Cluster "Survey"

rwryne · October 4, 2012, 15:26

Hello:

Just wondering if we could do a quick survey for those who work in industry and utilize a cluster...

Specifically I am curious:

1) Generally, how big is your company (i.e. hundreds, thousands, tens of thousands)
2) How much access do you have to your cluster (direct to hardware, remote desktop, terminal, web-based, etc.)
3) How big is your cluster core wise (in general terms..i.e. ~30 cores, ~100 cores, etc)
4) How many cells is your typical problem & maximum size problem?

...and anything else you think might be relevant.

SergeAS · October 4, 2012, 16:10

1 - > 100
2 - direct access (terminal, web and remote desktop also) for department cluster and terminal access to institute cluster
3 - 80 cores department cluster and 400 cores institute cluster
4 - 2-10M typical mesh size, 50M Max mesh size

stuart23 · October 4, 2012, 23:11

1 - University, number of users of cluster < 10.
2 - SSH Terminal / SFTP. No X11 (waste of processor power!), therefore no desktop
3 - 2 x 100 core + 2 x 200 core clusters
4 - Research projects, so vary. 25M LES up to 120M RANS

More details on my webpage in sig

f-w · October 5, 2012, 13:58

1) 20 (cluster users:1)
2) Direct
3) 24
4) 5-15M

evcelica · October 9, 2012, 08:39

Just like in the locker room, I have the smallest one!

1) 1 employee
2) Direct
3) 12 real cores on 2 nodes (HT off) - 64 GB RAM
4) 1-10M

...
I was talking about my small corner locker, Jeez what were you guys thinking!

abdul099 · October 13, 2012, 06:15

1) ~550 employees, ~50 in our office, ~25 cluster users
2) SSH / display forwarding
3) 1x 1024 cores, 1x 256 cores. Similar equipment in other offices.
4) 500K to 120M, but there are too many different applications to say what's typical

daveatstyacht · October 20, 2012, 18:03

1) 4 employees, 2 users
2) SSH
3) 256 cores out of a cluster with 1024 cores
4) 1M to 35M, typically 4-6M. marine applications

Eike · October 25, 2012, 03:19

1) ~1300 employees, ~5 in our office, ~2 cluster users
2) direct to hardware if necessary, remote desktop, terminal
3) 192 cores, 16 nodes Dell cluster
4) 1-7M typical mesh size, 15M max mesh size

Steel industry, using STAR-CCM+

pre and post on Dell T7500 Workstation w/ 2x 300GB SAS-HDD, 2TB SATA-HDD, 8 cores @3,6GHz, 48 GB RAM (up to 192GB)

October 4, 2012, 15:26	Cluster "Survey"	#1
rwryne Senior Member Ryne Whitehill Join Date: Aug 2009 Posts: 312 Rep Power: 18	Hello: Just wondering if we could do a quick survey for those who work in industry and utilize a cluster... Specifically I am curious: 1) Generally, how big is your company (i.e. hundreds, thousands, tens of thousands) 2) How much access do you have to your cluster (direct to hardware, remote desktop, terminal, web-based, etc.) 3) How big is your cluster core wise (in general terms..i.e. ~30 cores, ~100 cores, etc) 4) How many cells is your typical problem & maximum size problem? ...and anything else you think might be relevant.

October 4, 2012, 16:10		#2
SergeAS Member Serge A. Suchkov Join Date: Oct 2011 Location: Moscow, Russia Posts: 74 Blog Entries: 5 Rep Power: 14	1 - > 100 2 - direct access (terminal, web and remote desktop also) for department cluster and terminal access to institute cluster 3 - 80 cores department cluster and 400 cores institute cluster 4 - 2-10M typical mesh size, 50M Max mesh size __________________ OpenHyperFLOW2D Project

October 13, 2012, 06:15		#6
abdul099 Senior Member Join Date: Oct 2009 Location: Germany Posts: 636 Rep Power: 21	1) ~550 employees, ~50 in our office, ~25 cluster users 2) SSH / display forwarding 3) 1x 1024 cores, 1x 256 cores. Similar equipment in other offices. 4) 500K to 120M, but there are too many different applications to say what's typical __________________ We do three types of jobs here: GOOD, FAST AND CHEAP You may choose any two!

October 20, 2012, 18:03		#7
daveatstyacht Senior Member Dave Join Date: Jul 2010 Posts: 100 Rep Power: 15	1) 4 employees, 2 users 2) SSH 3) 256 cores out of a cluster with 1024 cores 4) 1M to 35M, typically 4-6M. marine applications Ralph M likes this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Preformance of Ansys CFX on a Linux Cluster	roued	CFX	3	September 30, 2012 07:07
Parallel rerun in cluster	Andy_bm	OpenFOAM Running, Solving & CFD	4	November 27, 2011 07:16
Parallel cluster solving with OpenFoam? P2P Cluster?	hornig	OpenFOAM Programming & Development	8	December 5, 2010 16:06
Linux Cluster Performance with a bi-processor PC	M.	FLUENT	1	April 22, 2005 09:25
To Jonas: Linux Cluster	Mark Render	Main CFD Forum	1	June 20, 2002 04:19

October 4, 2012, 23:11		#3
stuart23 Senior Member Stuart Buckingham Join Date: May 2010 Location: United Kingdom Posts: 267 Rep Power: 25	1 - University, number of users of cluster < 10. 2 - SSH Terminal / SFTP. No X11 (waste of processor power!), therefore no desktop 3 - 2 x 100 core + 2 x 200 core clusters 4 - Research projects, so vary. 25M LES up to 120M RANS More details on my webpage in sig

October 5, 2012, 13:58		#4
f-w Senior Member Join Date: Apr 2009 Posts: 154 Rep Power: 17	1) 20 (cluster users:1) 2) Direct 3) 24 4) 5-15M

October 9, 2012, 08:39		#5
evcelica Senior Member Erik Join Date: Feb 2011 Location: Earth (Land portion) Posts: 1,167 Rep Power: 23	Just like in the locker room, I have the smallest one! 1) 1 employee 2) Direct 3) 12 real cores on 2 nodes (HT off) - 64 GB RAM 4) 1-10M ... I was talking about my small corner locker, Jeez what were you guys thinking!

October 25, 2012, 03:19		#8
Eike Member Eike Join Date: Sep 2011 Posts: 40 Rep Power: 14	1) ~1300 employees, ~5 in our office, ~2 cluster users 2) direct to hardware if necessary, remote desktop, terminal 3) 192 cores, 16 nodes Dell cluster 4) 1-7M typical mesh size, 15M max mesh size Steel industry, using STAR-CCM+ pre and post on Dell T7500 Workstation w/ 2x 300GB SAS-HDD, 2TB SATA-HDD, 8 cores @3,6GHz, 48 GB RAM (up to 192GB)