Topic: [ANN][AUTO-SWITCH] Profit-switch auto-exchange pool: CleverMining.com - page 276. (Read 554401 times)

can anyone team viewer my rig and set it up?

I will compensate

Best
matt

-I 13 --queue 0 -g 2 -w 256 --gpu-powertune 20 --thread-concurrency 11200 --temp-target 75 --temp-overheat 95 --temp-cutoff 97 --gpu-fan 40-100 --auto-fan --gpu-engine 1020 --gpu-memclock 1500 -o stratum+tcp://us.clevermining.com:3333 --scrypt -u 18MEFUcfHiQdEDMe3GKfW16FTAFhX6WhBA -p x --failover-only -o stratum+tcp://middlecoin.com:3333 --scrypt -u 1MbyxkJGxQzpzPCWa6dXnQYJvtNei1h9ny -p x

This is a exact copy of thje config on my r9 280x windforce(gigabyte), my shappire works on a similar config but I have a his r9 280x that need --cpu-engine 920 to give an average decent hashrate(650).

p.s.: dont forge the max_gpu_aloc and the other things

could u please paste your settings?

I saw this fight before in MC thread, In theory it should be equal, But if you take reject shares as waste of hashrate, it makes difference.

I don't think so, try to set lower intensity. I'm still doing tests but it seems to work.

As far as I understand this behavior should be good: http://d.pr/i/N6Kc

People that have a absurd ammount of reject:

Something is wrong with your config , internet or something else.I live in Brazil, so I think it would be fair to say I have , in average, the same ping from you guys in EU and I dont get absurd ammounts of reject.Maybe I cant get 100% perfect tweaked config but I have been averaging 5% rejects in a 24 hour window for some time  and my reject are in 70-80% of the time lower than the pool's,as seen on the stats page. Just for a comparison my pingtest to a server from california averages 200 ms

I switched to this pool about 6 hours ago. So far I'm getting only about 6-7% rejected. Initially it was a higher percentage, but I lowered the intensity on some of my cards that were running 19 or 20 down to 16 or 17, and that seems to have cured it. I'm not straining my cards so much, and I'm still making more money here than mining DOGE on multipool, so all is good

Nice front end to the site - very impressed.

Best of all, I don't have to mess around converting DOGE to BTC anymore, it's all done for me and all I have to do is sell my BTC.

Oh, I use r9 280x, so all my gpu's are set to -I 13. Terk, and other people, said that if you use I-20 normally maybe it is better to lower the intensity a little bit(one or two points). I can not comment over this because I didn't need to do it.

According to cgminer readme -queue is the number of work that your workers have downloaded to the server and didn't just started, hence the name queue.I had previously set it to 1 , because 0 didnt make sense to me, but a guy who appeared very knowledge on the matter said to put it to 0 a few pages back. I did and it certainly isnt hampering the performance, I hope it is helping but I cant really be sure because tweaking can be a tiresome procedure so I didnt follow a cientific approach to it: Change something, leave it running for X hours, see the result, start again.

I get over 50% Stale/invalid.
When i used wemineltc, i got better hachrate, and less then 1% Stale/invalid.

Do you think it's because i use GUIminer-script?

This is a good spot to start.  When you express things like "include VARDIFF and this pool will be better" then you are expressing a fundamental misunderstanding of how mining works.  Your hashes are not part of some global race to a solution, they are each individually capable of winning the lottery (aka share/block reward).  

Higher difficulty causes more variances, but if you are mining for more than a day or two, everything evens out and some days you will have lucky days where you get more shares than you should have with any given hashrate.  This stays the same whether the difficulty is 1 or 10,000.

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

Woohoooo, 1,015.7 MH/s! Nice one Terk 

where? i could not find it

Difficulty does not matter.

Go read up on how mining works since you clearly do not understand.

Could u please clarify what is the effect of "queue" exactly?

About the intensity parameter do you think? Are you using the maximum working for your gpu or do you think should be lowered to achieve a lower percentage of rejected?

I have been mining on this pool for about 10 days, with ocasional switches to other pools to test proftability, so I think I am gonna list some things that may help you.

- I am using the exact same .bat file that I used on other pools with exception that I put -queue 0 and to submit shares.I have tried --expiry and --scan-time but I couldn't notice any real difference so I just delete them form the .bat .

- the best way to measure reject rates is to leave the miner for about 1-2 hours then look at the site on your user stats, A good reject rate is one that accompany the pools reject but is always 2-3% lower. I gave up on trying to measure reject and stales by cgwacher and cgremote, I couldn't figure everything out.

- Vardiff will be implemented soon.It lauched that way but it got disabled to help with the troubleshooting.

- the fee is now at a fixed 2%.Terk is going to study a variable fee, so the in the good days everyone wins more and in the bad days everyone wins less,think this as a perfomance fee.IT IS NOT YET IMPLEMENTED and expect a proper announcement before it is.

-Terk is working in the site to reflect the recent changes in the pool, fixed fee and other things.

-Dont worry if you reject goes up in a interval, the pool sometimes mine fast coins.The things you must look are the long term rejects(should be close to 5% or less) and your rejects compared to the pool's.So if your reject goes up to 10% but the pool's is 15% you are fine.Even thought Terk doesn't reveal the formula of coin switches he said that it does take in account the high reject rate implied on mining fast coins.

-The pool now provides graph of reject rate of each user, just search your btc adress in the homepage.

With exception of the 2 first topics , I think everything else was said by Terk(the pool owner) somewhere in the thread .

Hope it helps

It will have an automatic vardiff of 16-1024.

Waffle's choice starts at 64, not 16. 16 Would bring too much load om the servers I guess. As soon as clever will give this option to set a fixed diff I'll be back.

so no-submit-stales should be false?

the thing is that on a regular pool without coin switch, my reject is bellow 1%, here i just tested again and got average of 8%.

on waffle now its bellow 1% (but waffle seems very unprofitable so far, havent tested for long, that sucks), i want go back to clevermining.

please set difficulty like waffle, when you can choose from 16 to 1024, it would be perfect

Please take note that stale shares are still paid. The 512 diff will be changed to 256 soon, and vardiff will be back up soon.