Folding Forum

Hey all, I'm just wondering about Work Units and Points. Last couple of days I received what seems like the wrong number of points with WU's. I've been getting pretty much 480 points each
WU, but the last day or so, it's hit or miss. i.e. 960 points with 2 WU's and then 960 points with 4 WU's. Or 480 points with 2 WU's. When I look at my log, they all go to 100% with no early work ends, or shutdowns.

Anyone else experience this?

Thanks, T

PS, Here is my last 24 hrs


Last 24 Hours Production
Time Points WUs
01.10, 6pm 960 2
01.10, 3pm 960 4
01.10, 12pm 1,070 3
01.10, 9am 1,070 3
01.10, 6am 1,070 3
01.10, 3am 1,623 4
01.10, 12am 960 2
01.09, 9pm 480 2

The points on a given WU changes depending on what is needed. Certain WU's give higher points because of the hardweare demands it makes on your system. I have seen points from a low of 42 to a high of 2150.

Thanks for your reply. Maybe
I didn't explain myself correctly. According to FAHMon, all of my work units were 480 points except for the few WU from my PS3 which allowed me to have 1070 points with 3 WU's. I'm referring to to 480 and 960 points with different associated WU's. Again-i.e.960 points with 2 WU's and then 960 points with 4 WU's. And Or 480 points with 2 WU's. By the way 460 times 2 is 960. Please look at the production and you will see the obvious descrepencies. Starting off 6PM 960 points with 2 WU's, 9PM 960 points with 4 WU's. What gives? According to FAHMon all of the WU's were worth 480 points. Am I recieving too little point or too many WU's

T


Last 24 Hours Production
Time Points WUs
01.10, 6pm 960 2
01.10, 3pm 960 4
01.10, 12pm 1,070 3
01.10, 9am 1,070 3
01.10, 6am 1,070 3
01.10, 3am 1,623 4
01.10, 12am 960 2
01.09, 9pm 480 2

It looks to me looks you're turning in some WUs for 0 points. This could be due to duplicate WUs being assigned due to misconfiguration of the client. If for instance you have two GPU in one machine and both have the default machine ID of 2, the Assignment Server sees them as the same GPU instance. Check your configurations for duplicate IDs.
The default assignment behavior when an instance requests a new WU without having returned the last assigned is to resend the last assigned. If two instances have the same user and machine IDs, both will get assigned the same WU and you will only get credit for the first turned in.

Thanks, Could be cause I tried to set up a service a couple of days ago. Crashed the computer numerous times. Had to delete it. Maybe this is residual effects.

T

Chas, from what I have experienced with a crashed box, it appears that the Stanford servers must see your mac address and uses that to do the assignments. I had a crash that lasted over three days with both the program and my system. During the restarts with a clean slate, the server sent me the same WU.

pmasley wrote:Chas, from what I have experienced with a crashed box, it appears that the Stanford servers must see your mac address and uses that to do the assignments. I had a crash that lasted over three days with both the program and my system. During the restarts with a clean slate, the server sent me the same WU.

Stanford does not use your mac address. They assign a unique value called a UserID to each machine. Check FAHlog.txt. The values you enter are xxxxxxx and yyy. Stanford assigns zzzzzzzzzzzzzzzz.

Thank ya Bruce, I wondered how they kept us straight!