Posts by Bill

1) Message boards : News : Special Fundraiser for Parkes Data Store and GPU development system (Message 1912604)
Posted 7 days ago by Profile Bill
Post:
*Edit* Was just putzing around on Boincstats, and I noticed on the Hosts Overview page of SETI's stats, on 12/10, there was a huge blip in the number of new hosts. It appears the average of the month surrounding that date was in the neighborhood of 2-500 a day, and on that day the total was 2844. Not sure if this fund drive had anything to do with it or not, but I did find it an interesting coincidence.


This reminds me...and I hope to not go too far down a different rabbit hole, but I thought I saw a post or received a notice on BOINC a few months back stating that more WUs were going to be in the hopper soon. The point of the notice was asking us to spare a few more CPU cycles. Did I just dream that, and if not, has there been a noticeable increase in WUs being crunched?
2) Message boards : Science (non-SETI) : A new prime number has been found! (Message 1910848)
Posted 14 days ago by Profile Bill
Post:
I know that this is mathematics and not science, but congrats to the people involved in finding the largest prime number! Fivethirtyeight ran an article, and subtly gave S@H some press as well:

https://fivethirtyeight.com/features/we-have-a-new-prime-number-and-its-23-million-digits-long/
3) Message boards : Number crunching : SETI/BOINC Milestones [ v2.0 ] - XXIX (Message 1909689)
Posted 19 days ago by Profile Bill
Post:
I noticed over the weekend that I am in the top 25% of average credit! After not using S@H for about 12 years, I'm a little surprised to have reached this milestone after crunching on and off for the past six months with one computer. As of right now I have a RAC of 1,953.

Congrats to everyone else who has reached a milestone!
4) Message boards : News : Special Fundraiser for Parkes Data Store and GPU development system (Message 1906227)
Posted 10 Dec 2017 by Profile Bill
Post:
A 6:1 match is amazing, count me in! Its great to see the enthusiasm, it surely has drawn me in!
5) Message boards : Number crunching : "infinite" runtime on Intel GPU tasks (Message 1903736)
Posted 30 Nov 2017 by Profile Bill
Post:
Ok, I think I get it now. I'll stop running iGPU and see if anything changes. Everything else has been running fine before...I think? We'll see if anything improves. One last thing and I'll let this go: when my iGPU work units are running, it says it is using a fraction of a CPU + the iGPU. The last WU I ran, for example, is using 0.00314 CPU + 1 Intel GPU. I assumed that for that work unit that is all the unit needed to crunch. My four CPUs are all running tasks, one with the NVIDIA, two dedicated to their WUs, and I am assuming the last one just decides to run one non-GPU WU, but keeping a fraction of its power on the side for the iGPU.

Thanks again for the help!
6) Message boards : Number crunching : "infinite" runtime on Intel GPU tasks (Message 1903694)
Posted 29 Nov 2017 by Profile Bill
Post:
Thank you for the input. I apologize in advance if I'm asking any dumb questions. Its been 13 years since my CS minor...I know my computer smarts are somewhere in my brain, they just haven't been exercised to this degree for some time.

Are you saying that most people don't run Intel GPUs for S@H? I just jumped back in to S@H a few months ago, and I have not had any problems with iGPU workunits up until this point. I suppose I can turn off the processor in my settings...it just seems like I'm having him sit on the bench for no reason.

Also, if this issue has come up before, and you happen to know of a thread off the top of your head, could you point me in the right direction? I did a search of the forms and came up empty.

As for the restarting, I'm not sure what that is all about. I use the computer during the day, but lock it out at night, and my power settings do not go into sleep/hibernation mode. Yes, the monitor will turn off, but it should be staying off until the next morning. I also just tried updating the driver (version 8.15.10.2639) and had something weird happen. I grabbed the latest driver off of Intel's website, and it said it couldn't be installed, that I needed one specific to my computer. I went to Lenovo to get the driver, and it said that my computer doesn't meet the requirements...even though I used their auto-detect for my hardware. Not sure what to do there, but I'm not going to turn this thread into a "help me install drivers" thread.

For what its worth, I noticed that if I open/close BOINC, any work done on an iGPU unit resets. The elapsed time is the same from when the unit hadn't started yet, and the elapsed time begins at zero...its as if the work progress isn't saved somehow? With all that said, I suspect something is wrong with this "batch" of Intel work units, more than something being wrong with my setup...not that I have anything to prove that.

Thanks again,

Bill
7) Message boards : Number crunching : "infinite" runtime on Intel GPU tasks (Message 1903337)
Posted 28 Nov 2017 by Profile Bill
Post:
I have had a few tasks that seem to run forever without completing. Right now I have one that when it started running yesterday estimated 4 hrs 45 minutes to complete, but this morning after running for 14 hours it says it has 1,194,684 days left to run!

I had this happen to one or two other tasks earlier, and they all happened with an Intel GPU tasks. Any idea what's wrong? This has only started happening in the past 7 days.

The task I am having problems with, I think, is this one: http://setiathome.berkeley.edu/result.php?resultid=6152615131

(there are a few I have that identical timestamps, this one's name is the closest, but the last few digits don't match on BOINC).

I have not had problems for the past few months since I've been back on S@H, so I'm not sure if this is my problem or a data problem. Thoughts?

Bill





 
©2018 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.