The alternative to all those small payouts, which is what causes you to require a larger fee to move your BTC, is that Kano keeps a hot wallet on the pool side.  A hot wallet makes a hacking target and opens everyone up to the possibility of a large loss.  Kano has said he doesn't want to keep a hot wallet so that's why we get the small payouts which in turn means a larger fee for us to spend it.  What I do is keep a mining wallet and when it hits about 1BTC I move that to another wallet where I keep most of my BTC.  So when I spend, I spend from my larger wallet so I can use a small fee and still get a quick confirmation on the transaction.

One thing that might be useful (if we got more blocks on a regular basis) is a consolidated payout.  It seems like we often get paid for 3 or 4 blocks at a time, so getting that all paid in one lump sum rather than 3 or 4 separate payouts all at the same time, would help a little.  However, the trade off there is that it would become more difficult for the miners to keep track of getting paid for every block.  Maybe if Kano put a txt note on the transaction that included block numbers?  

EDIT:  I see others made the same point while I was typing.

And the gui for the miner looks like all 3 boards are still hashing?  Based on the actual hashrate it looks like one of the boards stops working.  If the miner gui says everything is working normally and shows a hashrate for all 3 boards then I would be inclined to suspect it was a controller problem and not a hashboard problem.

I have a miner with a problematic board (an earlier batch, like 3 or something) and I have to reboot it many times to get a decent hashrate from the problem hashboard.  If the reboot process took a lot longer, that sounds like a pain in the ass.  Unless this process actually guaranteed a good boot from that board.  I am skeptical though.  

Supposedly new firmwares are going to be available on the website for download soon so I would guess they will be releasing this kind of thing to buyers of earlier batches.

I emailed them about it a few days ago and was told this:


@jstew  I know you we asking for more recent but I have these:

6-24 version of the 600 freq firmware 

low voltage version of 6-22 firmware for 650 freq

They may not be selling them for this reason, too expensive for them to make to sell at a price that would allow a reasonable ROI.  At this point the BM1387 is the highest efficiency consumer available chip there is and not likely to see an improvement for a while.  Caanan may have a competing chip in the works that releases soonish but that remains to be seen.  It is unlikely that bitmain will have a more efficient chip in less than another year, but more likely it will be 2 or possibly as many as 3 years before something smaller than 16nm hits the consumer market.  So the difficulty now and for the foreseeable future is dictated by how many chips bitmain puts out there.  Unless Caanan really does have a 16nm chip in the works it is in bitmain's interest to not flood the market too fast.  If they sell at a pace that doesn't drive difficulty up too fast, they can continue to sell for some time.  If they sell as many as they can make, they will drive the difficulty through the roof and then nobody will buy S9's to say nothing of the R4.  I wouldn't be too shocked if the R4 appeared once more but I expect they have already ended any serious production on this one.

If Caanan releases something then a price war may ensue, but Caanan has always had fairly low production runs so all indications are that bitmain could produce many more units much faster.  It would be in the interest of Caanan to not try and undersell bitmain and not create a price war.  I think Caanan would still sell whatever they could make just based on the reliability issues with the latest antminers, even at a price point that was slightly higher than bitmain's price per Th.

The only other chip producers that I am aware of that can compete with the BM1387 are bitfury and BW and neither of them are selling to the consumer market at all.  I would guess that their chips are either not as efficient and so they couldn't compete on price, or their manufacturing ability is just too small, in which case again, they can't compete on price with bitmain.

EDIT:  When I wrote the above I hadn't seen this: https://twitter.com/BitfuryGeorge/status/786391526632935424 so maybe there is competition on the way in which case you really, really don't want a miner that can't ROI for a very long time if ever.  Bitfury says in there they are shooting for 30% of network hashrate on their chips.

My guess is they did pull a fast one on you intentionally.  I am not sure what returns on aliexpress look like so maybe you have some recourse there, but since you did not buy it direct from bitmaintech I would say there is no warranty you can rely on (plus the bitmaintech warranty is questionable anyway).  Bitmainwarranty based out of Denver (not to be confused with bitmaintech based out of China) could probably tell you if that dent is likely to be the issue and could also possibly repair the board (depending on the actual problem with it) for a fee.  So yes bitmainwarranty, despite having the word warranty in the name, is a fee based repair business.  But on the bright side they operate with a high degree of professionalism, they are timely with repairs and communications, and they are based in the US, so for US miners they are a very good option for stuff that isn't covered by the actual bitmaintech warranty.

As long as it is reliable it really doesn't need much bandwidth at all.  I have run 3-5 miners all on a 1.5MB connection which was also shared with a 20 or so PC's.  The bandwidth for mining is almost nothing but the latency is very important.

Honestly it sounds to me like it could also be a PSU overheat issue.  Usually if something works fine for a period of time and then stops I think overheat.  Do you have another PSU to try? 

I am thinking they already swapped it and sold the S7's.  They just didn't fully stock and fire up the S9's so they wouldn't explode the difficulty.  There appears to be a very deliberate and steady diff increase going on these days.  Someone fires up a ton of gear towards the end of each diff jump and then lays off mid cycle before doing it again to produce a jump each time.  This next one looks to be a good size jump but whoever is doing it seems to have more and more hashrate with each diff step, which makes me think it must be bitmain.  There are a couple other miner manufacturers but they all seem more limited in how much gear they can produce (apparently not enough to sell any anyway).

I wonder if the extended unavailability on the website means a board redesign is in the works.  Judging by the network hashrate though I would bet large orders are still going out the door.

I have been lucky so far.  Other than one problematic, low hashing board, both of mine have been doing fine.  If I was to predict a board failure it would be that one board since it sometimes shows random X's that come and go if I try and run the unit at stock freq.  That one is a batch 3.  The other one I have is a batch 4 and it seems solid so far.  In fact with the cooling temps of autumn, I OC'd it to 606 and it has been doing fine with a solid 12.98 Th.  I was using a script to change freq on the B4 twice a day for the past few months without a problem.  I even had a utility power failure a week or so ago and everything came up fine after the power was restored.  But other than that power failure I haven't power cycled them more than once or twice each.

I did have one weird instance where both of them dropped two boards each.  One showed all boards fine but with two of them with zero hashrate, the other showed two boards all X's and zero hashrate.  But a simple cgminer restart worked, no power cycle was necessary.  Since they both puked at the same time I am inclined to believe it was either a power issue or a network issue, but a power problem seems less likely since none of my high-end UPS's in the same room logged anything weird at the time.  The miners are not actually on the UPS's since their power draw is a bit much and I need the UPS's for servers and the like.

I want to upgrade my last remaining S7 but I am holding out for something else at this point.  This S7 is one of the original 54 chip board ones and it OC's like a champ.  I have been getting over 5 Th from it from almost the very beginning (although the error rate has crept up over the months).  So I am not eager to let it go for something that seems likely to have a failure.  The S7 pricing and then the S9 release so soon after caused me to almost not get 100% ROI on the S7's before I started replacing them, but this one is making my S7 profit slowly grow beyond just a break even.

If bitmain came out with a new miner, even using the same chips, I would consider it after some time if the failure rate looked ok.  But right now I am thinking whatever Canaan comes up with is the one to wait for.

I do have to say though, bitmainwarranty in Denver has really increased my confidence level in bitmain overall lately.  Those guys have been doing what bitmain in China should have been doing all along in terms of responsiveness and customer service.  Even so, the S9 reliability looks too poor for me to get another right now.

And if luckyant has brought that much weight to the party where did all the other hashrate go since we are still down overall?

Cool.  Thanks.  Appreciate it.

I use local proxy and then mine on local proxy with genoil and all of my miners were down as well.  Thankfully back up now though.

Do those work for remote access screen resolution issues?  I notice that I have screen resolution issues when I remotely access the miner if I don't have a monitor plugged into at least one graphics card, and that is true for both RDP and teamviewer.  Windows wants to default to something like 800x600 when it thinks it can't detect a monitor - also have display issues with MSI afterburner in remote access sessions which is what I think you are referring to.  I can sometimes use DVI plugs for that purpose but they don't always work (depends on the card I guess).  

Also where do you get those?  I see some on amazon but they are pricey.

If I am not mistaken it looks like what you got there was the beaglebone and you need the I/O board to go with it.  

I thought I saw somewhere that someone had said the newer S9's did not use the beaglebone any longer.

With BTC dropping I imagine we will see a new batch soon but I'll be damned if I am getting any more of these with the reliability issues.  I'm glad I haven't seen a serious problem yet but still months and months to go before break even.

Mine looks nearly identical except the freq is lower.  Top board has the highest hashrate, middle board is close to it but slightly less, and bottom board (which is also chain 4) is always atleast 100 GH/s less than the top board.  And that's what I get when I get a good boot which sometimes takes several reboots to achieve.

I have this exact problem with a batch 3.   I installed the low voltage firmware which seemed to help slightly (fewer restarts needed to get the 3rd board to hash normal speed) and I also downclocked which also helps in the same way.  But I still sometimes need 4 or 5 restarts to get it working right and even then that low hash board always hashes less than the other two.  This miner has been this way from the day I got it.  I have spent 20+ restarts trying to get it to hash normal at full freq before but I don't even try anymore.

Meh.  Glad to see the option but too expensive.  Three-hundred extra for a 180 day warranty is kind of obnoxious.  I hope that isn't an indication of what they think about the reliability of their own product.

These guys have some options: http://www.signalandpower.com/Stocked-Europe-Power-Cords/?gclid=CMGTsabPks8CFYuVfgodptwAPw

EDIT:  Oh I think I see your problem.  The 2mm is referring to the width so something like the European equivalent to American wire gauge.

EDIT2: http://www.cablestogo.com/product/03138/2.5m-14-awg-european-power-cord-cee7-7-to-iec320c13