[bfromcolo]

I was already running WCG on my old server, will add more to it. This is a tough project for us to get points in.

 

[bfromcolo]

I hope they give SETI one last go, even if we would never have a chance at points there.


Looks like Universe
03/19/2020 21:00 (UTC) - 03/22/2020 20:59 (UTC)
[email protected]

 

[bfromcolo]

330 hours of CPU time for 0 points, 85 error out and 11 got rejected when they were uploaded, another 100 complete that aren't uploading. This is on 4 different computers, Linux and Windows. I give up.

 

[bfromcolo]

Finally got some points, the BHspin seem to be running OK and they generated a bunch of them.

 

[bfromcolo]

Yes plenty of work, and no failed tasks since switching to BHspin only. But I have a bunch of uploads with problems.

 

[bfromcolo]

I agree with the Ryzen recommendation, I have 2 1700x and a 1600 I bought at Microcenter as motherboard bundles. They were cheap ($139 and $79 for the CPU) and work fine on inexpensive B450 motherboards since you don't need a lot of features for BOINC. And the included cooler with the 1600 is adequate at stock clocks for 12 threads of you don't mind some fan noise.

If I was building a cheap BOINC/[email protected] machine today I would probably be looking at the 1600AF, 12 threads with a serviceable cooler for $85.

And yes the RAC after this Sprint is meaningless, the ULX tasks were a mess (I had 386 fail and another 54 invalid) and then the download/upload issues.

 

[bfromcolo]

This weekends scheduled Sprint is cancelled.

 

[bfromcolo]

If you set it to "no new tasks", it will still process what you have already downloaded. It should switch back and forth between projects until all the queued work is complete. To handle work already downloaded you could suspend until after the Sprint, or cancel them altogether, but you have to do this on individual work units.



If it is in fact downloading work after you told it to stop, thats a bug I have not experienced.

 

[bfromcolo]

Doh! Will get something going.

 

[bfromcolo]

All my wanless2 tasks are still right here and are now all past the due date. But at least its just a couple hundred, one guy on the forum was claiming to have 12k.

 

[bfromcolo]

There is plenty of GPUGrid work now and its pretty stable. We could get a point with some more users.

WEPM2 event was canceled.

Running what I have, is there any benefit messing with multiple WU per GPU?

 

[bfromcolo]

speaking of late....

 

[bfromcolo]

I'll get something up and running now.