IIRC, it takes around 3% of your hashrate as a payment for the distribution. I may be wrong, or thing may be changed since I looked at that. Check for yourself.
Still, would be good to have a way to build such USB-based system from scratch. I am now doing own version to run fully from RAM with USB stick set to r/o after boot. But I am afraid I can't take money for free software, and am too lazy to write detailed manuals for free (this needs much more time than I can afford). I like some distributions, but I definitely want to build it myself, not use prepackaged one.

Would agree with you but do you really write on your window: "You can do anything you can in the room if you are in?". THAT is the case with TheDAO, its own contract allowed him to get some ETH technically and their own agreement explicitly allowed to do anything legally what the contract code permits. This is the key difference unlike, say, banks where your interest is declared by contract and any error must be fixed by support. Please notice also that he was not a person aside, he was legitimate TheDAO investor who used contact as it was written. There is no documents that declare his profit on DAO split, that code was not even commented well enough. It is their fault, and any fault should have consequences. In addition, you tell me about the spirit of TheDAO, but what about the spirit of Ethereum? Again, I don't support that he has done, but formally nothing was broken or hacked. And that was the only thing I disagreed. He just read the code better then authors or security auditors. You don't ask Intel to fix a CPU to recover an error in an accounting program. Why you expect the Ethereum Virtual Machine (the CPU) to be fixed when it worked as expected, and the flaw was in upper level layer, TheDAO (accounting program)... Well, no more offtopic from me. But I am ok to discuss in PM if anyone has arguments, not emotions.

Your answer regarding new miner was correct. But your statement that it was a hack, and ETH was stolen was NOT. In fact it was non-ethical use of the contract code as it was written by TheDAO, used in accordance with contract agreement that explicitly stated that the code is law (unlike bank service contracts etc that do not contain such statement and cannot be used as precedents).

So I quoted only related part of your message I disagreed (overquoting is a bad habit) and explained what was wrong with that part of the message, and why any Ethereum fork is a very bad thing since it makes a precedent that irreversible blockchain may be reversed if "necessary", even if it does not contain bugs in Ethereum platform. There were no any references in the reply because anyone interested can find them, and all I wrote were my personal thoughts and words. If your answer contained a quote from some other source, then a reference to that source should be nice to have as well. In that case there will be no questions to you, but to the author of that incorrect statement.

And that all was written by me because a lot, if not most, of pool miners use Claymore's one. If they believe that something really depends on their choice (to fork or not to fork), they at least should know what has really happened and make a right choice voting. Your statement gives no such knowledge, but tells without references that TheDAO was hacked (wrong - no code changes), ETH was stolen (wrong - use of contract code by one of many TheDAO token holders - means investors) by hackers (wrong - was no hack). And, BTW, TheDAO is NOT a part of Ethereum project, it is just one of thousand of projects based on Ethereum platform, just biggest one where curators happened to be Ethereum founders. But it seems that there is no any real choice if you are a pool miner (I explained why). So I personally switched to Ethereum fork due to all that mess.

Here you repeat that lie that mass media tell you that you believe.

Hack means illegal code change to get some advantage. But TheDAO was just one (but biggest) of many projects based on Ethereum. It used smart contracts written in a language that run them on Ethereum Virtual Machine (EVM) - that is, on all miners. The contract agreement tells you that "The contract code is law, and if there are different human interpretations, the code runs as it runs, and that is the final decision".

The guy or group called hacker is in reality a smart guy who studied the open source code of contract and took advantage of some "features" to get some ETH tokens from TheDAO account. From ethical point of view he was wrong. From law and technical point of view he did not hack or hijacked it. It was the LAME CONTRACT CODE written by TheDAO that allowed him to do so. And their own agreement tells that it was done in full accordance to the agreement - the code on EVM run exactly as written, and the code is law.

Please notice that TheDAO knew about this code vulnerability and in the beginning of June they wrote that "nothing bad can happen" and did nothing. It happened 10 days later. EVM run as programmed, the flaw was in TheDAO contract only.

Also notice that main TheDAO token holders are the same people as Ethereum Foundation creators or affiliated. So takes place the conflict of interests.

Finally, TheDAO was just one of investing projects, and every such project has risks. Sometimes they simply fail, and any investor should check and accept risks.

Now those people suggest to do a hard fork that means to take ETH from the DarkDAO account (where ETH was transferred to) and return them to original investors. From ethical point of view it seems to be ok. BUT from technical point of view such fork to FIX JUST ONE BIG PROJECT OF hundred other projects IN FAVOUR OF 3RD PARTY means BREAKING THE MAIN IDEA of Ethereum as a machine that runs code that cannot be arbitrary controlled by a 3rd party - people, groups, governments, etc.

If that was another smaller project - no one will break rules that blockchain is irreversible. But TheDAO is too big to fail, and Ethereum Foundation affiliated with TheDAO token holders. So they break own rules and actually KILL THE ETHEREUM by reverting the blockchain history. Saving one big project, they kill the Ethereum ecosystem.

Now they should make you believe that it is not an arbitrary change, but a network consensus. But please take a look how it was achieved. Take for example, the vote results of DwarfPool. They asked people do they support the fork or not. Then they will install or not the modified Ethereum code. But heh, here are the results: http://dwarfpool.com/eth/voting  They tell you that network decided to fork because 71.80% pro-fork, and 28.20% contra and install forked code. So it was a network decision. What they do not tell you that only 6% of people or 4.37% of hashrate voted at all. The rest basically ignored the voting and that should be treated like they do not want to change anything. Or, if you like it more, they do not know about voting. That is the trick to make you believe that forkers do the "right thing", while actually they do not. The good voting should be like on coinotron: they have two different ports, and no one of them should be the default port. So to mine there people MUST make a change and choose one of options, not be silent. In that case it will represent the network consensus.

TheDAO is just one of many projects based on Ethereum. The EVM worked exactly as it should. The Ethereum code has no flaws and worked as it should. TheDAO contract (badly written) worked exactly as it was written. So TheDAO requres fix, not Ethereum blockchain. And it is wrong to make Ethereum as NoTrustCoin just because one big project failed. The right thing is to accept losses, fix TheDAO and go ahead. And do not break Ethereum platform. But big money talk. And (my speculation) it is easy to pay something to maybe 10 pool operators covering 95% of market to make them install the fork.

Think: how big should be loss in another project to make fork for them? 1M? 10M? 1K? Or? Who should be project owners? What if some government will be interested to block some addresses? How long should be the black list of "undesirable" accounts? What if some "secret service" will be interested to fork it and/or block someone? Many questions. But definitely known that the Ethereum slogan on the 1st page of its site:


after any fork to "fix the damage" will be wrong. The code of TheDAO contract run as programmed, but innocent Ethereum code and blockchain will be changed by 3rd party interference. And Ethereum will be NoTrustCoin.

Last note: assuming that TheDAO was most hyped project based on Ethereum, the ETH price maybe raised due to it. And is possible that after fork the same people getting ETH back to the failed investment project and breaking own contract agreement (code is law) will keep the price high. Maybe. But it will not be that Ethereum running unstoppable code. And I can predict a lot of similar projects that tell you explicitly that they will not censorship own blockchains in any way. And they win. Even now some ETH forks are more profitable - the fork is planned on 20th regardless of my or your opinion, and the future of Ethereum depends on the results.

PS. Sorry for offtopic, but it was a reply to question. I hate when some mass media "prepare" people for the move that some 3rd party players are interested in, but trying to make you believe that it is a network consensus when just 10 pools and few exchanges take decisions, in fact, any other miners accept what their favorite pools prepare for them (that's why no change in Claymore's miner). No one of Bitcoin specialist told anything good about the fork. And that's why Bitcoin is a trusted currency. Sad I can't say that about ETH after fork if it happens.

Confirmations are block finds, not a time lasted. Current difficulty is a bit high comparing to the network hashrate. So it takes more then 24 hours to confirm. When there was no dual miner it was vice versa - confirmations took 15 hours, not 24 or more.

Due to a lot of people going to dual mining ETH+SIA the profit every day decreases. The more miners, the less profit. It is yet profitable, but not as much as was 10 days ago. And don't forget that it takes much more power that means increased electricity bills, higher power supply loads and higher temperatures.

Hardware problem with GPU0 that hangs, maybe due to overclocking/downvolting.
Either fix it or add -r 1 and a reboot script file to reboot the rig. Miner restart usually does not help in such case.

Actually, unlike nanopool, this pool claimed payments starting from 2000 every 6 hours and did them. But recently it had serious delay with payments, so I was thinking to jump back to nanopool (that also had delayed payments couple days ago). Maybe that was due to PPS scheme where it must pay coins before they confirmed by network and actually paid to pool.

Anybody knows how admin of siamining.com can be reached? Forum/thread? I didn't find it, and no direct contacts on the site.

As I already wrote you, you can update whole config and restart the miner remotely (Claymore suggested to use wireshark to catch JSON requests/responses for API requests). But I believe there is no way to change -dcri remotely without restarting the miner. It might be done for you if you ask him nicely

That is weird since OpenCL code was not changed. But see below.

I can confirm this issue with 380X card and Windows rig. For my case I do not restart miner, but instead just change clock options using Afterburner once, and immediately set them back. And hashrate jumps up by 1MHs exactly as you tell. I remember someone already wrote here about such effect. So you may as well try to change clocks twice using atitweak with some delay in between. Maybe it helps.

It should not be difficult. I use a nodejs utility to fetch stats from miner rigs and provide JSON data for web page . It's possible to add a code to update miner config since the API exists. I think, miner restart will be necessary after each change, thought, so fetching stats will take some time. But the idea is great. Hope the calc is open source.

You may be surprised, but those "USB" risers have nothing common with USB 3.0 standard as well as with USB at all. The only common is that they use USB 3.0 connectors and USB 3.0 cables (more pins and wires comparing to USB 2.0 and below). They only extend PCI-e wires using high-quality cables normally used for USB 3.0. Also due to limited number of wires in USB cable they do not supply 3.3 PCI-e voltage to GPU and need 5V separate power apart from 12V (both supplied with Molex connector).

So I would not expect that even cables from different "USB" riser manufacturers will work swapped (even if they should work being pin to pin wired).

- if I enable internal Intel CPU video, miner doesn't see any AMD cards...

I can assure you that at least one guy (me) reads all documents and checks what's new after new releases. Even if not use all, it is good to know what is possible to do with it. But FAQ is necessary just to point people to it like "Read here, question #100500"    Helps very well.

You aren't alone

It's time to write a FAQ for questions asked on every page:

- what command option X means?
  RTFM, page 1, message 1. This is a copy of README.txt for the miner.

- how to mine using pool X?
  RTFM, page 1, message 1, examples section.

- why wrong temperature displayed?
  RTFM, page 1, message 1, KNOWN ISSUES section. Known issue of different ordering for mining and overdrive, thanks to AMD.

- can miner stop overheated GPU?
  Yes, -tstop option.

- why miner does not stop overheated GPU but exits?
  Probably, wrong -di option. See question above about wrong temperatures.

- why this command line doesn't work (escaping '&')?
  Char '&' in command line means command separator, to use it in command line either quote string with "", or escape '&' (use ^& on Windows).
  No need to do this in *pools.txt or config.txt.

- how to mine Decred or Sia ONLY with this Ethereum Dual miner?
  No way. It is Ethereum miner with extra bonus coins. To mine extra coins only use other miners.

- why Ethereum hashrate in Dual mode is higher than in Single mode?
  Hardware feature, accept it as an extra bonus.

- is 15.12 driver mandatory?
  Usually latest drivers work well. But there are some reports of people where they don't. So 15.12 is known good version.

- will newer drivers have higher/lower hashrate?
  Usually no, but it depends... Check for yourself.

- why miner does not show temperatures for RX 480 cards?
  They use newer overdrive API which is not yet published by AMD.

- why miner on Linux with stock card settings gives lower hashrate than on Windows?
  This probably is the difference in time calculations on both platforms. In reality the accepted hashrate is usually the same.

- why -cclock/-mclock options do not work?
  Sometimes they do not work. Use Afterburner or Trixx on Windows, atitweak and other tools on Linux instead.

- why my GPU is 10C hotter in Dual mode?
  This is a price for the extra work done. It also consumes more power, so make sure your PSU has sufficient power.

- can the temperature be lowered in Dual mode?
  Yes, -dcri option. Also see -tt and -ttdcr.

- how can I undervolt my cards on Linux?
  Usually only by flashing modified GPU BIOS. Unfortunately, no standard way of doing so.

Please add more by your taste.

Read this thread, page 1, message 1, "KNOWN ISSUES".

Usually either CDM detects invalid ETH share by itself and does not send it to pool (number of such shares is shown in console miner stats). Or if something was really bad then OpenCL call (to GPU processors) hangs, miner detects this and restarts itself (or reboots the system depending on -r parameter).

Thank you for hint, just switched to it to test.
Connection params are the same as for nanopool, just replace host name.

The answer is how long do you mine SIA on nanopool :-)

This was an issue with the pool that stopped/delayed responding to network requests. Miner probably had no explicit timeout check and decided that some execution thread was abnormally stopped. So it had to restart. On Windows some network errors detected much sooner (and as a side effect, Windows loses TCP connections instead of trying to wait a bit and recover). On Linux such connections still trying to recover for some time that exceed miner thread execution watchdog timeout. So there was no issue on Windows machines (miners just reconnected), but Linux miners aborted and restarted until pool recovered.

This definitely needs a fix because any network issue with the secondary coin pool should not abort mining the primary coin.

Yes, I had this on Linux too.

I was unable to send you FULL logs since I run diskless and before reboot my script saves only last 50 lines of log, and the rest in RAM was discarded on reboot. Anyway, I was sure that it is a network bug, so you are able to find it without full logs.

As I wrote, it was an issue with pool (confirmed by pool operators), and it is definitely the bug in miner. But it did not bother me a lot (just once), so I did not enable logs after that (it was not repeated either).

I am also able to test it, but I think it will be difficult to reproduce. So just add a timeout and hope it will be resolved.

@Claymore:

what does this mean in logs (I set -r 1 to reboot in case of issues):


This issue occured on both my Linux rigs at the same moment. It was repeated (rebooted) 3 times before continued mining. I also found that at the same time nanopool had hashrate drop, so it looks like some issue with the pool.

What concerns me, that an issue with the secondary coin pool resulted in reboot (restart) when it should be sufficient to restart some threads (they are not GPU/OpenCL related). Maybe makes sense to restart just miner if it is not a hardware/OpenCL issue, and reboot only when there are problems with GPU?

Could you please give some numbers?

For me ETH+SIA is more profitable at the moment than SIA only using other miner. Did not try, but assuming that SIA only gives doubled hash rate comparing to SIA hashrate in dual mode. Just because in ETH+SIA the 1st part gives more income than the 2nd. If prices change it may change too.

So what is your hash rate for SIA only (gominer), ETH+SIA (CDM), or you compared to ETH only profit (CDM)?