Hello EVGA Crunchers!
 
This is the official thread for April's Project of the month : Moo! Wrapper
What's Moo objective, why we crunch for them ? http://www.distributed.net/RC5
 
Date: 24 to 28 April 2017
 
Run the project should be easy, it runs both on CPU and GPU (AMD and Nvidia) and there are a lot of WUs ready to be sent.
 There are apps for Windows, Linux, Android and MacOS so everyone is welcome!
 
 
Our team is only 65th by RAC and 117th by Total credit. We should show to the other teams what our little but strong team could do in few days !!
 
 
The only real problem with this project is ..... NO BADGES
 
 
I'm not sure if HT should be on or off, I have only 2 Intel i5 and an AMD PII x3 720 so I can not try, sorry !
The GPU app uses a full core of your CPU so be sure to left one core free for every GPU you are using in this project.

 
 
35m33s and 1536 credits for a GPU wu with my 560ti 2GB and 372.90 driver
19m08s and 6144 credits for a GPU wu with my 980ti and 372.90 driver
15m26s and 6144 credits for a GPU wu with my r9 280x Toxic
45 min and 72 credits for a CPU wu with my Intel i5-6600k@4.1
2h49m and 72 credits for an ARM wu with my Nvidia Shield k1

I'm game.... but..but.. NO BADGES!??         I know a few of us have posted on their message boards asking for them.   Falls on deaf ears it seems.. There was a few concepts for them but nothing became of it. 

It's great to have a choice on gpu or cpu because of how fast these other challenges come up.
There will be a slight overlap with the next Formula Boinc Sprint. (04/27/2017 21:00 (UTC) - 04/30/2017 20:59 (UTC)) and with luck they will name Moo!.

My GPUs will be ready to go!

After a couple of days of work the average results are that in this project my old Sapphire R9280x Toxic is faster than my MSI GTX980ti :
19m08s and 6144 credits for a GPU wu with my 980ti and 372.90 driver, Intel i5-6600k with Win7
15m26s and 6144 credits for a GPU wu with my r9 280x Toxic, AMD PII X3 720 with Win10

bcavnaugh
Great Choice, Bummer the timing is off for me.
My GTX 980 is down because I am changing out Computer Cases. I run 2 to 1 task on this rig and Moo! Runs Great.
 
 (4 GPUs)
<app_config>
  <app>
    <max_concurrent>2</max_concurrent>
    <name>dnetc</name>
    <gpu_versions>
        <gpu_usage>2.0</gpu_usage>
        <cpu_usage>0.5</cpu_usage>
    </gpu_versions>
  </app>
</app_config>
 
This to may be one of my last Hard Run GPU before Summer time hits us.
I will do my best to get it moved by Sunday Night.
 
 (2 GPUs)
 <app_config>
  <app>
    <max_concurrent>1</max_concurrent>
    <name>dnetc</name>
    <gpu_versions>
        <gpu_usage>2.0</gpu_usage>
        <cpu_usage>0.5</cpu_usage>
    </gpu_versions>
  </app>
</app_config>
 
  (1 GPU)
<app_config>
  <app>
    <max_concurrent>1</max_concurrent>
    <name>dnetc</name>
    <gpu_versions>
        <gpu_usage>1.0</gpu_usage>
        <cpu_usage>0.5</cpu_usage>
    </gpu_versions>
  </app>
</app_config>
 
It should give us a push over on Formula-Boinc as well.

Hope you get the computer up and running soon.
 
Thanks for the config file data,, just did my first GPU run, OMG 10mins completed with 3x 970's, very nice...

planetclown
Here's my app_config.xml for a single GTX 1070 running 3 GPU tasks fed by 1 CPU thread.

<app_config>
<app>
<name>dnetc</name>
<max_concurrent>3</max_concurrent>
<gpu_versions>
<gpu_usage>0.33</gpu_usage>
<cpu_usage>0.34</cpu_usage>
</gpu_versions>
</app>
</app_config>

How long per task is your 1070 taking??

bill1024
I am thinking since your R9-280x is faster than your GTX980Ti Moo must use DP computing.
So an original Titan 6gb should do well also with DP turned on.

I will test this on my Titans, that is DP Enabled.
 
Added
Tested no real difference.

planetclown

bcavnaugh

HK-Steve
How long per task is your 1070 taking??

planetclown (Linux running on an AMD x86_64 or Intel EM64T CPU1.04 (cuda70))
2,933.94 / 91.14 / 6,144.00 Distributed.net Client v1.04 (cuda70)  GTX 1070

Yes, bcavnaugh's right, just around 40min/task.  I'm now testing whether 2 tasks gives better throughput.

2 Tasks on 1 GPU?

HK-Steve
I am currently running 1 task on each GPU combined (2 or 3 GPU together multitasking) and single task per CPU core....,
 
2x 980Ti's with an i5 2500 cpu = 11mins GPU, 1hr 10mins CPU task
3x 970's with a AMD 9590 = 10mins GPU, 45mins CPU task..

Wonder why are your GTX980Ti so much faster than Dans?
Different size tasks, different clock speed. It is a BIG difference.
 
After a couple of days of work the average results are that in this project my old Sapphire R9280x Toxic is faster than my MSI GTX980ti :
19m08s and 6144 credits for a GPU wu with my 980ti and 372.90 driver, Intel i5-6600k with Win7
15m26s and 6144 credits for a GPU wu with my r9 280x Toxic, AMD PII X3 720 with Win10

HK-SteveI do not know Bill.
Am happy that my times are faster, more points
My temps are getting higher here now, were cooler this morning.,
30"c in my workshop currently,,, and rising as summer is coming fast here like you also.. I have so many fans running to push air around all my computers. maybe that is the secret????
 
Really happy my 3x 970's are working so well, great points.
 
 
 
first time moo'ing, so far so good..  did that sound right...... I do live in the land of chocolate and cheese.....

Hi Steve, you live just a little more northern than me, I'm in Verbania, near the Swiss border and its starting to be hot in my house too!
I'm looking at the differences between yours wu's report and mine and yours always said "[Apr 24 02:31:44 UTC] 2 crunchers ('a' and 'b') have been started." and mine "[Apr 24 01:14:19 UTC] 1 cruncher has been started.", you are working on every wu with 2 Gpus ,right ? That would explain why I'm so slow.
 
I'm trying bcavnaugh's config for 1 GPU now (BTW, thanks Bcavnaugh for posting that!) and my 980ti seems to be slighlty faster.

bill1024Testing right now a single 6gb titan with DP on.

Nice, I'm curious to see the results, WUs seems to really likes AMD GPUs in this project.

I do believe it is not DP that makes the AMD run this project well.
Seems to me it is open_CL vs cuda. AMD can do open_cl tasks better than Nvidia cards..
My Titan 6gb is not the best at this project.
AMD r9-280x - 90x or maybe the 480 or better would be real good for these tasks.
 

I am just running GPU at the moment. 

Wow.     That is faster than my 1080 FTW.