qtMiner, eth-proxy and genoils miner all connect to the same backend logic. All of them are using vardiff shares so either one miner is more efficient than another (I doubt that as they all use the same opencl kernel) or it is an issue with your local setup. How long have you been mining with either miner? Usually it takes several hours to display an accurate hashrate average.
Been using genoil for about 6h, and when i posted i was using qtMiner for about 1h, i just checked now and it display roughly the same MH/s on website so you are right.
Yes, while cudaminer gives a better hashrate on nvidia cards the performance should be the same for AMD cards. Both are preferable compared to eth-proxy as they do not use polling.
@dr_pra If your running an eth-proxy on each rig, so you're only polling to localhost, do you still see qtminer or Genoil's stratum-enabled miner as measurably better? That also gives you a better failure scenario than one central eth-proxy for all the rigs on your local network. Thanks!
@dlehenky yes, even if you only poll to localhost you have a polling interval which is larger compared to the push feature of the stratum protocol that sends new work nearly instant to the workers.
Ok, so which one is the best for your pool, I just connected yesterday through qtminer as recommended, should I change now to eth-proxy to use the new server as i am in europe? Or should i stay on qtminer and us server?
@zorvalth Are you asking should you use both? If so, the answer is no. You use either eth-proxy or qtminer. They both use stratum connections to the pool. Does that help?
No I'm asking which one is preffered, i know you like stratum more because of the less log, but i want to hear his opinion, is there any difference in the rate as the qtminer is without proxy?
I just got mine first block but its 3.7125, which means its uncle... Isnt there a better way uncles to be distributed?...
@zorvalth The big difference between eth-proxy and qtminer, which favors qtminer, is that eth-proxy, although it has a stratum connection to the pool, still has to be polled (farm-recheck) by 'ethminer'. With qtminer, the stratum connection to the pool is built in to what is essentially an 'ethminer' clone, so there is no polling for work by the miner, there is no farm-recheck. When qtminer receives new work from the pool, it pushes it to the miner code internal to qtminer. HTH.
You got the uncle because it was your turn for the next block, which happened to be an uncle. I've gotten two in one day, although that's all I've gotten over 5+ days (60+ blocks).
@zorvalth our uncle rate is one of the lowest among the large pools and while it is still bad luck to receive an uncle it will happen to you exactly as often it will happen to you when you do traditional solo mining.
you should use qtMiner if you are using AMD cards for the reason mentioned by @dlehenky . If you have NVIDIA cards you can either use Genoils latest beta miner which has built in stratum support or Genoils latest stable cudaminer in combination with eth-proxy.
Once @Genoil releases his stratum enabled version as stable it will be the recommended miner for both AMD and NVIDIA cards.
Hi. Just switched to this pool using qtminer. Share is submitted every 2 seconds or so. Is that correct or can I/ should I try to optimize that to 1 second or 200ms, etc? Thanks
Valid shares should be submitted every 1-2 minutes, as after each valid share your miner will pause for a brief period. Vardiff should take care about that and will adjust the difficulty so that it matches your miners hashrate.
Hi. Just switched to this pool using qtminer. Share is submitted every 2 seconds or so. Is that correct or can I/ should I try to optimize that to 1 second or 200ms, etc? Thanks
You are mistaking submitting share with the time of the miner refreshing the speed.
Hi. Just switched to this pool using qtminer. Share is submitted every 2 seconds or so. Is that correct or can I/ should I try to optimize that to 1 second or 200ms, etc? Thanks
You are mistaking submitting share with the time of the miner refreshing the speed.
That's helpful and constructive mate. You really lost me there :-( . Why? How? How would you do it?
Anyone else showing less hash rate on here than mining on other pools or dierctly solo? I'm running about 20% less overall hash rate, even though ethminer shows quite higher. (I'm using the genoil 1.0.3 driver.
I will try the stratum proxy per rig directly to the pool but haven't set it up yet.
Anyone else showing less hash rate on here than mining on other pools or dierctly solo? I'm running about 20% less overall hash rate, even though ethminer shows quite higher. (I'm using the genoil 1.0.3 driver.
I will try the stratum proxy per rig directly to the pool but haven't set it up yet.
Same problem with my rig on coinotron as well for some days. Don't know what is going on. Switched to ethpool and will have a bit more patience with ethpool then decide. Huge fluctuations as well. The only one that was kind of close is nanopool. Maybe the reporting system is flawed because I actually submit very close to my expectations?
Dwarfpool showed the right amount (maybe 3-5% less than what ethminer averaged, but now ethpool is about 20% lower.
On the other hand, I now see what people noticed on pools with the crazy high fluctuations. Ethpool shows a 140MH difference at times on a 300MH rig I'm using to test it.
Hopefully the stratum proxy will fix this issue.
I tried qtminer but stopped after an hour as it was consistently showing 15% less hash rate locally than ethminer.exe did.
Kind of having the same conclusion. Don't quite like it. Just done the math on coinotron and the payouts reflect the drop... so charts should be accurate. Let me know how it goes for you?
ethpool.org shows the effective hashrate that is calculated based on your submitted valid shares. If one of your workers produces invalid shares your displayed hashrate will be less. Fluctuations are normal as the share difficulty is quite high to keep your miners busy. We could lower the difficulty to increase the accuracy of the displayed hashrate but this will decrease your miners efficiency (each start/stop causes some lost hashrate). After several hours the displayed average hashrate should match your local one. Also the payouts should match to what you would get when solo mining. Please note that nearly every mining calculator available does not take uncles into account which will cause it to underestimate the overall network hashrate and to overestimate your expected mining revenue.
Another question popped up in my mind. Right now the difficulty is 14.987 but in credits the highest miner is 16.300 pts. second one is 16.100. So the first one will get the next block and will be reset to 0.200 pts. Which makes 1.100 pts totally lost... its like 7% of mining lost. Isnt it better when the miner with 16.300 gets the next block to be deducted with the current difficulty and set to 16.300-14.987 = 1.313 pts? I understand that if the pool got more luck in the moment sometimes the miner will start with minus credits but still seems more fair... Correct me if my thinking is wrong...
We calculated both methods, in your example, if the pool is lucky the miner will get negative credits which we are unable to enforce (he can simply switch to a new address). The current method is fair during both lucky and unlucky periods. During periods where the pool was lucky no miner complained that he did receive a block even though his credits were way below the current difficulty.
@zor Depending on the pool's "luck", sometimes the highest miner will get the block *before* his credits get to the current network difficulty. It works both ways. I've gotten blocks when my credits were 1T below the current difficulty. As in all mining, there is, and always will be, variance. That's just reality. As far as the current formula for resetting the credits, that's simply the way it works, and I don't think your suggested change would be any better, personally.
@dlehenky, the only problem is that with the current scheme a person can keep getting lucky or unlucky over and over again. You have a big hashrate and it really doesnt matter. Mine is ok around 350-400mh/s and i will be probably ok and our luck will get even over the time faster. But if a miner with less then 100mh/s who needs 2-3 days to get a block got 3 times in a row unlucky that sucks. With the method of deducting current difficulty the good and the bad luck will be fixed right after every block and not by the time which could take weeks or months for a miner with small hashrate. I really didnt though that if somebody got minus pts can change the address but i personally doubt hat somebody will put a new address on every new block but its still possible. The other pools dont really have this problem because they split the ether immediately and your luck doesnt depend on the very moment when you are completing a new block which as i said for a miner with small mh/s could be a problem..
Its probably just me fell little out of luck, my first block was uncle and the second will come around 16.*** on difficulty a little less than 15.
@zorvalth Yes, I've been a little down lately, too. It all evens out in the end. Just got an uncle on my last block, but ethpool does average less uncles than the other pools.
Comments
Been using ethpool too and it's great, i like the concept.
But i don't know wich way to mine is the most efficient;
When i use qtMiner i get 80MH/s stable on console, and 65MH/s average on website (?).
With genoil 1.0.3 and eth stratum proxy i get 85/90MH/s on console, and 100MH/s average on website (?).
Are vardiff shares (the only benefit from gtMiner IIRC) really worth it?
Thanks!
So it's all down to personnal preference then?
Thanks again!
-Best Care
David
Please note that the new server only supports stratum enabled mining. Please switch to a stratum enabled miner for an optimal mining performance!
I just got mine first block but its 3.7125, which means its uncle... Isnt there a better way uncles to be distributed?...
You got the uncle because it was your turn for the next block, which happened to be an uncle. I've gotten two in one day, although that's all I've gotten over 5+ days (60+ blocks).
you should use qtMiner if you are using AMD cards for the reason mentioned by @dlehenky . If you have NVIDIA cards you can either use Genoils latest beta miner which has built in stratum support or Genoils latest stable cudaminer in combination with eth-proxy.
Once @Genoil releases his stratum enabled version as stable it will be the recommended miner for both AMD and NVIDIA cards.
Thanks
(I'm using the genoil 1.0.3 driver.
I will try the stratum proxy per rig directly to the pool but haven't set it up yet.
Don't know what is going on. Switched to ethpool and will have a bit more patience with ethpool then decide. Huge fluctuations as well.
The only one that was kind of close is nanopool. Maybe the reporting system is flawed because I actually submit very close to my expectations?
On the other hand, I now see what people noticed on pools with the crazy high fluctuations. Ethpool shows a 140MH difference at times on a 300MH rig I'm using to test it.
Hopefully the stratum proxy will fix this issue.
I tried qtminer but stopped after an hour as it was consistently showing 15% less hash rate locally than ethminer.exe did.
Just done the math on coinotron and the payouts reflect the drop... so charts should be accurate.
Let me know how it goes for you?
Its probably just me fell little out of luck, my first block was uncle and the second will come around 16.*** on difficulty a little less than 15.