Thanks for the advice. I'll do it as soon as I fix the machine. A little bit ago there was suddenly a sharp bang and now it's not running. It's always something!
Regards,
Steve
A sharp bang!? That can't be good. I can't imagine what could cause that other than an RPG :>)
uh-oh, sounds like a capacitor may have popped/blown up. check to see if there are any swollen/burnt capacitors on your motherboard, PSU, or video card.
That seems to have done it. I changed both entries to "0."
This is probably the best number to use unless you have other CPU intensive apps that need all resources instantly for proper operation. BOINC is designed to run in the background and get out of the way if necessary so it's not usually a problem.
Quote:
The notation still remains after each GPU WU about 0.5 cpu etc. but I can't figure out where it's coming from.
Did you read the application configuration link that Richard gave you? The values you see are the defaults as supplied by the project for cpu_usage and gpu_usage. You shouldn't be concerned by these as the defaults seem to work quite well but if you want to refine them you can set up an app_config.xml (as explained in the link) which allows you to override the project defaults by tweaking those (and other) values. You need to read and understand all the instructions and be aware of the potential consequences of ill-chosen values.
In an earlier message you said:
Quote:
One follow-up question. I notice that E@H is displaying the notation (0.5cpu and 0.25ati Etc.) after the gpu units. I know how to adjust the GPU Wu's but can't find where to change the CPU fraction to increase it since E@H no longer uses app_info.xml files.
By "know how to adjust the GPU Wu's", I presume you mean "know how to adjust the GPU utilization factor to allow crunching of multiple concurrent tasks per GPU". This is a convenience preference setting, designed and implemented first here, which allows easy configuration of multiple tasks per GPU. It allows you to quickly choose a standard setting without having to create an app_config.xml. It effectively gives you certain default values for the resources needed by each GPU task that is running concurrently. You mentioned 0.5 CPUs and 0.25 ATI GPUs so that tells me you are attempting to run 4 concurrent tasks per GPU. As you have 2 GPUs (7850 and 7950) you should see 8 concurrent GPU tasks and (at 0.5 CPUs per task) this should automatically reserve 4 full cores from your Phenom II x6 for GPU support duties, leaving just 2 cores running CPU tasks. Is this how your system was running?
Those resource requirements should be reasonable but you should check to make sure. My experience with Phenom II x4s tells me that they run OK at 0.5 CPUs per GPU task. This is when running 4x on a single HD 7850. In your scenario, the HD 7950 might need more support so that you may find better GPU performance if you free up more CPU cores. To do that (ie override the project supplied defaults) you would need to set up app_config.xml. If you set cpu_usage to 0.65, that would reserve 5 full cores (8x0.65=5.2) and 0.75 would reserve all 6 cores. People often say they get best GPU performance by not running CPU tasks at all and this is probably the case with 2 GPUs, one being reasonably high end.
Be aware that a Phenom II x6 is quite power hungry, as is a HD 7950, with the HD 7850 being a lot less so. Running 8 GPU tasks and 2 CPU tasks would have put quite a strain on your PSU so I presume that may have been the cause of the 'bang' you experienced, particularly if the PSU was 'original', ie getting on in age. Also, you need to be concerned by temperatures inside the case. Have you had a look inside to see if it needs cleaning?
Another point to consider. In a recent message robl made a suggestion about using less than 100% of the CPUs. Please be aware if you do use less than 100%, this will be on top of whatever CPUs are excluded by either the GPU utilization factor or the use of app_config.xml. It will also affect all projects to which you are subscribed and not just Einstein. It's therefore 'better' to control Einstein CPU usage through the cpu_usage parameter in an app_config.xml which only comes into play when Einstein GPU tasks are running, thus allowing other projects to have different settings for their own tasks, if necessary.
Another point to consider. In a recent message robl made a suggestion about using less than 100% of the CPUs. Please be aware if you do use less than 100%, this will be on top of whatever CPUs are excluded by either the GPU utilization factor or the use of app_config.xml. It will also affect all projects to which you are subscribed and not just Einstein. It's therefore 'better' to control Einstein CPU usage through the cpu_usage parameter in an app_config.xml which only comes into play when Einstein GPU tasks are running, thus allowing other projects to have different settings for their own tasks, if necessary.
Thanks for pointing this out. I was not aware of this.
Thanks for the advice. I'll do it as soon as I fix the machine. A little bit ago there was suddenly a sharp bang and now it's not running. It's always something!
Regards,
Steve
A sharp bang!? That can't be good. I can't imagine what could cause that other than an RPG :>)
Given that it was crunching E@H at the time, I was thinking it was a really big gravity wave!!
Ended up being the power supply which has now been replaced and is running so I'll make the change you recommended and hopefully that will be it.
That seems to have done it. I changed both entries to "0."
This is probably the best number to use unless you have other CPU intensive apps that need all resources instantly for proper operation. BOINC is designed to run in the background and get out of the way if necessary so it's not usually a problem.
Gary,
Thanks for the suggestions. Now that this machine is back in action, I'll incorporate them. My BOINC machines do nothing but BOINC.
trying to use my amd r9-280x in this machine but getting
"Einstein@Home 08/11/2015 23:24:08 Jobs for ATI GPU are available, but your preferences are set to not accept them" from scheduler.
How do I accept ati in preferences, I dont see any option other than nvidia or intel?
Account page -> Project preferences -> 3rd setting from the top.
If you use different 'venues', change the setting for the venue the host belongs to. Use 'Switch view' to see all venues side by side if you have multiple venues.
Further down the page is a GPU utilization factor setting if you want to run concurrent GPU tasks. Your card can produce more results per day if you run more than one at a time.
RE: Robi, Thanks for the
)
A sharp bang!? That can't be good. I can't imagine what could cause that other than an RPG :>)
uh-oh, sounds like a
)
uh-oh, sounds like a capacitor may have popped/blown up. check to see if there are any swollen/burnt capacitors on your motherboard, PSU, or video card.
RE: That seems to have done
)
This is probably the best number to use unless you have other CPU intensive apps that need all resources instantly for proper operation. BOINC is designed to run in the background and get out of the way if necessary so it's not usually a problem.
Did you read the application configuration link that Richard gave you? The values you see are the defaults as supplied by the project for cpu_usage and gpu_usage. You shouldn't be concerned by these as the defaults seem to work quite well but if you want to refine them you can set up an app_config.xml (as explained in the link) which allows you to override the project defaults by tweaking those (and other) values. You need to read and understand all the instructions and be aware of the potential consequences of ill-chosen values.
In an earlier message you said:
By "know how to adjust the GPU Wu's", I presume you mean "know how to adjust the GPU utilization factor to allow crunching of multiple concurrent tasks per GPU". This is a convenience preference setting, designed and implemented first here, which allows easy configuration of multiple tasks per GPU. It allows you to quickly choose a standard setting without having to create an app_config.xml. It effectively gives you certain default values for the resources needed by each GPU task that is running concurrently. You mentioned 0.5 CPUs and 0.25 ATI GPUs so that tells me you are attempting to run 4 concurrent tasks per GPU. As you have 2 GPUs (7850 and 7950) you should see 8 concurrent GPU tasks and (at 0.5 CPUs per task) this should automatically reserve 4 full cores from your Phenom II x6 for GPU support duties, leaving just 2 cores running CPU tasks. Is this how your system was running?
Those resource requirements should be reasonable but you should check to make sure. My experience with Phenom II x4s tells me that they run OK at 0.5 CPUs per GPU task. This is when running 4x on a single HD 7850. In your scenario, the HD 7950 might need more support so that you may find better GPU performance if you free up more CPU cores. To do that (ie override the project supplied defaults) you would need to set up app_config.xml. If you set cpu_usage to 0.65, that would reserve 5 full cores (8x0.65=5.2) and 0.75 would reserve all 6 cores. People often say they get best GPU performance by not running CPU tasks at all and this is probably the case with 2 GPUs, one being reasonably high end.
Be aware that a Phenom II x6 is quite power hungry, as is a HD 7950, with the HD 7850 being a lot less so. Running 8 GPU tasks and 2 CPU tasks would have put quite a strain on your PSU so I presume that may have been the cause of the 'bang' you experienced, particularly if the PSU was 'original', ie getting on in age. Also, you need to be concerned by temperatures inside the case. Have you had a look inside to see if it needs cleaning?
Another point to consider. In a recent message robl made a suggestion about using less than 100% of the CPUs. Please be aware if you do use less than 100%, this will be on top of whatever CPUs are excluded by either the GPU utilization factor or the use of app_config.xml. It will also affect all projects to which you are subscribed and not just Einstein. It's therefore 'better' to control Einstein CPU usage through the cpu_usage parameter in an app_config.xml which only comes into play when Einstein GPU tasks are running, thus allowing other projects to have different settings for their own tasks, if necessary.
Cheers,
Gary.
RE: Another point to
)
Thanks for pointing this out. I was not aware of this.
RE: RE: Robi, Thanks for
)
Given that it was crunching E@H at the time, I was thinking it was a really big gravity wave!!
Ended up being the power supply which has now been replaced and is running so I'll make the change you recommended and hopefully that will be it.
Regards,
Steve
RE: RE: That seems to
)
Hi Guys, Trying to install
)
Hi Guys,
Trying to install Boinc on a Linux Mint computer.
Messages: Can't install due to missing Character set.
AMD Athlon 4400+
AMD Radeon X1950GT
Any suggestions?
George
Hi everyone, trying to use
)
Hi everyone,
trying to use my amd r9-280x in this machine but getting
"Einstein@Home 08/11/2015 23:24:08 Jobs for ATI GPU are available, but your preferences are set to not accept them" from scheduler.
How do I accept ati in preferences, I dont see any option other than nvidia or intel?
Account page -> Project
)
Account page -> Project preferences -> 3rd setting from the top.
If you use different 'venues', change the setting for the venue the host belongs to. Use 'Switch view' to see all venues side by side if you have multiple venues.
Further down the page is a GPU utilization factor setting if you want to run concurrent GPU tasks. Your card can produce more results per day if you run more than one at a time.
Cheers,
Gary.
Thanks ! not obvious :) got 7
)
Thanks ! not obvious :) got 7 wus right away :)