Graham: Difference between revisions

From Alliance Doc
Jump to navigation Jump to search
No edit summary
No edit summary
Line 70: Line 70:


<!--T:7-->
<!--T:7-->
All of the above nodes will have local (on-node) storage ranging from 300GB to 960GB SATA SSD drives.
All of the above nodes will have approximately 1TB of local (on-node) storage provided by SSD drives available in /tmp.
 
<!--T:8-->
Compute Canada is not currently able to disclose the specific CPU models or specifications.
 
<!--T:9-->
Graham 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 completed system is expected to have around 35,000 cores.


<!--T:11-->
<!--T:11-->

Revision as of 17:54, 12 January 2017

Other languages:


Graham (GP3)

GRAHAM is a heterogeneous cluster, suitable for a variety of workloads, and located at the University of Waterloo. It is named after Wes Graham, the first director of the Computing Centre at Waterloo. It was previously known as "GP3" and is still identified as such in the 2017 RAC documentation.

System evaluation is not yet completed as of 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 Cedar's. There is a slightly different mix of compute nodes.

Attached Storage System

$HOME

Standard home directory
Small, standard quota
Larger requests should be on $PROJECT

$SCRATCH
Parallel High-performance filesystem

Approximately 3PB usable capacity for temporary (/scratch) storage.
Aggregate performance of 35GB/s. Available to all nodes.
Not allocated
Purged - inactive data will be purged

$PROJECT
External persistent storage

Provided by the NDC.
Available to compute nodes, but not designed for parallel I/O workloads.

High Performance Interconnect

A low-latency high-performance fabric connecting all nodes and temporary storage.

The design of Graham is to support multiple simultaneous parallel jobs of up to 1024 cores in a fully non-blocking manner.

Node types and characteristics

"Base" compute nodes 800 nodes 16 cores/socket, 2 sockets/node, 128 GB of memory.
"Bigmem512" nodes 24 nodes 16 cores/socket, 2 sockets/node, 512 GB of memory.
"Bigmem3000" nodes 3 nodes 16 cores/socket, 4 sockets/node, 3 TB of memory.
"GPU" nodes 160 nodes 16 cores/socket, 2 sockets/node, 128 GB of memory, 2 NVIDIA P100 GPUs.
"Large" nodes 56 nodes 16 cores/socket, 2 sockets/node, 256 GB of memory.

All of the above nodes will have approximately 1TB of local (on-node) storage provided by SSD drives available in /tmp.

The delivery and installation schedule is not yet confirmed.


Retrieved from "https://docs.alliancecan.ca/mediawiki/index.php?title=Graham&oldid=7534"