Welcome to MilkyWay@home

Get 'Waiting to Run' when running a 'MilkyWay@Home N-Body Simulation 1.18 (mt)' WU


Advanced search

Message boards : Number crunching : Get 'Waiting to Run' when running a 'MilkyWay@Home N-Body Simulation 1.18 (mt)' WU
Message board moderation

To post messages, you must log in.

AuthorMessage
Steve

Send message
Joined: 31 Jan 13
Posts: 1
Credit: 12,758,436
RAC: 4,454
10 million credit badge9 year member badge
Message 59001 - Posted: 20 Jun 2013, 21:14:54 UTC

Hi,

I have a 6 core AMD processor, and a Radeon 7850 GPU. I have my app_config file setup to run 2 GPU tasks at once, as I find I get a greater throughput that way.

<app_config>
<app>
<name>milkyway</name>
<max_concurrent>8</max_concurrent>
<gpu_versions>
<gpu_usage>0.50</gpu_usage>
<cpu_usage>0.05</cpu_usage>
</gpu_versions>
</app>
<app>
<name>milkyway_separation__modified_fit</name>
<max_concurrent>7</max_concurrent>
<gpu_versions>
<gpu_usage>1</gpu_usage>
<cpu_usage>0.05</cpu_usage>
</gpu_versions>
</app>
</app_config>



The issue I am having is that while my 2 GPU tasks are running as well as expected, my CPU task (which is running a MilkyWay@Home N-Body Simulation 1.18 (mt) WU on all 6 CPU's) almost constantly is set to the state 'Waiting to Run (6 CPUs), and sits there suspended.

By snoozing the GPU tasks, the CPU tasks restart. If I resume the GPU tasks, the CPU task may either hit the 'Waiting to run' state immediately, or progress for a couple of minutes before hitting the 'Waiting to run' state.

I don't recall getting this issue until the mt WUs came along.

What can I do to get my CPU and GPU tasks to run simultaneously?
ID: 59001 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profilemikey
Avatar

Send message
Joined: 8 May 09
Posts: 2817
Credit: 463,357,032
RAC: 11,950
300 million credit badge13 year member badgeextraordinary contributions badge
Message 59012 - Posted: 21 Jun 2013, 11:26:16 UTC - in response to Message 59001.  

Hi,

I have a 6 core AMD processor, and a Radeon 7850 GPU. I have my app_config file setup to run 2 GPU tasks at once, as I find I get a greater throughput that way.

<app_config>
<app>
<name>milkyway</name>
<max_concurrent>8</max_concurrent>
<gpu_versions>
<gpu_usage>0.50</gpu_usage>
<cpu_usage>0.05</cpu_usage>
</gpu_versions>
</app>
<app>
<name>milkyway_separation__modified_fit</name>
<max_concurrent>7</max_concurrent>
<gpu_versions>
<gpu_usage>1</gpu_usage>
<cpu_usage>0.05</cpu_usage>
</gpu_versions>
</app>
</app_config>



The issue I am having is that while my 2 GPU tasks are running as well as expected, my CPU task (which is running a MilkyWay@Home N-Body Simulation 1.18 (mt) WU on all 6 CPU's) almost constantly is set to the state 'Waiting to Run (6 CPUs), and sits there suspended.

By snoozing the GPU tasks, the CPU tasks restart. If I resume the GPU tasks, the CPU task may either hit the 'Waiting to run' state immediately, or progress for a couple of minutes before hitting the 'Waiting to run' state.

I don't recall getting this issue until the mt WUs came along.

What can I do to get my CPU and GPU tasks to run simultaneously?


I believe they already are, here at MW the I am pretty sure gpu units use the cpu cores to assist the gpu in its crunching, meaning a unit doesn't crunch on the gpu alone. And it will use multiple cpu cores to make it crunch even faster.
ID: 59012 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Alinator

Send message
Joined: 7 Jun 08
Posts: 464
Credit: 56,639,936
RAC: 0
50 million credit badge13 year member badge
Message 59014 - Posted: 21 Jun 2013, 13:04:42 UTC
Last modified: 21 Jun 2013, 13:05:27 UTC

The issue here is due to the way the CC handles the scheduling of MT tasks on the CPU.

The problem is since you have your General prefs set to use all the cores, BOINC won't start the MT task until it has all cores available to use on the task. If ANYTHING needs/must use the CPU, it will suspended the MT task until it can have all of it again.

Unfortunately, you cannot use app_config to modify this behavior at this time.

So your options are:

1.) Change your general prefs to not use all the cores.

2.) Override the general prefs locally to not use all the cores.

3.) Limit nBody to use fewer cores with an app_info file. Of course you'll have to set up all your other MW apps as well.
ID: 59014 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote

Message boards : Number crunching : Get 'Waiting to Run' when running a 'MilkyWay@Home N-Body Simulation 1.18 (mt)' WU

©2022 Astroinformatics Group