EbDj2V1zvtAZD2VXo8paF6YTKDcy9VhMAE

thanks!

'priv-i-yet'

I find it interesting having people posting in other languages occasionally.

Either I can see what bits I can understand, or just have the browser translate, depending on my mood/the language at the time...


There was an exchange in Chinese recently on the bfl forum which had some interesting opinion, although I am not sure of the significance of pumpkins!!!

Hi joolz,

just speculating, of course! To get a real handle on it we'd probably need to find the standard deviation of the distribution, then we could comment on how likely it is for any particular result to be  a particular percentage out, and indeed any particular *random* sample of results (rather than a selected group that catch our eye)...

Looking at my 7-day moving average from the stats page, however, it's followed quite closely my increase in hash rate and the difficulty change of 30.04, although the last couple of days aren't there yet, so I'm not concerned (at the moment...)

Doesn't preclude a temporary loss, of course. I did notice my cgminer report a missing connection briefly a few hours ago...

The first 3 were a bit lower for me, the last two a bit higher. May just be normal variability with c=210 (as per those maths links posted a few pages ago... )

Also I see pool hash rate now over 9100G - so I'm expecting lower payouts but more frequent payouts compared to a day or so ago when we were on low 8- something Ghash... (same daily payout on average until next diff change, of course).

FWIW: I am running a 7870 and a 7970 in the same PC. They are running from guiminer with just -v -f0 and -w 128 or 256 flags, but I set the clock performance from Catalyst Control Centre.

7870 is giving 430MHash/s at 1170/1200 engine/mem, with fan set to 60% and power -5%. At engine 1200 it crashes fairly quickly.  Runs at 66C, VRM1 at 71C

7970 is giving 645MHash/s at 1125/1000, 60% fan, -5% power. Runs at 67C, with VRM1 at 88C

One side is off the computer, the PSU is sitting on top, and it's in the cellar where it's nice and cool. Plus we're about 700ft/210m above sea-level, in a rural so that's generally less warm all round!!!!

Maybe time for a quick visit to the Recycle Bin to restore it - it may be useful!!!

Errm, now I'm lost! If you didn't go to AppData/Roaming/poclbm - what did you delete?

If you open a command prompt it puts you in C:/Users/YourUsername, so you should be able to do directly:

cd appdata
cd roaming
cd poclbm

and then a 'dir' should list the files, including poclbm.

and a  'del poclbm' command should get rid of it.   (Caveat - it is 'del' in windows isn't it, not 'rm'?)


Addendum: Or as auseeker has jsut posted, show those hidden folders!!!!

If you are on windows7, the file 'poclbm' is in the directory C:/Users/YourUsernameOnTheComputer/AppData/Roaming/poclbm.

It contains the information saved using 'File, save settings' about your various workers as well as the window position when saved. You can delete the file (thereby losing the saved setups, but also getting rid of any strange window positioning), and this is a safer option that trying to edit the file directly! It's easy to setup the information again for Slush!

Edit: assuming the window appears in a normal position when you restart it!

Note: Use the 'close' option on any gui-miner icons in the righthand tray thingy as you don't want to be restoring the mystery missing instance of the program!

End Edit

The 7970 has run pretty constantly for 5 days, the 7870 for 10+, with a few minor interruptions when I've decided to play with cgminer and options rather than let gui-miner run them. Not sure how long the 110C episode was - I normally check on the cards with GPU-Z a couple of times a day, and now especially after playing with cgminer because of the 110C event!

Have both at -5% power from Catalyst options, but this didn't seem to affect hash rates, temperatures or voltages.

Right-clicking on the task-bar icon hopefully gives three options - pause, restore and close. Restore should open it up for normal interaction.

I'm on Slush's pool, and that option should be in the 'server' drop down box. This will put the relevant info into some of the boxes for you.

Fill out your user.worker name e.g. psipower.worker1, and your password for that worker (which you have created in your account with Slush), select the relevant device - presumably your graphics card - and click 'start mining'. You may see it connect to api2.mining, but get transferred automatically to the stratum servers (more efficient, less fees).

You can save this configuration using File, save settings. Then if you re-start the program anytime, you just need to click 'start mining'!


Then it's time to play around with the other bits and find out what gives you the best rate!!!!


I found setting a CPU affinity box improved my rate.


You may find this helpful for various flags to improve your rate:

https://en.bitcoin.it/wiki/Mining_hardware_comparison

1 x 7970 running about 645Mhash:

1125/1000

GPU 71 C
VRM1 94 C
VRM2 62 C

fan-speed 55%/2552rpm manual setting. Can drop the VRM to 88C with 60%. On auto fan, lets the VRM1 temp go to about 110C.



1 x 7870 running about 420Mhash:

1160/1200

GPU 69C
VRM 79C

fan is auto, running at 50%/2050rpm.

Occasionally, as rewards are confirmed and being processed from 'unconfirmed' to 'confirmed', they disappear for a short while from 'unconfirmed' (and therefore the total) before then appearing in 'confirmed' (and therefore the total). Seen this frequently!

Very occasionally, a block fails confirmation (invalid, orphaned?), and thus disappears from the 'unconfirmed' without obviously being added to the 'confirmed' a bit later. Seen this a couple of times!

in fact ditto and double-ditto, muttley...

ditto!

bitcoin-qt, litecoin-qt, terracoin-qt, PPCoin-qt, plus namecoind.

Wallets encrypted and backed up to a RAID1 system.