Posts by HAL9000


log in
1) Message boards : Number crunching : adding to rig with r9 280x (Message 1594446)
Posted 34 minutes ago by Profile HAL9000
Since BOINC won't by default use both GPUs when different, there is an option in cc_config.xml to force it to use all gpus:

<use_all_gpus>1</use_all_gpus>

make sure you include that when you use both cards.

And make sure both cards are included in the BOINC startup messages.



Thanks jravin for the reminder.

From what has been said I probably could go for the r9 270x instead, for another $10.

Thanks Hal

11-13% more performance for only $10 is a pretty good deal. Sure you don't want to splurge and just get a 2nd R9 280X? :P
2) Message boards : Number crunching : adding to rig with r9 280x (Message 1594417)
Posted 1 hour ago by Profile HAL9000
Would there be a problem adding a r9 270 drawing 150 watts to my i7 4790K with a r9 280x? I just checked my watt meter and I'm only drawing 310 watts with boinc running. I have an 850 watt gold PSU.

I have been told there are some various technical problems with combining different amd models on the same machine.

Thanks

The issues when mixing different speed GPUs are normally when they are drastically different in speed. The R9 270 is about 60% the speed of the R2 280X.

Also at present whatever app settings you make to tune one GPU will apply to the other. Which may not be ideal for one GPU & can cause it to generate errors. You may have to set the tune settings for a middle of the ground settings. So that both GPUs will be happy.
I think Raistmer said a new app version will allow per card settings, but it is not current released so no need to worry about using it yet.

Powerwise you would probably be OK. With 400w of GPUs that is 33.3 Amps. With the other 12v items in your machines that would probably put the load up near 42A or so. What is the total 12v output on your PSU?
3) Message boards : Number crunching : AstroPulse Work Fetch Thread (Message 1594399)
Posted 2 hours ago by Profile HAL9000
Yesterday after AP work started going out I checked my two i5-4670K systems with BOINC 7.2.42 to see how they were getting work. Most requests would only send 3 or 4 tasks at a time. Sometimes they would be all CPU, sometimes they would be all ATI, but most of the time it would be a mix of CPU & ATI. I didn't have any extra debug flags set as this was more of a "let's see what BOINC does with how I'm setup now" test.
When I go home for lunch I'll setup a new BOINC instance with debug flags set to see how it operates from the perspective of a new host.
4) Message boards : Number crunching : Panic Mode On (91) Server Problems? (Message 1594159)
Posted 15 hours ago by Profile HAL9000
When will I get APs for NVIDIA? I really want some.

Thursday the 18th at 17:22 UTC.

Looks like you did have 36 on one of your machines but they all got marked as timed out just a short bit ago.
5) Message boards : Number crunching : AstroPulse Work Fetch Thread (Message 1594006)
Posted 20 hours ago by Profile HAL9000
It's odd. Now that we are looking at it. My host 5837483 received 4 GPU tasks in a work fetch request before the CPU limit was hit.
6) Message boards : Number crunching : resetting memory clock for amd r9 280x (Message 1593845)
Posted 1 day ago by Profile HAL9000
I has a problem with the memory adjustment not working on my HD 5750. I set it to minimum, applied it, waited for it to show in GPUz, & then moved it back up to the value I wanted.
Probably a bug in the GUI. I would guess related to the occasional crashing of CCC.
7) Message boards : Number crunching : AP V7 (Message 1593842)
Posted 1 day ago by Profile HAL9000

Did anyone ever submit a reasonable reason for not just having Plain APs as you suggested a while back? I can't think of any reason to have device restrictions.

What do you mean??? There is no difference between CPU and GPU AP tasks at all. Not even some scheduler rules as for VLAR MB...

The VLARs are the only problem, since those are kept from the GPUs. What I mean is why do I have AstroPulse v7 Anonymous platform (ATI GPU) And AstroPulse v7 Anonymous platform (CPU). Why not just AstroPulse v7 Anonymous platform. Then the Manager wouldn't have to limit the tasks to certain devices. That way you wouldn't get stuck with 79 CPU tasks that would take over 10 days to complete on the CPUs or only 11 hours to complete on the GPUs. The tasks could be sent to the GPUs or CPUs depending on which device needed them. I'm pretty sure that's what Claggy was suggesting back in one of those old CreditFew threads.

Hmmm, seems I have more CPU resends.

Ah you just want a general pool of tasks on the client end and then let the client process them on the hardware as it sees fit. Other than the issues that the BOINC devs would have to overcome for that to happen, of which I imagine there would be a few, it seems like a logical way to do things. I think several people have suggested BOINC work like this since GPU crunched started. How other projects would handle such a change is really the issue.
8) Message boards : Number crunching : Lunatics Windows Installer v0.43 Release Notes (Message 1593618)
Posted 1 day ago by Profile HAL9000
Wedge & Raistmer,

Thanks for the additional info! FYI: I haven't done any experimenting yet - still in getting "up to speed" mode.

I did a quick read through of the links posted here regarding TdrDelay, etc. and found some discrepancies between MS's info and my version of Windows - which I don't understand yet. That is, when I use RegEdit to look at TdrDelay, I noticed that my Windows doesn't have most of the other Tdr related variables listed in the articles. Not sure if this is because the MS articles were mainly referring to Win8 and I have Win7 or if there may be another reason - e.g. I used a Norton Utilities type program a while back to "fix" some registry problems. I also noticed that my TdrDelay was set to 8 - not the default value of 2 - but I have never tinkered with these settings before.

Haven't gotten to the "Read Me" file yet but I appreciate the guidance relative to N value settings. I am sure it will save me a lot of time experimenting.

I have a lot on my plate for the next few days, so I probably won't get back to this until next week sometime. But I will keep the current WU suspended until I do. And I will report back on any successes and/or failures.

Again, thanks to everyone who responded for the help.

Stick

Any missing values are running at predefined default levels. Unless an entry is specifically created to change the value. Only items you want to change would need to be created. Instead of adding them all.
It is possible that your Norton software, or another application, added some values or changed some other settings from default.
9) Message boards : Number crunching : Lunatics Windows Installer v0.43 Release Notes (Message 1593612)
Posted 1 day ago by Profile HAL9000
I also thought thousands was a bit excessive but I think someone on the boards suggested it at that time. I'll try what you suggested should APs run out again.

I tend to stick with latest stable drivers on the APUs as well. I didn't get to the point of BSoDs, just driver restarts.

I think the primary cause of video driver restarts in Windows is over aggressive crash detection. It is a bit of the "There was nothing wrong but we fixed it for you" I think.
I imagine if the video driver was actually crashing as often as Windows thinks it was I would have lockups or BSOD happening now instead. Which I don't.
10) Message boards : Number crunching : display driver stopped and now wu is stalled (Message 1593497)
Posted 2 days ago by Profile HAL9000
Test copy to SIV64 -- using 2 wu's with the gpu:

Well I tried to click the X button to close that. I should go back to bed... if only I wasn't at work. Defiantly going to one of those kind of days.
11) Message boards : Number crunching : AP V7 (Message 1593263)
Posted 2 days ago by Profile HAL9000
...I would have set it up to request AP work, but knowing the AP coffers are dry that would have not been very fruitful.

I can repeat this once we have AP to spare again. However, I do suspect that it will take more than 3 request for work to fill the host queue to the limit with AP. Looking in my logs the most AP I have had in one request looks to be 7 in the past several weeks.

Yes, MBs are not APs. I was going to ask, then decided to check myself. I just checked two of your Hosts. One only had a couple resends, the other looked just like my Host, http://setiathome.berkeley.edu/results.php?hostid=5255585
All the resends since the well ran dry went to the CPU. I'm going to check a few others. Well, checking other hosts is too much work. Apparently quite a few had full CPU caches using MBs. My CPU cache has room for resends, the GPU cache has much more room though. Anyone else have room in both caches and receiving just CPU AP resends?

Only 3 of my hosts are configured for CPU & GPU AP only. Hosts 5837483, 5255585, & 7324426. I was going to do some benchmarking on host 7395996, but then we switched to AP v7. On my notebook 5838375. I can run CPU tasks or GPU tasks, but not both due to overheating. Maybe if I stick it outside in the winter...
Everything else runs CPU only for MB or AP. With the exception of 6727898. Which runs CPU & GPU MB with an old NV 8500 GT.

Host 5837483 has has surprisingly bad luck getting any resends. While 5255585 has been filling up. With AP luck of the draw is a wildcard in the calculation as well.
12) Message boards : Number crunching : Lunatics Windows Installer v0.43 Release Notes (Message 1593237)
Posted 2 days ago by Profile HAL9000
Stick & Wedge,
You might just need to tweak some of the Windows OS video watchdog settings.

See more details here: http://setiathome.berkeley.edu/forum_thread.php?id=75324&postid=1553652#1553652

If bumping up the TdrDelay doesn't work you might want to just disable it with the TdrLevel setting.
13) Message boards : Number crunching : iGPU tuning (Message 1593235)
Posted 2 days ago by Profile HAL9000
Are there tools to monitor cache thrashing or memory bus saturation?

You may try PerfMonitor
http://www.cpuid.com/softwares/perfmonitor2.html
http://www.cpuid.com/softwares/perfmonitor.html

Unfortunately PerfMonitor does not even display the cache info for my BayTrail system. Which I'll be using as a basis of comparison to my Haswell system. It also seems to have an issue showing information sometimes. Much like HWMonitor will sometimes be missing some of the readings until it is restarted a few times.
14) Message boards : Number crunching : AP V7 (Message 1593227)
Posted 2 days ago by Profile HAL9000
So I setup a new instance of BOINC using 7.2.42. I attached to SETI@home & here are the results of the work requests
27-Oct-2014 21:49:17 [http://setiathome.berkeley.edu/] Master file download succeeded
27-Oct-2014 21:49:22 [http://setiathome.berkeley.edu/] Sending scheduler request: Project initialization.
27-Oct-2014 21:49:22 [http://setiathome.berkeley.edu/] Requesting new tasks for CPU and intel_gpu
27-Oct-2014 21:49:24 [SETI@home] Scheduler request completed: got 1 new tasks
27-Oct-2014 21:49:24 [SETI@home] General prefs: from SETI@home (last modified 27-Oct-2014 21:46:53)
27-Oct-2014 21:49:24 [SETI@home] Host location: none
27-Oct-2014 21:49:24 [SETI@home] General prefs: using your defaults
Received 1 task for CPU.

27-Oct-2014 21:55:46 [SETI@home] update requested by user
27-Oct-2014 21:55:49 [SETI@home] Sending scheduler request: Requested by user.
27-Oct-2014 21:55:49 [SETI@home] Requesting new tasks for CPU and intel_gpu
27-Oct-2014 21:55:52 [SETI@home] Scheduler request completed: got 67 new tasks
Received 67 tasks for iGPU .

27-Oct-2014 22:03:11 [SETI@home] update requested by user
27-Oct-2014 22:03:13 [SETI@home] Sending scheduler request: Requested by user.
27-Oct-2014 22:03:13 [SETI@home] Requesting new tasks for CPU and intel_gpu
27-Oct-2014 22:03:17 [SETI@home] Scheduler request completed: got 132 new tasks
Received 132 tasks for CPU & iGPU filling the host queue to its limit of 200 tasks.

The task list shows matches times for the work requests with my timezone of UTC -4(US EDT) this time of year.

The venue settings are for a cache of 10 & 10 (20 day cache) & MB only for this test. I would have set it up to request AP work, but knowing the AP coffers are dry that would have not been very fruitful.

I can repeat this once we have AP to spare again. However, I do suspect that it will take more than 3 request for work to fill the host queue to the limit with AP. Looking in my logs the most AP I have had in one request looks to be 7 in the past several weeks.
15) Message boards : Number crunching : remaining AP v6s (Message 1593117)
Posted 2 days ago by Profile HAL9000
My old iMAC that can't run AP v7 has not received a new AP v6 task to crunch since the 25th.
The end is neigh! The end is neigh! The end is neigh!

However I do know of at least one task that isn't going to be returned on time and will be sent out again on 2014-11-05 due to a failed machine.
16) Message boards : Number crunching : Panic Mode On (91) Server Problems? (Message 1593113)
Posted 2 days ago by Profile HAL9000
And then I had no more work...

27/10/2014 22:09:56 | SETI@home | Reporting 1 completed tasks
27/10/2014 22:09:56 | SETI@home | Requesting new tasks for AMD/ATI GPU
27/10/2014 22:09:56 | SETI@home | [sched_op] CPU work request: 0.00 seconds; 0.00 devices
27/10/2014 22:09:56 | SETI@home | [sched_op] AMD/ATI GPU work request: 90720.00 seconds; 1.00 devices
27/10/2014 22:09:58 | SETI@home | Scheduler request completed: got 0 new tasks
27/10/2014 22:09:58 | SETI@home | [sched_op] Server version 705
27/10/2014 22:09:58 | SETI@home | No tasks sent
27/10/2014 22:09:58 | SETI@home | No tasks are available for AstroPulse v7
27/10/2014 22:09:58 | SETI@home | Tasks for CPU are available, but your preferences are set to not accept them
27/10/2014 22:09:58 | SETI@home | Tasks for NVIDIA GPU are available, but your preferences are set to not accept them
27/10/2014 22:09:58 | SETI@home | Tasks for Intel GPU are available, but your preferences are set to not accept them

27/10/2014 22:09:58 | SETI@home | Project requested delay of 303 seconds
27/10/2014 22:09:58 | SETI@home | [sched_op] handle_scheduler_reply(): got ack for task ap_30oc11ab_B4_P0_00108_20141026_31709.wu_0
27/10/2014 22:09:58 | SETI@home | [sched_op] Deferring communication for 00:05:03
27/10/2014 22:09:58 | SETI@home | [sched_op] Reason: requested by project

It's the third time in a short amount of time I've seen this specific "No tasks sent, no tasks available for your hardware" message. Hope it's only me. ;-)

If you mean those messages I highlighted. Then they are nothing new. I have a whole log full of "Tasks for AMD/ATI GPU are available, but your preferences are set to not accept them" messages on a host with no ATI GPU.
Also sometimes I get the message about "there is no usable entry for Astropulse v7" in your app_info.xml on MB only machines or the opposite on AP only machines.
17) Message boards : Number crunching : @Pre-FERMI nVidia GPU users: Important warning (Message 1593108)
Posted 2 days ago by Profile HAL9000
Someone mentioned there were other drivers with computational errors like ver 295.73.

Is there an updated white or black list of drivers pinned somewhere? I had the misfortune to have these 340.52 drivers installed on 2 machines, wondering why I was receiving no GPU units, and when trying to correct the issue dropped down to another blacklisted driver.

IMO, the best Driver for the older cards is 266.58 WHQL. That driver doesn't use the Full CPU core to run AstroPulse tasks the way newer drivers do. No need for the -sleep switch or anything else, great for Dual core CPUs. It also works fine with CUDA versions up to 3.2. This host is using it; All tasks for computer 6813106 It also works fine with my GTS 250.

The XP Link is here, http://www.nvidia.com/object/winxp-266.58-whql-driver.html
You'll have to search for the other OS versions.

Or just use http://www.oldapps.com/nvidia.php
18) Message boards : Number crunching : AP V7 (Message 1593060)
Posted 2 days ago by Profile HAL9000
Here is a recent example, "It keeps loading me up with cpu tasks"
I've noticed the same behavior, so have others. I notice it every few days because my Mac is just about out of work every few days. It use to download the GPU tasks first, now it doesn't. Just as others have noted...

So. That's what we're looking for. What makes that decision? What criteria does it use to decide? Please start thinking about some answers, rather than just re-stating the question.

Not that is has anything to do with AP v7. Maybe this stuff could be moved to a new thread about the subject?
I have noticed a change in behavior reporting work on a machine with a full queue limit. Previously with 100 tasks in the queue if a completed tasks was reported a new one would be sent in the same request. Now a task is not sent & the message "This computer has reached a limit on tasks in progress" is displayed. Then on the next request for work a new task is sent.
19) Message boards : Number crunching : Some questions about BOINC for Android... (Message 1593050)
Posted 2 days ago by Profile HAL9000
Don't know if this has already been asked here but can a smartphone also crunch when it's not connected to the main current? In other words, can an android crunch only on its battery?


Yes sure but the battery wont last long probably.

I set my Galaxy S4 to stop at 20% battery & then let it go from a full charge. It did manage a few hours before pausing crunching at 20% battery level.

Ran a test to track more precisely how long my S4 will run on battery while crunching. With 3 tasks running at once I'm down to 60% from a full charge after 01:30:15 of run time. So probably close to 3 hours of run time before hitting 20%. As I was using my phone for some activities while I was at lunch it could be a bit more than 3 hours if left untouched. I'm going to charge the battery back up and try with 2 tasks next.
20) Message boards : Number crunching : Some questions about BOINC for Android... (Message 1592972)
Posted 3 days ago by Profile HAL9000
Don't know if this has already been asked here but can a smartphone also crunch when it's not connected to the main current? In other words, can an android crunch only on its battery?


Yes sure but the battery wont last long probably.

I set my Galaxy S4 to stop at 20% battery & then let it go from a full charge. It did manage a few hours before pausing crunching at 20% battery level.


Next 20

Copyright © 2014 University of California