I used public forum when I had an issue with shares. After that I posted details in PM, shares were recalculated and the issue was fixed (not for me only - for the EU server users, it was an accident case when stitching data from worldwide servers with Ethereum).

Also when my friend had an issue with CoinMine, he created a ticket using pool site and did not receive an answer. Then I post a PM to pool operator on forum and received almost instant reply that solved the issue. I don't like that, but seems that social networks and forums sometimes provide better support than emails/tickets. Don't ask why, maybe due to public nature of those requests...

I use it with ASrock H81 Pro BTC board, it has 6 PCI-e slots, one of which uses this splitter, and two cards are connected via it. So 7 cards in total. All 7 are Sapphire RX470 Nitro+ 4GB Hynix memory, and all 7 mine at exactly same ~212 Sol/s. This rig uses Windows 10. I've heard Win10 can use up to 10 cards, I do not know. I had issues with RX470 cards with Win7, no patch helped, 5 cards were maximum, the rest did not work at all (even 6th). Now I use Win10 To Go booted from USB flash, and it works perfect. It also works great with Linux. No drivers were installed by hand, Win10 found the PCI-e bridge automatically and installed GPU drivers.

That's exactly why few people wrote here that hash rate race is not good if no explicit reasons for that since it does not increase profit but decreases it due to power bills. No one listened, now it is the result: more power, more expensive PSU setup, etc. And not more profit :-)

Don't mix a warm with a soft (mining with shopping). When ETH dual miner was created, Claymore changed its default intensity from 16 to 8 to save few ms only. These ms cost some stale shares since after share is calculated, it should be delivered in time. Using -i 16 created larger jobs from CPU to GPU which can't be stopped until finished, and even those ms were important. Now we talk about Zcash, it has slower block time, but solutions itself are huge comparing to ETH ones (around 1300 bytes vs 50 for ETH), and all this now is encrypted in addition. So you have exactly the same case, but for some reason no one is concerned with it like it was for ETH.

Someone tells no one will see their addresses via SSL. If you are paranoid, you should not use exchange addresses. Install zcashd, mine to t-addr and move all to z-addr, send to an exchange t-addr, and no one will know who you are. No one except you will even know from what z-address funds came to the exchange. This is what Zcash was created for. You do not need SSL for that at all. Use right tools for particular task. I understand that 90% people here are home miners, they never read about Zcash: what is it and how it works. Still, it provides 100% security if you are paranoid and use it right. Without any encryption or closed source. That is the brilliant part of Zcash that is undervalued yet.

I used NiceHash when I had no time/wish to watch Zcash prices. Since its orders were paid for current Equihash (ZEC/ZCL) prices depending on the market, NiceHash paid in BTC more comparing to mining a particular currency (but took 3% of that in BTC before payment, that's a lot). Its customers watched for better algo, switched to it and paid more in BTC.

I actually hopped between CoinMine and NiceHash because when the payment rate @ NiceHash was high, I mined there and received BTC. When the payment rate was low, I mined @ CoinMine to exchange Zcash later.

I tried and stopped using automatic pools like MiningPoolHub because it does not give you exactly what you expect. It selects an algo for its instant profitability (network difficulty, current price, etc). But after you mine it, it matures, moved to an exchange, things may be quite different. So I preferred NiceHash when it paid well instantly, and CoinMine, when price was low (to sell when it goes high).

Now I left CoinMine, one of reasons is its PPLNS system since Dec 8th. It works bad in such scenario. Never liked PPLNS and always used DwarfPool for ETH. So CoinMine lost my hash rate instead of keeping it, since as said, PPLNS is not good for hoppers, and there is a reason for it as explained above.

I like DwarfPool due to its full stats and ability to talk to operator if you have questions. Every share is logged, so any issue always can be fixed by recalculation. It is small (because it did not enter Zcash race in time). It provided 50% to 30% of Ethereum mining, but only 2-3% of Zcash at the moment. The more people will use it, the less will be variance in payments.

Of course, people already shout here about missing SSL - I've heard it is in progress on DwarfPool. I use it just because I like it as it is.

I am seriously think that even if someone knew how to cut off the dev fee, he should not shout about it on forums. So he's lamer because he did that, and even more - he was trying to sell the software for this.

Really? But I don't.

Of course. But how? It goes far out of the topic of this thread, like Open Source vs Closed, Free software vs Paid, and so on. But do you agree with my points that it added nothing to end miners?

It is NOT about increasing the fee later. It is about creating a bad precedent.

Not at all. It is not an issue for real hacker to decrypt the code and "fix" it. It just adds some more trouble to him, and increases his payment for such work from big farms.

There are two points of view.

First is about dev fee as such and its protection. I do not argue about it, I agree that every work should be paid somehow.

Second is what good was added for mining and crypto world by introducing SSL layer on top of plain stratum. There is nothing good:
- more CPU overhead (without reason) - no direct impact;
- more latency for bigger packets (without reason) - there is an impact on stale shares, people already complained here, read the thread;
- more work for pool operators (not a lot, but also with no reason);
- more obscurity and doubt in general.

So this only adds DRM, and nothing more for miners. Actually I expected something like this after some lamer posted how to break dev fee with Optiminer. But I expected to see a private channel to some dev server (proxy) to collect dev fee. It would be harder for developer, but much better for community. But the way it is done now, sigh... Next step would be to close the source of cryptocurrency... Why not? Bitcoin was open and is open, and it has a value. Coins with something to hide (being that miners or core) are way to nothing.

Absolutely no.

1. SSL encryption increases data size and thus latency. Even if you have ultra fast CPU, it still increases it. So no win here.
2. SSL prevents packet editing. This is not a concern for miners, but it protects dev fee packets.
3. SSL is not enforced literally, but having 2% fee vs 2.5% enforces it anyway. Not miner dev requests it from pools, but people. Smart trick.
4. SSL does not allow you to check what is sent on wire as was said already. You can trust Claymore, but other may (will) follow the example.
5. If Claymore thought about miners' protection he would better add a simple password protection for management port. Seems few people had issues when found their rigs mining for someone else. Easy to set it to read-only, but adding a password is also easy comparing to extra work to add an SSL layer and actually introduce it for open protocol using trick with the fee.

Let me guess next possible steps (I do not insist they are planned, just are possible):
1) Remove plain stratum and keep SSL only (less probable).
2) Reduce mining speed for plain protocol (no fee mode) and still provide full rate for SSL channel (and ALL will demand it from pools). Why not, it's exactly the same as 2% vs 2.5% trick, only better.
3) In addition to channel protection (to keep transit data safe) add also pool certificate check. This means that only few pools with known certificates will be usable with particular miner. If pool changes it (in rare case) - new version of miner is created. This way no one will be able to remove the fee outside of the miner (even using SSL proxy). This is for those who wrote above it's impossible. It's quite easy, actually, and I wonder if it already is in place.
4) Step 3 means people will hack the code to remove certificate check or rather to remove the fee at all. No matter how good a protection is, it may be broken. I've heard about hacked miners on Chinese forums, this time they will be much more distributed.
5) Step 4 means more viruses in hacked binaries. Of course, we use only original binaries, but it is possible. I've heard someone had BTC stolen after downloading some Windows Zcash wallet after Zcash launch.

So to make things clear: SSL DOES NOT help miners or their security in ANY way, but MAY reduce profits due to increased packet sizes, thus, more stale shares. SSL DOES protect developer fee only, but introduces extra closed layer in open protocols that is a very bad idea. And finally, using tricks like different fee, it uses people (like sheeps, do you like it or not) to put pressure on pool operators to add encryption (for less fee or soon for full rate?) that is useless for mining as such.

Same is true for Claymore's miner. The difference is that Claymore's one shows you full rate with nofee option, but obviously drops some percent of solutions (seems to be around 10% according to pool reports) instead of sending to the server (that is documented). This miner, at least, shows you true rate.

While I appreciate Optiminer's software, let me comment this from a side point of view.

Plain fact: the devfee was first mentioned in the README only in version 0.3.1 (0.3.0 and below did not contain it).
Plain fact: the size of dev fee still is not officially reported.

I went to Optiminer after Claymore stopped Linux support. I found the name of the miner and looked at github to try it. I did not look at the source and was thinking that it is an open source (since github was made for open source projects, not binary releases). There was no mention on github (original distribution source), neither in README, that it contains any dev fee. I was amazed with its performance  and even more amazed when I read on forums that there is a dev fee in addition to hash rate shown. I can agree that it maybe was stated on forums, but shouldn't it be stated on the original source site and/or README, at last, if someone concerned about it?

If you look at the github, it contains original README evolved since version 0.1.1. Only in 0.3.2 change was declared "reduced dev fee" - then it was reduced from 15% to 10%. Please take a note that this percent still is not reported and never was in the README or forums by the author, and is the result of evaluation by curious guys.

So while I am not concerned about dev fee exact value, I should confirm that something said by your opponent was correct. I am not on someone's side, but for the true facts.

Talking about competition, at least now the statement is also correct after Claymore's v9 release. It mines at 212-215 Sol/s on my 470 4GB Hynix cards (downvolted, 915W from the wall for rig of 7 cards). Older cards have even better results: 234 for 7970 (and only 185 +10% fee = 203.5 for the same card with Optiminer), 200 for 7950, 312-315 for 390X. Claymore's miner was worse on Polaris vs Optiminer, but seems it is the winner now. At least while 0.7.0 is not released. Still, since Claymore doesn't like Linux, there is no other public alternative to Optiminer yet. Since hash rates were pushed to the top, all have to follow.

There is a case where you may want low difficulty shares. If you have bad internet connection where an established connection may be lost, you may lose share because when you send it, it will not be delivered (in time or at all). In that case a share that costs a lot may be worse case than few low cost shares. Of course, TCP should guarantee the delivery, but in real life you still may have "connection lost" reports. So not being a recommendation, this could be a reason for static difficulty port.

Agreed with you. Actually, the miner contains some countermeasures against this kind of attack. But anyway, it does not matter what OS someone uses, if one wants to cut the fee, he will do that.

It is not a driver issue, the version 1.1 worked very well with the same speed as a windows one.

He said many times that he specialises in Windows and does not like Linux (not because it's bad, but because it has own tips and tricks to be learned). But this time he hated people who said that he stole a big part of the code from another open source miner (silentarmy) because he used some part of source and did not give credits (this was explained as a mistake from his part). But since it was from "Linux people", he decided do not support it anymore and suggested to use silentarmy instead.

Another issue was that someone showed a disassembled part of his Linux miner code. While there is a lot of EXE file encryptors on Windows, it is not so popular on Linux since much software comes in sources anyway. His miner 1.1 for Linux was not protected (maybe by accident), and he disliked attempts to look into it.

ZCL has no 20% dev fee, and I don't really know what exactly is nicehash fee. It provides our hashing power to customers who pay for hashrate and point to pools of choice (ZEC or ZCL). In reality it might be more or less comparing to the actual exchange price at the moment, but in average it seems to be the same as mine and dump ASAP.

Still I am sure that all would benefit if there is no "fastest miner competition". Now any Equihash coins are like a dust. But power consumption is high due to high GPU loads. Nobody won.

It was 15% couple versions ago, now 5% added to the user, 10% left to developer. But still, when you see double performance, you even do not think that can be anything ABOVE that rate as a fee. Now solution rates are more or less similar. Not counting the dev fee, Claymore's miner seems to be faster on Polaris, but Optiminer might be wins with other cards (not sure, do not like Windows so did not try latest v6).

Actually, I (and a lot of people) would use Claymore's miner for its embedded monitoring feature. For Optiminer I had to write scripts to provide the same API for web-based monitor. But since there is no choice, I use that miner on Linux.

Claymore decided to leave the Linux competition.
Use other miners for Linux, say, this one (the rates shown do not include 10% dev fee of that miner, so its actual speed is even higher - around 200 Sol/s for R9 390, but you receive exactly what is reported):

My rigs use G3240, it is enough of it for all miners.
Low end was not mine, some single core Celeron 2800MHz, don't know more - was not interested, just helped to run the rig.

It is quite stable, there are issues with kernel error messages in logs, but it usually does not stop mining.
No yet watchdog if some card stops mining, but it is planned.
There is a Windows version (posted 2 hours ago, early alpha).


I have different cards on my rigs (7950, 7970, 390/390X, 470). I had 1555 sol/s with SA5, but 2525 sol/s with Optiminer. DevFee is really 15% but it is above hashrate shown, giving total performance of 2900 sol/s - twice from SA5 and Claymore's 4.0.
No one complained since it is not even written in the README about DevFee (IIRC), not everyone sees that. Also unlike Claymore that mines for some time and it gives rate drops, this miner mines devfee in parallel with main jobs. So there is no any drops in rate visible.
Pool reported hashrate is equal to reported by the miner. DevFee is not included and is "outside" of that rate.

Miner is slow on low end CPUs. I tried it on one such machine and had around 80 sol/s for RX470 x 2, but 155 with SA5. At the same time with a better CPU the same 2 cards give with Optiminer ~240 sol/s.

Quoting myself:


It's time now to think about usability, not only boost. Anybody with python skills wants to help with that (see original post).
If not, I will do it myself. But I would be ashamed to show my code in python to share since I have almost zero experience with it.

Actually, it depends...

pp_dpm_sclk shows current peroformance level and when mining it usually always 7.
It is possible to write 0-20 to pp_sclk_od and/or pp_mclk_od (power play boost in percents).
Written to pp_mclk_od value then gives increased mem clock in pp_dpm_mclk. But for one miner it really increased rate, for other I saw no changes and finally had to mod bios.

Confirmed. Added some stats to my post.
Included are: