Graham: Difference between revisions
No edit summary |
(Marked this version for translation) |
||
| Line 4: | Line 4: | ||
<translate> | <translate> | ||
</noinclude> | </noinclude> | ||
===GP3 (Waterloo)=== | ===GP3 (Waterloo)=== <!--T:1--> | ||
====Introduction==== | ====Introduction==== <!--T:2--> | ||
<!--T:3--> | |||
GP3 system evaluation is not yet completed (as of early November 2016). Anticipated specifications, based on Waterloo's RFP and bids received, include the following. This information is NOT GUARANTEED and might not be complete. It is provided for planning purposes. | GP3 system evaluation is not yet completed (as of early November 2016). Anticipated specifications, based on Waterloo's RFP and bids received, include the following. This information is NOT GUARANTEED and might not be complete. It is provided for planning purposes. | ||
<!--T:4--> | |||
The parallel filesystem, interconnects and external persistent storage (NDC-Waterloo) will be the same as [[GP2]]. There is a slightly different mix of compute nodes. See next section for details. | The parallel filesystem, interconnects and external persistent storage (NDC-Waterloo) will be the same as [[GP2]]. There is a slightly different mix of compute nodes. See next section for details. | ||
====Node types and characteristics==== | ====Node types and characteristics==== <!--T:5--> | ||
<!--T:6--> | |||
{| class="wikitable sortable" | {| class="wikitable sortable" | ||
|- | |- | ||
| Line 25: | Line 28: | ||
|} | |} | ||
<!--T:7--> | |||
All of the above nodes will have local (on-node) storage. | All of the above nodes will have local (on-node) storage. | ||
<!--T:8--> | |||
Compute Canada is not currently able to disclose the specific CPU model or specifications. | Compute Canada is not currently able to disclose the specific CPU model or specifications. | ||
<!--T:9--> | |||
GP3 will have GPU's, but Compute Canada is not currently able to disclose specific details. The RFP used NVIDIA K80 as a baseline specification. | GP3 will have GPU's, but Compute Canada is not currently able to disclose specific details. The RFP used NVIDIA K80 as a baseline specification. | ||
<!--T:10--> | |||
The total GP3 system is anticipated to have over 25,000 cores. | The total GP3 system is anticipated to have over 25,000 cores. | ||
<!--T:11--> | |||
The delivery and installation schedule is not yet known. | The delivery and installation schedule is not yet known. | ||
====Name==== | ====Name==== <!--T:12--> | ||
<!--T:13--> | |||
For naming details of the new systems see [[Migration2016:New_System_Names]]. | For naming details of the new systems see [[Migration2016:New_System_Names]]. | ||
<!--T:14--> | |||
<noinclude> | <noinclude> | ||
</translate> | </translate> | ||
</noinclude> | </noinclude> | ||
Revision as of 14:47, 17 November 2016
GP3 (Waterloo)
Introduction
GP3 system evaluation is not yet completed (as of early November 2016). Anticipated specifications, based on Waterloo's RFP and bids received, include the following. This information is NOT GUARANTEED and might not be complete. It is provided for planning purposes.
The parallel filesystem, interconnects and external persistent storage (NDC-Waterloo) will be the same as GP2. There is a slightly different mix of compute nodes. See next section for details.
Node types and characteristics
| "Base" compute nodes | Over 500 nodes | 128GB of memory each |
| "Large" compute nodes | Over 100 nodes | 256GB of memory each |
| "Bigmem512" nodes | TBD | 512GB of memory each |
| "Bigmem3000" nodes | TBD | 3 TB of memory each |
All of the above nodes will have local (on-node) storage.
Compute Canada is not currently able to disclose the specific CPU model or specifications.
GP3 will have GPU's, but Compute Canada is not currently able to disclose specific details. The RFP used NVIDIA K80 as a baseline specification.
The total GP3 system is anticipated to have over 25,000 cores.
The delivery and installation schedule is not yet known.
Name
For naming details of the new systems see Migration2016:New_System_Names.