You have ZERO chance at RoI with anything at that high of an electric cost - even the 0.23/KWH is prohibitive.

 Mining is just not for where you're at.


 By the time "more efficient" gear shows up, the diff will have increased too much to make it worthwhile - do keep in mind you're trying to compete with large farms that commonly have electric in the under 7 cent/KWH and frequently in the under 3 cent/KWH range.

 If you mean "falling diff", that isn't going to happen. Ditto "flat diff".

 Don't forget you don't GET the thing 'till sometime in December.

 My standard calc shows this thing has no more chance at RoI with anything but VERY cheap electric than the 4.66 or 4.86 batches did - they did NOT drop the price enough to account for the lower TH and the LONG DELAY IN SHIPPING.



 The only advantage to this unit is that the higher volts/chip should let it overclock and be more reliable on meeting it's specs - at the cost of efficiency (I do NOT believe that 0.25w/GH spec, but had to use it for my calc as it was what was available).

 Oh, you believe in that alleged dude that supposedly changed water into wine?

 And I do agree that churches are not needed for that belief - they didn't exist to any real degree for a couple centuries or so after Christianity got it's start after all.

Mostly it looks like it will depend on how expen$ive the final 14/16nm generation is, and how many folks end up selling that generation to the general public.
 Diff increases SHOULD stableise out for a long time once that generation has been around for a while, as Bitcoin ASIC will have caught up to the "state of the art" at that point (I'm guessing somewhere around December 2016, give or take 3 months - but won't be shocked if I'm off a few months from that guess).

 At that point, it's actually practical to plan for achieving RoI on hardware in time measured in years, instead of the commonly less than a year and rarely year-and-a-half any miners up to now have had to be profitable (assuming AVERAGE US electric prices on that statement, we all know that cheap electric extends "profitable" lifetime on mining gear).

 Right now, you pretty much have to figure on achieving RoI by July 2016 on any gear you buy NOW, as the halfing will make almost anything currently available either nonprofitable or VERY VERY low profitable after that point - it's actually a bad time to be introducing the next generation.


 Last month, we got a bit lucky on diff increases (probably caused by Bitmain's S7 shipping issues for the most part) - this month is already going to be a lot higher than we've seen for most of a YEAR (downside of the current pricing spike).

 You think that the price, which is so high as to make it likely impossible to achieve RoI on these things at any electric cost greater than about 3 cents/KWH, is right in line?

 Dunno what you're smoking, but keep it away from me - it's making you halucinate.


 Original Blockc price was $1300, then they went to $1400, haven't bothered to look since then - it was way too expensive (worse so than any of the S7 varients) at $1300.


 I'm still in "wait and see what Lketc/BW.COM come up with for a price, and what the actual specs on their miner(s) end up being" mode, while Ocing my S5s and pushing my SP20 as hard as I can with the cooling weather to maximise income while Bitcoin prices stay high.


 I'm starting to think I might manage to RoI those miners ANd their PSes after all. 8-)

Never mind, I was thinking S1 for some reason.

 That doesn't sound right at all, that's very close to what my S5 units are generating (.0095 right now theoretical max) - the S2 is a LOT lower hashspeed than the S5.

Your limit on "free electricity" is probably going to determine the most you can mine.
At this point, you would want to go with Antminer S7 units as they are the most efficient.

On the other hand, if you have enough "free electric" and not enough cash to buy enough miners to max out your power, you might want to look into older miners where you can buy more TH per $ because they're less efficient.


 There are too many variables for us to give "One True Answer", other than "It Depends...".

Looks like it's back into a slow sink, each roller-coaster peak and valley gets lower.

 Dunno if it will get back down to $250, but looks like we won't be seeing $500 this year.


 I do wish I'd waited a day or so to cash out what I had on hand, made a fair bit but it could have been more....

 Check the JohnnyGuru review, they usually check to see which connectors tie to which rails, then make sure each of the hash boards in on a seperate rail and see if you can tie the controller board connector to the 4'th rail - though at 50 amps a rail you should be able to get away with running the controller from one of teh same rails that one of your hash boards it tied to if the 4'th rail is CPU/Peripheral specific or some such.

 For that kind of money though, I'd spring a little extra for a Seasonic 1200 platinum, or drop down to the Seasonic X-1250 or EVGA 1300 G2 gold units. Did the math a while back, and the extra cost of a platinum PS takes many YEARS to pay off unless your electric is expensive (and in that case you'll probably be losing money on mining anyway).

 Still have the issue with the limit of the PS on whatever +12v bus powers the "peripheral" connectors, and on an S7 you ARE talking about
 (1) Appx 400 watts or well over 30 AMPS per hash board, works out to about 13 AMPS per PCI-E if you're using all 3 of them per Bitmain specs.
 (2) limit of DO NOT POWER A SPECIFIC HASH BOARD FROM MORE THAN ONE PS.

 The issue isn't JUST the molex connector.
 In fact, it's more about limits of the Power Supply in this sort of case, though pulling over 6 amps through a peripheral-type Molex is marginal for reliability.
 It's not smart to assume the current draw will work out exactly equal, parallelling connectors tends to have a little imbalance on the contact resistances and makes the power draw somewhat uneven and makes those adapters even more marginal (to be fair, the imbalance would probably be on the order of 7 vs. 6 amps per connector *IF* both are making a good solid connection - but if one is NOT making a good solid connection, FRIED connector and probably fried S7 hashboard).


 You have to derate the connectors used in a PCI-E connector by quite a bit.
 Molex only specs them for 8 AMPs each max = 24 amps per connector = 288 watts in a PCI-E 6-pin connector, due to heat retention by the connector itself.
 Even though the pins THEMSELVES are rated for 13 amps max that's a FREE AIR rating.

 (Yes, I read ALL of those specs a while back - it's where Spondoolies came up with it's 288 watt max limit per connector in the SP20 web interface).

UPS on a miner is a waste of money, if you have reliable power.

 Electric being effectively "free" during the winter doesn't require you to have electric heat - miner heat generation cuts gas bills for heating quite well too.
 There is a limit though, if you have enough miner heat output to exceed how much heating you actually NEED.

Reality seems to have set in for a bit, dropped back under $400 (hit $370ish briefly on Coinbase) before it rebounded some.

 Gotta wonder when the next reality check and how big it will be.

Just buy the BTC, it's faster.

The other problem with the 100 minimum is that most folks will end up having to buy from a reseller - which will make the already TOO HIGH price WAY TOO HIGH.

 I'm also still very concerned about the longevity, given the serious issues reported by the only ones anyone has reported about "in the wild" to date.

For some of the older lower-power miners that molex-to-PCI trick would be tolerable. I ran one of my Gridseed blades on 2 molex per side for a while, but that was less than 4 amps or so per side so only 2 per molex, which is well within their rating and the capability of the wiring.
It would be practical to run a 6-pin PCI-E power connector that is running within the PCI-E specs from 2 molex, but most miners are pushing the connectors to closer to the limits of the CONNECTOR, which is way outside what most 4-pin molex peripheral connectors AND THEIR WIREING are designed to handle.

That doube-4-pin Molex to PCI-E adapter will almost definitely NOT handle the load a S7 would put on it - and it's doubtfull that your power supply puts enough current out on the "peripheral bus" +12V supply anyway.

 The wiring is WAY too light, will probably melt and short out.



 Garage should be fine, if it doesn't leak and if it's secure enough you don't have to worry about someone stealing your miner(s).
 You're also more likely to have at least one existing 220V outlet there to power an S7 with from a GOOD power supply.
 Many moons ago, I had a setup in a slightly modified storage shed (I added some airflow access, a breaker panel and outlets) running a bunch of computers - ran air in under the eves on the South side and out the East due to prevailing winds, worked fine and stayed dry and close to ambient temps. Solid metal walls, solid door, and a good lock in a backyard with a solid fence around it made the setup hard to notice, and the comps I had in it were older ones that weren't worth a lot anyway.


 Bitmain's site was up a couple minutes ago, they might have been doing a "price adjustment" and had it down for a short while earlier.


 *IF* the current Bitcoin price holds up, an S5 should barely be profitable after the halfing for a while, and an S7 somewhat but not greatly so for months, even with the diff increase rate I forsee between now and then - but neither is going to make a LOT after the halfing even at the current $400ish price, you had best plan to hit RoI on either by end of July IMO even if the Bitcoin price stays high.

 If the current Bitcoin price drops back down into the mid-$200s soon, the situation stays the same except RoI is a ton less likely and the S5 will probably NOT remain profitable at all after the halfing (the S7 will still be, but BARELY).

It hasn't been possible to break even on Bitcoin (or any other SHA256 coin) mining with GPUs for a few years now.

 It hasn't been possible to break even on Litecoin (or any other Scrypt coin) mining with GPUs for about 2 years now.

 It is possible to make a profit on mining other altcoins like Dash with GPUs, but you better have some fairly cheap electric and a well-optimised rig.

 If you're using the thing as a space heater (essentially free), Dash can make you a bit if you run it on Nvidea Maxwell-based hardware (like your 750 Ti), and a very small bit on recent AMD hardware. You might even be able to achieve RoI on the cards given essentially free electric, but that is iffy.

 Don't bother with trying to mine on CPUs at all. GPUs are tens or hundreds of times more efficient to mine with, if you can mine with the GPU at all.


https://bitcoinwisdom.com/bitcoin/calculator can give you some idea of the real numbers.




 Badly outdated info, CURRENTLY the Maxwell Nvidia cards are the BEST to mine with on most altcoins, the pre-Maxwell ones aren't as good as the Maxwell ones but are sometimes competative with AMD cards depending on the specific algorythm the coin uses and how recent of a version of the correct miners you use.

The real problem, just like they ran into on the GC3355, is that the SHA256 side is not all that impressive, and will be unprofitable far too soon to be able to RoI the machine - and the Scrypt side is reliant on a coin that seems to be losing any chance at staying power, making it really hard to figure if you will even RoI the scrypt side or not.

 The whole "dual miner" concept should never have been revisited, you end up with a machine that's halfway obsolescent far too soon for way too high a premium.

SHA256 or Scrypt coins with GPUs

 There, I fixed your statement.

 9-)