Topic: [ANN][Pool][Profit-Switch][Optional Auto-Exchange per Coin][Vardiff] ~ Hashcows - page 148. (Read 347329 times)

I don't know, this is making me crazy, I'm doing all the combinations of GPU from 900 to 1100 MHz vs memory from 1000 to 1600, in 10 MHz steps -> 1200 combinations. And this is only a 2-parameters experiment. 

Truly fucking unbelievable. I wish I could be of more help but mine are quite common and conservative settings (sans the overclock, without which I was getting ~620KH) so I maybe your card is faulty?

These shitty cards are amazing... With any other settings I get from my 7970 more than 540 KH/s (but no where near 700 KH/s), with your exact settings I get 413 KH/s. Fucking unbelievable...  

Ah, so this is overclocked. Are you also using the --verbose flag to see if you're getting any HW errors on that intensity?

I am getting 730KH on my 7970 using these:

--thread-concurrency 8192  --gpu-engine 1050 --gpu-memclock 1470 --gpu-powertune 10 -I 13 -g 2 -w 256 --lookup-gap 2 --verbose

Sku is 11197-03

intensity 20
thread concurrency 40960
shaders 2048
gpu-engine 1100
gpu-memclock 1500
powertune 0

> Not that I'm a big fan of Windows,  but I'm still resistant to trying to deal with linux at the moment,  since I really don't know anything about it.

BAMT is easy to run.. Just put on USB and Boot up... goto the file manager, edit the cgminer.conf file to your settings and off you go.
can't get simpler than that.

> Can't say that I agree with you on anyone being stuck at I:13 because of windows.  From what I understand, that's usually due to running an sha-256 kernel.   Perhaps I am mistaken.  I'm running a scrypt kernel since I am mining scrypt,  which allows me to go to I:20 with stock versions of cgminer.  
Changing from I:13 to anything higher will definitely show an increase in hashrate on my system. I can crank them up to I:20 and get 650 a card,  but temps, stales and rejects seem to go up and my WU-to-hashrate percentage is lower as compared to I:18.

Windows itself has the limitations/shortfalls.  I can set SHA256 to I 20 but get low hash rates.  First, find your I level you get most with, then tune your settings then retune the Intensity again to see which works best again.

Thats strange, I have never been logged off automaticaly, unless I have switched computers / browsers

Gentlemen,

Great job on the recent website upgrade !

One question / idea :
Is there a way to change the new auto logout interval time ?
I get auto-logged-out in like 15 minutes or so. Way too quick.

I'm browsing my tabs in about 60 minutes or so....

thanks for the reply nearmiss. i appreciate it.

also thanks for the info, not counting the rejects in hashrate calc makes sense.

is there any way to improve these high rejects?

Thats correct.  The high rejects also affect the pool's hashrate calculation, as we don't count rejected shares towards the hashrate.


Regarding TAG, it takes 200 confirms to mature, meaning 200 blocks have to be found on the network before we actually have the coins.  It has a 4 minute target time, meaning if the network is running at optimal hash/diff ratio, it takes > 13 hours for a block to confirm!  Generally speaking, it likely goes a bit quicker than this, but of course then difficulty will adjust upwards to even it out.  

While it can be annoying in the short term, tag has proven quite profitable for us since adding it when it launched.  We mine it fairly often, and so generally have an older round of it maturing at a fairly steady pace.

what about the varying hashrate?

I believe that those high reject rates are normal for some coins and the profit calculator algorithm takes that into account and so, even with those high reject rates, it's still profitable as expected.

It seems this pool is spending quite a lot of time on TAG but I don't see any payouts from it.

What's up with that?

I'm also seeing very high reject rates when on certain coins. and varying hash rates based on coin (as reported by your stats), but my hashrate reported on my miner seems pretty stable.

for example,
high hash rates (600-700KH) and low rejects (1-2%) on TAG
low hash rates (100-300KH) and high rejects (60%+) on FST

thoughts?

If you don't mind mining two pools at the same time you can use the --balance option instead of --load-balance, or --failover-only for when it loses connection to your pool forcing it to switch.  Balance will roughly split your hash power between 2 or more pools, load balance will keep the same amount of work going to 2 or more pools(ie. multipool and forced 64 difficulty giving 1 share, hashco.ws getting 4 shares at 16 diff) and failover giving all your hashpower to the backup pool when your primary crashes( I.E. hashco.ws crashes on tag, your backup pool of multipool, picks up all the slack and mines like a mofo until cgminer resumes connection on hashco.ws and then it switches back completely.  There is a round robin option as well that lets you set a time limit between pools, I E. switching every 12 hours, and another one which I've never used and can't remember.  I'm still trying to get cgwatcher to play nice some days.. Has a habit of resetting the computer after a driver crashes and never finishes the reboot. Your mileage may vary with different options, but I've never given a %age to my load balance and it works fine, and I usually use failover.  I'm just experimenting with balance and loadbalance to see which gives the best option between the two pools.  

As for intensities, a 7950 will mine all the way up to 20, but a 7970 will usually crap out above 13 in windows.. I think the poster above may have mixed up your cards and linux doesn't have voltage lowering programs unless you hardflash the cards with a lower voltage as far as I remember reading(this might have changed recently, I don't use linux).

Estimated earnings should come back in line sometime within the next hour or so.  No affect on payouts ect.

...well I got cgwatcher to take care of the last connectivity issue on it's own, at least. It noticed a loss of accepted shares after a couple of minutes and restarted cgminer automatically.

I also loaded cgminer 3.2.1,  just to see how that runs.

I feel a little defeated though because I cannot get cgminer to run smoothly by itself with the load balancing commands that you are suggesting above.  I put what you have up there for quota and the number and semicolon and it would not go. I checked the cgminer readme file and it said to go with "url" for the first pool and no number,  and "quota" in place of "url"for the second one, with the number and semicolon like you have it up there, but it was still rejecting the "quota" one at that point. Idk.  Kept telling me that there was a json typo for the quota one,  but I basically cut and pasted a known-working config file into 3.2.1 folder and did the few little changes with the urls to try to load balance and had problems.  I'll hit it again later on,  right now I'm just going to let cgwatcher t.c.o.b. .

It's just acting weird - if I add up my estimated earnings per coin since midnight, I have earned a lot more than the estimated earnings that is getting displayed (around an order of magnitude).  It's like a decimal is out of place or something.  No worries, even though the estimated earnings acts weird frequently, the end of day payout is pretty consistent!

Estimated earnings on the website are way down by over a full place through the past few hours as compared to the past few days, not sure why.

i left my miners with an average of 1800-2000 khash on for 20 hours on here, and all i got was 0.005 BTC...can you check whats wrong, username: bigd, i should get at least 0.02 putting it mildly...

Model and cgminer settings please.