Try unplugging the bad module.

 Actually, Litecoin price kicked UP quite a bit when it halved, and has been pretty stable - and the hashrate has been pretty stable - since then.

 It's was a better situation for folks to have a good chance at RoI and even overall significant PROFIT on their gear than Bitcoin has since about mid-summer.


 This might change when the A4 is released, but depends on how many of those Innosilicon gets built and how long it takes them. A2 units will still be profitable even if the Litecoin hash rate doubles and the price drops back down to $3 (it's been above $3 since a month and change before the halfing) even if you don't have VERY VERY cheap electric.

 I actually wish I'd bought some A2s instead of the S5s and SP20 I bought last summer - they'd be a lot closer to being paid off, and I am having serious doubts about the S5's ever managing RoI at this point DESPITE the recent Bitcoin price runup (diff increases have negated most of that, and I was figuring on holding my Bitcoin 'till the halfing runup anyway).

 The 110 units I have work very well on VARDIF on liteguardian - the 88 for some reason doesn't (TONS of rejects) but it works well on their fixed 1500 diff ASIC server.
 I suspect the 88 needs updated software, but as long as it's working I'm not inclined to mess with it.


 The stock PSUs in these units ARE known to be very marginal - I highly recommend replaceing them with Seasonic X1250 (Gold) units (the EVGA 1300G2 will work but it does NOT fit anywhere nearly as well).



 P.S. - Scrypt got a hair MORE profitable between early 2015 before the halfing runup, and afterwards - the price kicked up from the $1something range (as I recall) to over $3, more than making up for the block reward halfing. It's doing MUCH better than Bitcoin has over the last year.

I recently had to do a replacement on an A2 Mega power supply (Original was DOA on reciept from Zoomhash).

 I had an EVGA 1300-G2 on hand, tried that second (see below for first) - it fit sorta, but the PCI-E cables had NO clearance vs. the control cables, and the fan didn't line up very well at all with the care fan slots (it's offset about an inch on center, though the quarter inch open space between the PS and the case kinda made it work anyway).
 It worked and was a bit more efficient than the "unknown black PS" in the other Megas I have, but I didn't like the tight bends imposed on the power cables by the lack of clearance, and the fan not lining up with the intake slots on the case was a potential cooling issue.

 Swapped that PS out on my Antminers with one of my Seasonic X1250s I was using on them.

 The Seasonic is

 (1) a good quarter inch shorter than the EVGA
 (2) has PCI-E cable spots that FIT the empty spaces around the control board MUCH better
 (3) Showed efficiency that was identical, give or take measurement limitations, to the EVGA at 1000, 1100, and 1200 Mhz clock rates.


 Both of those PS are quite a bit less $$$ most of the time than the Corsair platinum, but have almost the same efficiency (they bloody near meet 80Plus PLATINUM efficiency specs in all the test reviews I've seen on them), but given the fit issues of the EVGA I have to recommend the Seasonic as a much better choice for an A2 MEGA Terminator replacement PS.

 1000 MH  -  93 MH/s   -  1040 watts
 1100 MH  - 102 MH/s  -  1150 watts
 1200 MH  - 113 MH/s  -  1260 watts

 All power measurments by a Brand 1850 power meter (which seems to read a bit high, though within it's 2% spec).


 And yes, I've noticed quite a range between my 4 MEGA units on power draw at the same clock - and some on hashrate - this unit seems to by my LEAST efficient unit.
 I had swapped the 1300 out of one of the others at one point for a quick test before I tried the EVGA, this unit drew almost 100 watts MORE than the unit the 1300 came from at 1000 MH clock but hashed 2MH/s more too.
 Seems in part to be an issue with how many WORKING cores each unit has, at a guess, more than anything else, but since the UI doesn't show the number of working cores on those units (one thing I wish they had carried over from the 88 non-MEGA firmware) it's hard to tell for sure.

Zoomhash has them at $800, BUT watch them like a hawk if you order from them.

 I ordered 5 of them a couple weeks back:

 I could not place the order as one order, as their site would not accept my debit card for 5 in one order despite my having plenty of money in the account.
 Had to place the order as 5 seperate 1 unit orders.
 This might be an issue with my card provider, though - and it did let me use their FREESHIPA2 code on each unit.

 Delivery took a week, which was reasonable for ground shipping.

 Of the 5 110MH MEGA units I ordered, they shipped me 4 as advertised and one was an 88MH non-MEGA unit.
 Upon reporting this to them in an email, Zoomhash_Michael promptly accepted my suggestion to refund me $160 as compensation for the wrong unit.
 (This prompt response to one issue was a GOOD thing.)

 ****************************************** BUT ****************************************************************

 In the SAME email, I reported that the power supply in one of the units (a 1000 watt Enhance PS that does NOT meet their stated "ships with an 1100 watt gold power supply") had DIED the first time I turned it on. I am currently running that unit on a spare Seasonic X1250 PS.
 I have recieved NOTHING back on my report about the dead power supply - no refund, no "return it for replacement" instructions, NOTHING.


 The fact of the power supply being the WRONG unit, and being DOA, and no response to my emails on the subject, coupled with previous issues with Zoomhash, make me disinclined to deal with them again even when they have a good deal going.

Won't happen. Gates isn't interested enough, Buffet thinks it's a bad investment, and pretty much anyone else that COULD do so has no interest.


 On the other hand, if someone DID start doing so, it WOULD be noticed and would push the price a fair bit higher over time as it got more and more concentrated.

It's "moron", not "manan".

Probability pretty much zero.

 The start of the push for block haling will happen a month or two BEFORE the actual halfing (AGAIN), not as an "instant" reaction to the halfing - and the halfing only cuts the reward for mining in half, it does NOT affect the LARGE EXISTING Bitcoin supply.


 Back when the $1200ish peak happened, the supply of Bitcoin was a LOT smaller. Bigger markets tend to move slower as it takes more money TO move them.

False.

 Bitcoin was in the $220-$230 range for quite a while this past summer, LONG after the $1200ish peak.

 There is NO inherent reason NOR DOES THE HISTORY SUPPORT an ATH to serve as a bottom.

EU writes laws regulating Bitcoin (by intent, or by "badly written law that ends up being TOO restrictive" like has happened so often in the past by well-intentioned lawmakers) to the point it's almost unusable.

 Other areas do the same - or outright ban Bitcoin (but I don't see a deliberate outright ban being likely, even China seems to have realised that THEY are making a ton off of it given how many Bitcoin manufacturers AND FARMS are China based).


 Internet "bubble" syndrome.

I suspect Bitcoin might touch $500 again this month - but eventually the current push will peter out as it's got NO real news to support it, and we'll drift back down towards $300 or so next year, before the halfing price runup kicks the price back up quite a bit again.


 There is nothing "rational" about $400 at this point. Only reason I don't see it getting back down into the low-to-mid $200s again is the massive amount of "new money" that came into the ecosystem when the Chinese exchanges started being able to accept fiat currency again....

 Sfards had that SF100 unit, but it appears to have been a very poor board-level design AS USUAL from the successor to Gridseed.

 Innosilicon A4 - sometime next year likely first half.
 Alcheminer is also working on a next-gen Scrypt chip, but doesn't seem nearly as well funded.


 
 I've been saying all along that probability of RoI on the S7 was pretty much zero unless your electric was VERY VERY cheap.
 Bad timing on the release and WAY too high of a price.

 Avalon 6, even worse - lower performance for MORE money.

 B-Eleven - going to have to be VERY cheap to have a prayer.

 There is hope on the horizon though.....

 Why should you be compensated for living in an area with high electric rates?
 That is YOUR choice.

 Nobody ever guarenteed that mining would always be profitable for everyone.

As a rough estimate, I'd say that probably 30-35% of the current hashrate is produced by gear in the 0.2-0.3 w/GH range.
I'd also guess that a small amount is produced by gear in the under-0.1 range (KnC small amount, Bitfury probably more) but doubt that's a HUGE number yet since neither Bitfury nor KnC have been showing hugh hashrate increases over the last few months.

I doubt there was a TON of old gear that had been turned off due to not being profitable last summer that is now turned on. 1-2% MAYBE.

 I suspect there is more gear that was "turned off for summer" or "undervolted/underclocked for summer" but still not a ton of hashrate. 2-3% perhaps.

 

Gee, price goes up hashpower follows it.

 WHAT an amazing conclusion!


 (IF your sarcasm meter didn't break from that commentary, it must already be broken.)

Other than the Chinese "we can now take fiat currency again" announcements that caused the initial spike, I'm still trying to figure out why prices have climbed.
There has been ZERO positive news for Bitcoin lately, and still getting those rumbles out of the EU about "planning tighter regulation for anti-terrorist reasons" crap.


 Only thing I can think of (other than crazy Chinese investor speculation, or a BIG pump-and-dump) is that the Chinese economy might be tanking more than their government censorship is letting anyone know about.

Innosilicon has ANNOUNCED full custom 14nm for their A4 Scrypt chip.
They haven't said one way or the other on the A3 - but I don't see ANY reason for it to be anything BUT full-custom as SHA256 is a simpler algorythm easier to design for.

 Note that they have indicated that they intend to bring the A3 and the A4 to market at the same time.

 Also note that their "search for investors" for the A4 seems to be more about risk management than anything - they're NOT cash-short after all the A1/A2 sales they've made, and they brought both of THOSE chips to market with no "pre-orders" or outside investors.


 Given the next diff increase is looking likely to just miss 20%, and that price has NOT been keeping pace since the initial surge (we're still not back up to that first peak despite the recent second surge, though might get there a lot sooner than I think)....



 Up side - price keeps going up even close to diff increases, I might actually achieve RoI on my S5s and SP20 after all. 8-)

Or they already took their losses, or they bought them as a LONG term investment, etc....

 The SP20 has internal limits at 288 watts/PCI-E connector, which is the CONNECTOR max spec (the contacts are rated higher but only as a free-air max rating).
 That works out to a bit over 300 watts per connector at the wall with a Platinum or Gold PS.

 *4* connectors on an SP20, 1200 watts = 300 watts/connector NOT 400. Your math was wrong.


 The PCI-E spec says 75 watts for a 6-pin, 150watts for an 8-pin on the connectors, but that's very very conservative especially when the only thing the 8-pin connector adds is 2 more ground wires (IMO they should have added 1 each ground and +12, 5 grounds 3 +12 makes ZERO sense).


 "at the wall" watts depends quite a bit on the efficiency of your power supply.
 

 After the next generation - though for BitFury and KnC it appeares they are already building that generation, with Innosilicon looking likely to join them in a few months.

 When full-custom 14/16nm becomes widespread, we'll be "caught up" on current tech and mining innovation should slow down lots - and difficulty should flatten out into a long "slow rise" period.

 At this point I'm not betting on diffuculty stabilizing in 2 months. Or 3. Perhaps in 4.

 I don't see any chance of the current generation (S7. Avalon 6, SP50) achieving a positive RoI unless you have VERY VERY cheap electric - and I'm starting to wonder if you'll be able to RoI those machines on 3cent/KWH electric.


 The rumblings are underfoot, A3 is comming and it looks like it's going to be the REAL game-changer for folks not running sub-20NM BitFury or KnC farms....