Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.
Comment: Updated node count in UI-GPU due to rental

Table of Contents

...

If your job does not use the system openmpi, or does not use MPI, then any desired core binding will need to be set up with whatever mechanism the software uses. Otherwise, there will be no core binding. Again, that may not be a major issue. If your job does not work well with HT then run on a number of cores equal to half of the number of slots requested and the OS scheduler will minimize contention. 

new SGE utilities

While SoGE is very similar to previous versions of SGE there are some new utilities that people may find of interest. There are manual pages for each of these.

...

Table plus
sortColumn1
columnTypesS,S,I,I
QueueNode DescriptionQueue ManagerSlotsTotal memory (GB)
ANTH(3) standard memoryAndrew Kitchen168384

ARROMA

(8) standard memoryJun Wang4481024
AS(5) mid memory

Katharine Corum

2801280

BARNHART

(1) standard memory

William Barnhart

56128

BIGREDQ

(4) mid memory

Sara Mason

2241024

BIOLOGY

(1) mid memory

Matthew Brockman

56256

BIOSTAT

(1) standard memoryGrant Brown56128
CCOM(18) high memory
5 running jobs per user 

Boyd Knosp

10089216
CCOM-GPU(2) high memory with P100 accelerator

Boyd Knosp

1121024

CGRER

(10) standard memory

Jeremie Moen

5601280
CHEMISTRY(3) mid memory

JJ Urich

168768

CLAS-INSTR

(1) mid memory

JJ Urich

56256
CLL(5) standard memory

Mark Wilson
Brian Miller 

280640
COE(8) mid memory

Matt McLaughlin

4482048

DARBROB

(1) mid memory

Benjamin Darbro

56256

FLATTE

(2) high memory
(3) standard memory 

Michael Flatte

2801408

FLUIDSLAB

(8) standard memory

Mark Wilson
Brian Miller

4481024
GV(2) mid memory

Mark Wilson
Brian Miller

112512
HJ(10) standard memory
(1) high memory with P100 accelerator 
Hans Johnson6161280
IFC(10) mid memory 

Mark Wilson
Brian Miller

5602560
IIHG(1) mid memory

Diana Kolbe

560256

INFORMATICS

(12) mid memoryBen Rogers6723072

INFORMATICS-GPU

(2) mid memory with P100 acceleratorBen Rogers112512

INFORMATICS-HM

(1) high memoryBen Rogers56512
IVR(4) mid memory
(1) high memory 

Todd Scheetz

2801536
IWA(11) standard memory

Mark Wilson
Brian Miller

6161408
JM(1) high memory
(1) mid memory with P100 accelerator 

Jake Michaelson

1121024
JP(2) high memory

Virginia Willour

1121024
MANSCI(1) standard memory

Qihang Lin

112128

MORL

(10) mid memory

Ann Black-Ziegelbein

5602560
REX(4) standard memory

Mark Wilson
Brian Miller

224512
REXHI(1) high memory

Mark Wilson
Brian Miller

56512
SB(4) standard memory

Scott Baalrud

224512
UDAY(4) standard memory

Mark Wilson
Brian Miller

224512
UI(15) mid memory 8403840

UI-DEVELOP

(1) mid memory
(1) mid memory with P100 accelerator
 112512
UI-GPU

(3) mid memory with P100 accelerator

Note: Only two nodes available from 3/27 to 4/3/2017

 2241024
UI-HM(3) high memory 1681536
UI-MPI(19) mid memory 10644864
all.q

(76) standard memory
(111) mid memory
(7) mid memory with P100
(31) high memory
(3) high memory with P100
(1) high memory with K80 

 1282457856
NEUROSURGERY(1) high memory with K80

Haiming Chen

56512
SEMI(1) standard memory

Craig Pryor

56128
ACB(1) mid memoryAdam Dupuy56256

...

There are many resources that SGE keeps track of and most of them can be used in job submissions. However, the resource designations for machines based on memory and GPU are more likely to be used in practice. For the most part, machines with different amounts of memory and GPU capability are segregated by queues. However, the all.q queue contains all machines and when running jobs in that queue it may be desirable to request specific machine types. The following table lists these out. They would be selected with the '-l resource' flag to qsub. These are all Booleans.

Full Resource NameShortcut Resource Name
std_memsm
mid_memmm
high_mem

hm

gpugpu
gpu_k80k80
gpu_p100p100

For example, if you run a job in the all.q queue and want to use a node with a GPU, but do not care which type,

...