If anyone already made an app_info.xml with S6 WU's included, than it will be highly appreciated if one post it here.
Both you and Dirk need to send a PM to Claggy and/or rroonnaalldd, they are both experts with app_info.xml files.
Dirk how much memory does a use gpu workunit use right now? You will need twice that plus a little bit to fit more than one in memory, you need a little bit for the overhead.
If anyone already made an app_info.xml with S6 WU's included, than it will be highly appreciated if one post it here.
Both you and Dirk need to send a PM to Claggy and/or rroonnaalldd, they are both experts with app_info.xml files.
Dirk how much memory does a use gpu workunit use right now? You will need twice that plus a little bit to fit more than one in memory, you need a little bit for the overhead.
100MB is used by windows it seems, 1 cuda WU from einstein adds around 300MB. So if I'd run 4 on my GTX480 it'd use roughly 1300 out of 1500MB. Edit, it'll probably be around the 1350MB mark just like for the op of this thread.
I'll pm one of them soon, thanks for the heads up!
Below, there is my app_info.xml for linux, running quite well.
I am not sure if everything is correct, (too lazy to go through that carefully) but I have not experienced any trouble, yet.
On these lines:
0.25
0.25
should be value 0.2 - I am still playing with that number a bit
I do not use binary radio pulsar search app. for CPU - only for GPU.
1WU = approx. 320MB of graphics RAM
2WU = approx. 570MB of graphics RAM
3WU = approx. 860MB of graphics RAM (my system is not usable with 3WU - I have GTS250 with 1GB)
einstein_S5GC1HF
Global Correlations S5 HF search #1
Here's a complete app_info for windows with all the latest apps in it, just change the count value to 0.5, 0.33, or 0.25 depending if you want to run 2, 3 or 4 Cuda Wu's at once:
einstein_S5GC1HF
Global Correlations S5 HF search #1
Just a little update, running 4 WU's on the GPU now. RAM usage is fluctuating between 1200-1300 MB. But the core downclock bug kicked in with all the restarting of tasks so I'll restart now to see how stable it runs overnight.
Restarted now. Running 6 CPU tasks along with the 4 GPU WUs. Overall CPU usage is at ~93% so the GPU isnt being starved of CPU cycles. GPU usage fluctuates around 73% (it was almost 100% when the core was downclocked), GPU memory use is between 1200-1300 MB. Here's a screenshot of MSI afterburners monitoring graphs.
Hmm, I've run into a little problem. I can't get enough WUs in my queue. This is because apparently boinc thinks one cuda WU will now take me 34 hours to complete, despite evidence to the contrary. I remember reading somewhere how to change something that will help reduce that expected time but I just can't remember where.
I'd rather not increase my work buffer by too much so if anyone knows it's appreciated.
Edit:
Found it! Got the flops estimate from my client_state and added it to the app_info and changed the number by quite a bit.
16276995705.519375
Now the estimated runtime is 2 hours 40 mins which is good enough. I did manage to ruin my current batch of tasks though by fiddling around with the app_info, sorry for that! Also crashed my display driver when I exited boinc but I'm running the latest nvidia beta ones so it's not unexpected and it recovered just fine.
Last edit, all my latest tasks are called anonymous platform on my results page now. Is that normal? I don't think they were last night.
RE: I have an I7 CPU. It
)
I think this is just a case of the estimate being for a nonHT cpu.
If anyone already made an
)
If anyone already made an app_info.xml with S6 WU's included, than it will be highly appreciated if one post it here.
RE: If anyone already made
)
Both you and Dirk need to send a PM to Claggy and/or rroonnaalldd, they are both experts with app_info.xml files.
Dirk how much memory does a use gpu workunit use right now? You will need twice that plus a little bit to fit more than one in memory, you need a little bit for the overhead.
RE: RE: If anyone already
)
100MB is used by windows it seems, 1 cuda WU from einstein adds around 300MB. So if I'd run 4 on my GTX480 it'd use roughly 1300 out of 1500MB. Edit, it'll probably be around the 1350MB mark just like for the op of this thread.
I'll pm one of them soon, thanks for the heads up!
Below, there is my
)
Below, there is my app_info.xml for linux, running quite well.
I am not sure if everything is correct, (too lazy to go through that carefully) but I have not experienced any trouble, yet.
On these lines:
0.25
0.25
should be value 0.2 - I am still playing with that number a bit
I do not use binary radio pulsar search app. for CPU - only for GPU.
1WU = approx. 320MB of graphics RAM
2WU = approx. 570MB of graphics RAM
3WU = approx. 860MB of graphics RAM (my system is not usable with 3WU - I have GTS250 with 1GB)
einstein_S5GC1HF
Global Correlations S5 HF search #1
einstein_S5GC1HF_1.07_i686-pc-linux-gnu__SSE2
einstein_S5R6_1.01_graphics_i686-pc-linux-gnu
einstein_S5GC1HF
107
i686-pc-linux-gnu
1.000000
1.000000
SSE2
einstein_S5GC1HF_1.07_i686-pc-linux-gnu__SSE2
einstein_S5R6_1.01_graphics_i686-pc-linux-gnu
graphics_app
einstein_S6Bucket
Gravitational Wave S6 GC search 1.01
einstein_S6Bucket_1.01_i686-pc-linux-gnu__SSE2
einstein_S5R6_1.01_graphics_i686-pc-linux-gnu
einstein_S6Bucket
101
i686-pc-linux-gnu
1.000000
1.000000
SSE2
6.13.0
einstein_S6Bucket_1.01_i686-pc-linux-gnu__SSE2
einstein_S5R6_1.01_graphics_i686-pc-linux-gnu
graphics_app
einsteinbinary_BRP3
Binary Radio Pulsar Search
einsteinbinary_BRP3_1.08_i686-pc-linux-gnu__BRP3cuda32nv270
libcudart32_32_16.so
libcufft32_32_16.so
EULA.txt
db.dev.lin.3d35195e
dbhs.dev.lin.3d35195e
einsteinbinary_BRP3
108
i686-pc-linux-gnu
0.25
0.25
BRP3cuda32nv270
6.13.0
einsteinbinary_BRP3_1.08_i686-pc-linux-gnu__BRP3cuda32nv270
libcudart32_32_16.so
libcudart.so.3
libcufft32_32_16.so
libcufft.so.3
EULA.txt
EULA.txt
db.dev.lin.3d35195e
db.dev
dbhs.dev.lin.3d35195e
dbhs.dev
CUDA
0.499995
Here's a complete app_info
)
Here's a complete app_info for windows with all the latest apps in it, just change the count value to 0.5, 0.33, or 0.25 depending if you want to run 2, 3 or 4 Cuda Wu's at once:
einstein_S5GC1HF
Global Correlations S5 HF search #1
einstein_S5GC1HF_3.06_windows_intelx86__S5GCESSE2.exe
einstein_S5R6_3.01_graphics_windows_intelx86.exe
einstein_S5GC1HF
306
windows_intelx86
1.000000
1.000000
S5GCESSE2
6.13.0
einstein_S5GC1HF_3.06_windows_intelx86__S5GCESSE2.exe
einstein_S5R6_3.01_graphics_windows_intelx86.exe
graphics_app
einsteinbinary_BRP3
Binary Radio Pulsar Search
einsteinbinary_BRP3_1.05_windows_intelx86__BRP3SSE.exe
einsteinbinary_BRP3_1.07_windows_intelx86__BRP3cuda32.exe
einsteinbinary_BRP3_1.00_graphics_windows_intelx86.exe
cudart_xp32_32_16.dll
cufft_xp32_32_16.dll
db.dev.win.3d35195e
dbhs.dev.win.3d35195e
einsteinbinary_BRP3
105
windows_intelx86
1.000000
1.000000
BRP3SSE
6.13.0
einsteinbinary_BRP3_1.05_windows_intelx86__BRP3SSE.exe
einsteinbinary_BRP3_1.00_graphics_windows_intelx86.exe
graphics_app
einsteinbinary_BRP3
107
windows_intelx86
0.200000
1.000000
BRP3cuda32
6.13.0
einsteinbinary_BRP3_1.07_windows_intelx86__BRP3cuda32.exe
cudart_xp32_32_16.dll
cudart32_32_16.dll
cufft_xp32_32_16.dll
cufft32_32_16.dll
einsteinbinary_BRP3_1.00_graphics_windows_intelx86.exe
graphics_app
db.dev.win.3d35195e
db.dev
dbhs.dev.win.3d35195e
dbhs.dev
CUDA
1.000000
314572800.000000
einstein_S6Bucket
Gravitational Wave S6 GC search
einstein_S6Bucket_1.01_windows_intelx86__SSE2.exe
einstein_S5R6_3.01_graphics_windows_intelx86.exe
einstein_S6Bucket
101
windows_intelx86
1.000000
1.000000
SSE2
6.13.0
einstein_S6Bucket_1.01_windows_intelx86__SSE2.exe
einstein_S5R6_3.01_graphics_windows_intelx86.exe
graphics_app
If you're missing apps and dll's, here are some links:
http://einstein.aei.mpg.de/download/einstein_S5GC1HF_3.06_windows_intelx86__S5GCESSE2.exe
http://einstein.aei.mpg.de/download/einstein_S5R6_3.01_graphics_windows_intelx86.exe
http://einstein.aei.mpg.de/download/einsteinbinary_BRP3_1.05_windows_intelx86__BRP3SSE.exe
http://einstein.aei.mpg.de/download/einsteinbinary_BRP3_1.07_windows_intelx86__BRP3cuda32.exe
http://einstein.aei.mpg.de/download/einsteinbinary_BRP3_1.00_graphics_windows_intelx86.exe
http://einstein.aei.mpg.de/download/cudart_xp32_32_16.dll
http://einstein.aei.mpg.de/download/cufft_xp32_32_16.dll
http://einstein.aei.mpg.de/download/db.dev.win.3d35195e
http://einstein.aei.mpg.de/download/dbhs.dev.win.3d35195e
http://einstein.aei.mpg.de/download/einstein_S6Bucket_1.01_windows_intelx86__SSE2.exe
Claggy
Thanks so much Claggy! It
)
Thanks so much Claggy! It works :)
Running 2 WUs now. Also playing football manager so not gonna push it up to 3 or 4 for now.
Just a little update, running
)
Just a little update, running 4 WU's on the GPU now. RAM usage is fluctuating between 1200-1300 MB. But the core downclock bug kicked in with all the restarting of tasks so I'll restart now to see how stable it runs overnight.
Restarted now. Running 6 CPU tasks along with the 4 GPU WUs. Overall CPU usage is at ~93% so the GPU isnt being starved of CPU cycles. GPU usage fluctuates around 73% (it was almost 100% when the core was downclocked), GPU memory use is between 1200-1300 MB. Here's a screenshot of MSI afterburners monitoring graphs.
Thanks Claggy! 10 minutes,
)
Thanks Claggy! 10 minutes, every thing is allright.
Hmm, I've run into a little
)
Hmm, I've run into a little problem. I can't get enough WUs in my queue. This is because apparently boinc thinks one cuda WU will now take me 34 hours to complete, despite evidence to the contrary. I remember reading somewhere how to change something that will help reduce that expected time but I just can't remember where.
I'd rather not increase my work buffer by too much so if anyone knows it's appreciated.
Edit:
Found it! Got the flops estimate from my client_state and added it to the app_info and changed the number by quite a bit.
16276995705.519375Now the estimated runtime is 2 hours 40 mins which is good enough. I did manage to ruin my current batch of tasks though by fiddling around with the app_info, sorry for that! Also crashed my display driver when I exited boinc but I'm running the latest nvidia beta ones so it's not unexpected and it recovered just fine.
Last edit, all my latest tasks are called anonymous platform on my results page now. Is that normal? I don't think they were last night.