Separate names with a comma.
Discussion in 'Distributed Computing' started by Gilthanis, Aug 8, 2014.
"squeak"....I gots older GPUs.
I hadn't been paying close attention, but saw this post over in the news section from a few weeks back.
Thanks to Grandpa's recent shift from Primegrid to GPUGRID, we are currently #4 on the CRUNCHATHLON GPUGRID [SPECIAL EDITION 7D - RACE XII], competing with the SETI teams.
All GPUs are welcome in order to claim rank #2 (Sicituradastra. seems out of range ?) after 7 days of [H]ard competition !
(And congrats to Grandpa for his 150 M milestone btw )
OK, count me in.
Great TJ !
Currently #3 and chasing SETI.USA...
You do realize you are making me move things around, I had to remove a GTX 680 because I could not run the 770 Classified at even stock speeds of 1200Mhz without throttling due to heat, it is now at 1350Mhz with a nice cool 50c. Moved the GTX 680 to one of the Intel 4P rigs running Ubuntu, man that was a bitch getting Boinc to recognize the card, but I got it running FINNALY now I just need to get coolbits to work correctly
I am going to try and get a GTX 580 and a 460 going on a couple of the other 4P's but no promises there after the PITA it was to get the first one running.
Thanks for the congrats I also had another milestone recently which I was rather proud of, I crossed the 1,000,000,000 mark in total Bionic credits.
I noticed, GPUGRID is back above 1,3 petaflops computing capacity . Great ramp up, we are boucing between #2 and #3, with a worthy competitor.
For coolbits and Keplers, if you use coolbits = 16, you can play (not too much ) with voltage with 346.x drivers and above.
And congrats for the mega-milestone !
I placed second in one of the performance rankings for long runs.
Great ! We are #2, still 2 days and 21 hours !
(And... I passed 400 M at GPUGRID tonight, Glutamine badge )
It's mindblowing that one user has racked up 30,000,000 pts for the challenge.
You should recruit them....
One can't blame me for trying to troll Gilthanis?
Just keep it up and I will keep on LOL, we can always use a good chuckle.
We need a damn LOL emoc.
Anything would be an improvement versus a clapping pony.
Meanwhile, with a little over an hour left, we have a lock on 2nd place.
Yes, we achieved a great second place, congrats !
As already pointed out, maybe the greatest achievement was to have the [H]orde slightly overperforming RaymondFO (alone) by 0,5 million points
I tried so see what sort of stuff that RaymondFO was running but that info was not available. Maybe that's what the FO stands for.
Well, he is #2 in RAC for the project, quite close to #1 Stoneageman (0,25 Mpoints RAC gap). This guy is using 5 x 780 Ti, 4 x 780, 2 x 960, 5 x 750 Ti, 5 x 660 eg 21 cards on 10 linux hosts.
Fired up three of my Nvidia cards yesterday, in order to get to my 100 million point goal for this project. I currently have ~33 million to go.
Been tough getting any WUs for this project since they went on Christmas vacation. Hopefully things get back to normal soon.
Question: has anyone been getting beta WUs for their AMD cards?
I ordered one of those R9 Nano cards yesterday since they dropped the price to $499. I was going to wait for the new cards coming out later this year but needed something now that could run a triple monitor setup for my sim racing rig. It looks like I can run E@H on AMD but would prefer GPUGrid if possible.
And then you have
Ouch; that sounds not good and in line with the suggestion from DA to move to a different way ...
A lot of places are thinking that way because they don't have skilled individuals that stick around for more than a few years that can maintain the projects. In this case, they don't have someone that is familiar enough with the server software. Since BOINC really isn't being seriously developed right now, you take it for what it is. This means any new technologies thrown into cards may not be supported. New co-processors may not be supported. Too many of the new projects require quick short runs and so just need a quick fix to get their research. That is one of the hangups for WCG. The scientists have to guarantee a certain amount of work to justify all the man hours of onboarding the project in the first place. Universities should really consider utilizing their programming depts. and other IT areas to help with maintaining the platform. That could go a long way. However, it could also look like favoring select depts....
GPUGrid has not just had application problems, but also work unit shortages. I don't know what is going on behind closed doors, but seems like they are struggling on all side.
One guy from GG joined PG ;-)
and getting worst looking on the server stats page: mostly zero work; some occasional WU dripping in ... #sad
If cards running too long dry or for me meaningless stuff like collatz I need to rebalance back to FAH ...
If I remember correctly, work availability tends to run dry over the weekend more often than week days. Someone they need to turn on the work unit generator...
Let see if you will get some over the week days.
I used to run POEM which is GPU project for protein folding but the project has ended and they are willing to share the open source code if anyone is willing to pick this up.
seems GPUGrid is back; still lower level in supply but at least I got again some work. They said something like IT messed up some IP settings thats why something broke. .... as long work units are back I'm ok ...
Nice catch ...
This coleslaw guy cause trouble with the page rendering ...
Tooooo many badges
Shame on him.........
Yeah...they still haven't even cleared those out of my list...