As each Itanium contains 2 floating-point
units that are able to execute a combined floating multiply-add
instruction, in favourable circumstances 4 flops/cycle can be
achieved and a Theoretical Peak Performance of 6 Gflop/s per processor can
be attained. This amounts to a peak speed of 768 Gflop/s for a full
configuration.
Because the similarity between the both Superdome systems is so large, the
Integrity Superdome has the same ccNUMA characteristics as its relative.
It therefore supports OpenMP over its maximum of 128 processors.
As the Integrity Superdome is based on the Itanium 2 for which much Linux
development is done in the past few years, the system can also be run with the
Linux OS. In fact, because the machine can be partitioned, it is possible to run
both Linux and HP-UX in the different complexes of the same machine. The remark
as for the HP 9000 Superdome applies: in different complexes one can employ
cells with different types of processors, making the system a hybrid Integrity
and HP 9000 Superdome.
Measured Performances:
In [42] a speed of 1049 Gflop/s is
reported for solving a full linear system of unspecified size. This result is
achieved on a complex of systems with a total of 220 processors. As the
Theoretical Peak Performance of such a cluster is 1320 Gflop/s the efficiency
is 79%.