| |
The Sushi Cluster
|
|
|
We have a very nice cluster of machines to run our simulations on. These
were very generously donated by Intel.
The Sushi cluster consists of 20 dual-processor nodes. Each
node consists of two 3.2 GHz P4's. We have Hyperthreading turned on,
and so we have a total of 40 physical (80 virtual) CPUs to be abused by
our simulation runs.
We have also added twelve more nodes, each
containing dual Core 2 Duo processors (four cores per node),
each running at 3.0GHz. The higher per-core IPC and true multi-core
(as opposed to Hyperthreading) effectively more than doubles our
available compute cycles.
The latest additions include four more nodes, each with dual
Core 2 Quad (eight cores per node), each running at 3.0GHz
(these are the ``Penryn'' 65nm versions with 6MB L2 cache per two
cores).
The original menu includes:
| Sushi1: Aji |  |
Sushi2: Anago |  |
| Sushi3: Awabi |  |
Sushi4: Ebi |  |
| Sushi5: Hamachi |  |
Sushi6: Hirame |  |
| Sushi7: Hokkigai |  |
Sushi8: Hotategai |  |
| Sushi9: Ika |  |
Sushi10: Ikura |  |
| Sushi11: Kani |  |
Sushi12: Maguro |  |
| Sushi13: Saba |  |
Sushi14: Sake |  |
| Sushi15: Suzuki |  |
Sushi16: Tai |  |
| Sushi17: Tako |  |
Sushi18: Tobiko |  |
| Sushi19: Unagi |  |
Sushi20: Uni |  |
The 2nd-generation additions to the menu include:
| Sushi21: Akagai |  |
Sushi22: Aoyagi |  |
| Sushi23: Hamaguri |  |
Sushi24: Iwashi |  |
| Sushi25: Katsuo |  |
Sushi26: Kohada |  |
| Sushi27: Masago |  |
Sushi28: Mirugai |  |
| Sushi29: Sayori |  |
Sushi30: Shako |  |
| Sushi31: Shirauo |  |
Sushi32: Torigai |  |
The 3rd-generation additions to the menu include:
| Sushi33: Amaebi |  |
Sushi34: Ankimo |  |
| Sushi35: Kanpachi |  |
Sushi36: Sawara |  |
|
|
|