Litecoin has NOT been mineable at anything like a profit for years now - I forget exactly when the first Gridseed GC3355 "orb" came out, but it quickly sold enough to put GPUs out of the business forever. I *think* that was about 4 years back, but I could be off some.

 Your only VIABLE choices are ASIC - if you dig around, you can probably find some older-generation gear like the Innosilicon A2 available for fairly cheap, and it's QUITE profitable right now if you have low power costs.
 Ditto the Alcheminer, if you can find one or any of it's "clones" like the MAT units.
 Ditto the KNC Titan cubes (most of which are probably down to the cores that will last along time, those things had HORRIBLE "core dies, cube down to only 75% of it's capability" type issues the first 2 years they were being sold.
 
 Unfortunately, Innosilicon has declined to bring the pricing on it's A4 unit down to even close to a competative price to the Bitmain L3+, so I can't recommend those - a little over half the hashrate for a little MORE power usage and a HIGHER price just doesn't make sense. If they bring the A4 down to more like $800 it WOULD start making sense.


 Availability on the L3+, like on ANY current-generation mining gear, has been spotty due to limited foundry capasity available to mining chip makers.
 I believe the next batch of the L3+ is slated to be available sometime in JULY - though you will find some units available now from third parties at GOUGE pricing.

Rate where I was at in Iowa was complicated - the base rate was tiered in the winter but not in the summer, I had Time of Day rate, and there were a few "surcharges" that TOD rate SHOULD have applied to but did not.

 End result, 9 months out of the year at my usage level the last several months I was there ended up being a bit less than 8c/kwh, but the 3 months of the "summer" starting in mid-June going to mid-September it ended up being over 14c/kwh.

 My current place I'm on a "fixed" payment for electric I negotiated with my landlord as part of the moving process.
 Immediately previous place (same landlord) I was paying ballpark 4.5 (and will be paying a hair higher than I was in the new place, as the rate went up a couple percent during the time I've been in the current place).

 

Never seen that "tier" thing before.

I'm going by the fact that I've NEVER had any sort of a "sleeve bearing" fan, no matter HOW fancy the name or how long the warrenttee on the PS was, last more than 2 years and usually they don't make it to 1 year before they die.

 Fluid Dynamic is a "sleeve bearing" fancy name.

 *ALL* of the fancy-name varients on a sleeve bearing claim to be better (the ones that use a "rifle" type slot IN the bearing do a better lube job than standard non-slotted types and a little better than a "straight slot" type, so technically they ARE better), but the seals STILL leak and the lube STILL runs out and the fan STILL dies way too soon.


 I'm not sure on the Teflon thing though - Teflon is VERY slick (though it's not THE slickest substance known to science any more, I think it's still #3 by a fairly narrow margin) and pretty bloody tough, so if they're using a Teflon coating to lubricate the sleeve bearing that MIGHT manage to last a long time - and it's definitely not going to leak, the only question there is how long it will take to wear down the Teflon.

 Backblaze publishes specs on reliability for all of the drives they use in their production "vaults" - ballpark 40 THOUSAND of them as of the last posted report.

 Seagate models have consistantly been a lot lower reliability on almost every listed model (and lower but not by a LOT on the rest) than WD, which tend to be lower reliability than HGST (especially the HGST enterprise models) - but Backblaze doesn't use EVERY model that have been made by those makers.
 They do tend to indicate trends though.
 They also have some Toshiba drives, but I forget where those fell at, and the number was kinda low - I think they were close to the WD numbers but the total drive count was low enough to not be definitive.


 RAID on a BURST mining setup doesn't make sense.
 It's not "critical irreplaceable data" where losing it would cost you a ton - if the drive goes down you just replace it and replot the new one, not a big deal.



 I hope that Ryzen system isn't for a pure mining system - if it is, that's MASSIVE overkill.
 Ignore this comment if it's primarily a gaming system that mines when you're not actively playing on it.

My Gigabyte full-length 1070s have a TDP of 180 watts - which is the highest I've seen on a 1070, most seem to be at the "FE" level 155 watts TDP.

 A EVGA G2 1000 in a rig with a low-TDP CPU should handle 5 of them, and if you lower the TDP to under 160 watts (like most ZEC miners do) or use cards with a lower starting TDP (EVGA 1070 SC as an example) it should do so comfortably.

 1. 30% is a lot closer on traditional A/C - SEER 10 would work out to about 33% if you have ZERO ventilation at all - but most places you can manage to get to 80 F for a LOT of the year with no more than good ventilation.
     
  To quote from one of the articles I've read about Yahoo's Chicken Coop design, though:

 "Yahoo engineers determined that YCC yielded roughly two percent annualized “cost to cool” with evaporative cooling; meaning that free-cooling would have an even lower percentage. (cost to cool is the energy (kW) expended to remove the heat generated by the data center load as a percentage of the data center load itself.)"

 Their setup uses standard rack mount cases, though, but it uses the fans IN THE CASES to do all of the air moving - I suspect their actual total COUNTING those fans would be under 10%.

 They use a "flash evap" misting type setup, not traditional "aspen" or "rigid" media style evap cooling, which also would keep their costs down - and this IS in a building custom-designed and built for this work.


 3. no doubt rates will rise - but with the AMOUNT of hydropower available, and the ability to bump "for sale" rates up even more, local rates shouldn't go up very fast.
     This year's increase in the county I'm in was 2.1% or LESS for most classes, though the lower level of "Industrial" got tagged for a 5% jump - but the higher level of industrial got tagged for less than 1%.
     
     Most folks in the Electric industry still use a "one household averages 10KW" estimate - you might think that should have gone up over time, but it gets dropped due to a lot more folks being "single" and therefore the average household size has been dropping for a long time, which evens that estimate out.
     Rough ballpark 50k pop each in Chelan and Grant counties = 500 Megawatt local usage each (ballpark), probably add that much more at MOST for local business usage (not a lot of manufacturing in this area so that estimate is likely VERY high) - and you're up to about HALF the available Hydropower in each county from it's dams on what should be a rather pessimistic estimate of local usage.
 Douglas ends up at pretty close to the same, since it's pop is a LOT lower.
 Area pop also isn't growing very fast, nor is there any real reason for it to do so.
 
     I've seen a very FEW other areas that can get to 5c/kwh on industrial rates, but haven't really looked all THAT hard as I'm already in the "optimal area".

Asrock 970X Extreme 4 on socket AM3+
Asrock FM2A88X Extreme 6+ on socket FM2+ on my usual "mixed rigs with an APU"

 There are a LOT of similar options for Intel, and several others for AMD, but this pair tend to be the lowest cost or VERY close to it, and they *work*.

 4, you're looking at a very limited selection of Intel boards - the one that sticks in my mind is a Biostar "Racer" model with 4 x PCI-E 3.0 16x slots (but not all of them are electrically 16x, which isn't a significant issue for mining on).

 The MSI MB you linked to WILL support 4 dual cards - you just need to run it as an open rig, or use a case like the Thermaltake Vista V34 that has specific "8 peripheral slots" support.
 Cards do NOT have to overlap the motherboard if the case makes provision for that (most cases don't).

 Just don't try to run more than 1 of the Gigabyte AORUS cards on a setup like that, they won't physically fit (those cards are at least 2.5 slots wide and some apparently are 3).

 The MAJOR down side on 4 card rigs is cooling gets a lot tougher to make work. I doubt you will be able to keep a 4-card non-riser rig full of 1080 ti cards cool - it's hard enough by report to keep 4 card rigs like that with RX 470 or 480 cards cool unless you undervolt a lot and turn the TDP limit down, and those are only 150 watt OR LESS cards to start with.

 I end up using one of the Gigabyte "ITX" very short cards in slot 3 of my *3* card rigs in order to keep the middle card cool (the middle card gets ONE fan completely unobstructed and the other fan(s) almost completely blocked, vs the #1 card having all fans with a narrow space for airflow and therefore partly obstructed), running 3 x 1070s.


 Mining is NOT bandwidth-intensive to any significant degree, or riser rigs with cards running from 1x slots would show VERY bad performance on those cards.

 4 card rigs are a lot harder to keep the GPUs cool, unless you use risers.

 I personally prefer 3 card rigs, but current motherboards are either WAY too expen$ive to get proper 3 cards spacing for good cooling on 3 FULL LENGTH boards, or I can use the low-cost stuff and make a point of putting a "short" card in the 3'd slot (like the Gigabyte "ITX" 1070 model) to allow the middle card to have ONE fan completely clear to keep it cool enough.

 3 card rigs are a LOT easier to set up then riser rigs, in my experience tend to be noticeably more reliable than riser rigs (riser rigs ARE easier to keep the GPUs cool on though), and the cost difference is very small to nothing after you factor in the cost of the risers, the extra amount of connections needed on your PS (or the extra CABLES to add to it) to power the risers, and the problem that power supply cost per watt starts climbing noticeably faster past about the 650 watt level vs the "additional costs" of 2 MB, 2 CPU, twice as much RAM, 2 HD/SSD/USB keys.

 You also get the nice part that if you DO have a rig go down, you only lost HALF of your hashrate from 6 cards instead of all of it.

 If I could get reliable access to that Katana-model 1070 or if someone that DOES normally sell to the US builds a similar model at a SMALL premium to 2-slot card designs, I might start going 4-card rigs regularly, as THOSE things you can keep cool fairly easily on boards like the Biostar Racer board that Phil liked that have 4 slots properly spaced to mount double-slot cards ON the motherboard.

 NVidia and AMD make "compute" single-slot-wide versions of many of their higher-end GPUs but those things are CRAZY expensive definitely not a good choice for a mining machine (unless you can get a good used one for a LOT off new pricing).


 The really scary part - as of this morning, the only RX 4xx boards I could find for sale anywhere (I do NOT look at eBay, too many bad experiences there) were all priced HIGHER than the low-end of the GTX 1070 - and the RX 5xx cards I could find weren't much better.
 We might see a big short-term move of folks to NVidia because the "lower cost" part of the ROI question on the RX 470/480/570/580 isn't THERE any more vs the 1070.

nvOS if you're running Nvidia cards is free, but he only supports a few specific motherboards.

It probably wouldn't be hard to modify it (IF needed) to work with other motherboards (LINUX is very flexable that way).

 Or go with a standard distribution like Ubuntu or XUbuntu etc.


 simplemining is a ripoff IMO, paying a monthly fee for an OS is a very sickening idea to me.

I don't class the EVGA B3/G3/P3 series as worth even looking at.

 Their B2/G2/P2 series was quite good, but the newer series went to a JUNK SLEEVE BEARING DESIGN fan that is NOT going to last under the load a Cryptocoin mining rig puts on a power supply.

 I'd go for a G2 series for as long as you can find them (or the Seasonic X-series, ignore the Seasonic "Prime" series as they ALSO went to a bloody junk sleeve bearing fan in those).


 I'm not familier with the BQ at all (dunno if that series had a gold-rated varient).

R9 270 should match the performance of a HD 7870 on ZEC, or slightly exceed it depending on how high you can clock it.
 It has the SAME GPU, but somewhat faster RAM (1400 or 1450 Mhz vs 1250 on the older card).

 Last time I ran my pair of 7870s on ZEC I think they were pulling somewhere in the 130-150 sol/s per card range, with their best reliable performance overclock (1000/1250).

 At $50, I agree that they should ROI pretty fast as long as the high price is willing and the diff don't rise (too far)....

As a general rule, it's cheaper to just build a mining rig that to buy one of those "use a PCI-E GPU on a laptop" adapters - and you get to keep the laptop available for doing other stuff with!

 1. the "data center" model of needing 68 degrees doesn't even apply to DATA CENTERS any more, and was a legacy holdover from mainframe and BIG mini days that has been dying for a while now.
     Look up the Yahoo "chicken coop" data center design they've been using on ALL of their recent data center builds (specifically including Quncy WA), for a well known example of using AMBIENT airflow and flash evap misting for cooling.
     Microsoft and Google have similar concepts in place in their large Quincy-based data centers.

  Cryptocoin models don't NEED "old style data center" level cooling, and never have - if you're putting more than 10% into cooling, you are being stupid about it and going to lose a TON of money (perhaps 15% if you're in a VERY HOT area and have to use a lot of evaporative cooling).

 Cryptocoin mining rigs should be able to handle 90 degree F temps as a general rule without issues, *IF* you have adaquate ventilation. More than that, you either start downclocking or if you can go evaporative cooling.

 If you're going to put serious money into a cryptocoin BUSINESS, you don't set it up somewhere that is hot AND humid, BECAUSE that drives your electricity costs up to the point you can't make any money at it long-term.
 (this is another reason for the large cryptocoin mining presence in Central Washington - it gets fairly hot but not REAL hot here, and it's DRY year-round - average of 8" or so rainfall A YEAR in the Wenatchee area and parts east).

 2. Yes, a FEW folks have managed to get loans and such for a cryptocoin business - but most of them have been at the "BIG" business level with technology-orientated investment banks like SoftBank or venture capitol groups, or the loans have been "signature/personal" loans to someone with a very high credit rating, or in some cases (I know a few of these) folks putting a morgage on their home to finance starting up or growing a cryptocoin business.

 4. don't go by "statewide" rates, which is the only thing the EIA reports.
     Washington state, for the most important USA example, has different rates in different counties to my direct knowlage that range form a low of LESS THAN 3c/KWH FOR RESIDENTIAL USAGE (Chelan and Douglas counties, 4.5 for residential in Grant but the big-business and industrial rates can be LOWER than Chelan or Douglas by a thin hair and ALL 3 are well under 3c/kwh for large accounts) to well over 10 for INDUSTRIAL (the counties Seattle and Tacoma are in, and a couple of the adjacent ones).
 I think the EIA reports Washington as a state at 8.something cents (lower than the national average but nowhere near the lowest in the US) - but Washington also happens to have the 3 counties that have THE lowest rates in the USA, despite not being anywhere near the lowest rate on a state-wide basis.

 Chelan County PUD semi-recently made the news because of it's proposals for how to raise it's rates for "bitcoin mining operations" - but the final result was relatively minor, mostly they kicked up the "front end cost to pay for infrastructure" amounts quite a bit in their large user rates if you were setting up a "high density power consumption" operation - Douglas and Grant already HAD such provisions in their rate codes - and none of that applied to SMALL businesses, only to folks pulling in the MEGAWATT+ range.
 Smaller folks saw only the same very minor rate increases everyone else got, 1-3 % ballpark.
 I think MegaBigPower ended up getting hit with a 10-15% jump in total, and they're probably still THE biggest mining operation in the area - and I'm pretty sure they're STILL paying a fair bit less than 3c/kwh.


 Idaho has some hydropower in the state, and a low population that doesn't have a huge power demand - their lows don't go AS low as Washington, but their highs are nowhere near the SeaTac metro area on electric rates.





Setting up a mining operation as a business is in most respects the same as any small manufacturing business.
 
Your "sell coin into fiat" is your gross income.

If you use coin directly to buy equipment, that has to be counted at the then-current fiat value as income.
 
Your miners can probably be counted as inventory, which offsets income used to purchase them - and you have to count their sale value as income when you sell them.
  You DO have to sell them off sometimes for that, otherwise they have to be treated as computer equipment and depreciated as such - which sucks since ASIC mining gear loses value a LOT faster than most computer gear, and the IRS depreciation allowance for computer gear is ALREADY way too long.
 
Your rent, electric, water, sewer, garbage, and other costs for the space you put the miners into all count as operating costs.
 

Many pools offer you SOME SORT of "bonus" if you are the actual finder of a block - but the details vary, and it's not universal.

 Transaction fees on Bitcoin have gotten to the point of giving a pretty fair percentage of the block reward on an "average" block - I think I saw a figure posted somewhere of "19%" or close to that a few days back - so the folks worrying about block rewards halfings making mining unprofitable have a lot less to worry about.


 On the other hand, the recent wide agreement about "impliment Segwit, and 2MB blocks in the next 6 months" that came out of Concensus might make things interesting - *IF* it actually gets adhered to this time, the last such agreement never got implimented.

Short term, a price increase DOES increase mining income.

 It takes TIME for more mining machines to get built, get bought, get delivered, get set up/installed, and start mining.

 There is also the current issue in ASIC mining that ALL of the ASIC manufacturers are having a hard time gettting enough chips to build miners WITH, due to the limits on available foundry capasity and the HUGE load being put on it by the new generation of AMD and NVidia GPUs, AMD and Intel CPUs, the continuing growth in Smartphone sales and the chips needed to build THOSE.....


 The obvious reason Bitmain had had quite a bit of time in the last year with NO stock of current miners to sell is that they can't get the chips MADE as fast as they can sell the miners.
 It's pretty obvious at this point that the reason BitFury hasn't been selling many chips to third-party miner makers is that their own internal usage + any sales they have made to LARGE farms has exceeded their available chip making capacity.


 This situation seems to be easing some, but the current major shortage on AMD GPUs indicates it's about to tighten up a LOT again, at least for a few months, as AMD soaks every wafer they can get out of GlobalFoundries (AMD has a MAJOR priority there, only IBM might match the priority AMD can pull with GF, due to how GF came into being and the contracts of the time as amended since then).

 It says a lot that the latest AMD "amendment" to their deal with GF specifies terms of AMD being allowed to use other foundries - the implication is that GF might not be able to keep up with the AMD demand at least for a short while at some point in the future, which is a BAD SIGN for availability of GF foundry capacity to anyone else (except IBM).

 TSMC has it's own "overloaded" issues, to the point that NVidia has ALREADY moved some of their production of new cards to Samsung....



 Net result - a price rise DOES give a miner significantly more income for a significant period of time, ESPECIALLY NOW when "new hardware to add" is in short supply all around.



 On the other hand, there is no way the major farms will continue to add capacity if they don't see any likelyhood of such capacity ever achieving ROI - some SMALL miners might, but I don't see a significant probability of the BIG mines doing so (especially since a lot of them buy the chips themselves in bulk and build their own miners, as a significantly lower cost per TH then small miners have access to).

 Don't bother trying to mine Bitcoin at all.
 7 cents/ kwh is too high for a large farm - and the 14 cents of the "base" rate is TOTALLY out of the question for a major ASIC-based farm.

 At that cost of electric, you would want to go 100% GPU rigs or a mix of GPU and Baikal and possibly some Bitmain L3+ as the profitability of the S9 would be VERY VERY marginal after a few months (any other Bitcoin miner model is even worse) with a fairly high risk of never achieving ROI much less any sort of a profit at 7 c / KWH electric cost (14c is OUT OF THE QUESTION) - and even the mix I suggest would be somewhat iffy on ROI achievement, but would at least have a fairly good CHANCE at it.


 EVERY major Bitcoin-specific major farm I am aware of is in an area where their cost of electricity is LESS than 4c / KWH, and MOST of them are paying less than 3.

 Information on non-Bitcoin-specific major farms is a lot harder to find, but the few I am aware of are ALL in 3c /KWH or less areas - on the other hand, altcoin mining has a lot more leeway AT THIS TIME vs cost of electric than Bitcoin mining itself does, and GPUs don't go worthless just because they can't mine at a profit any more nearly as quickly as ASIC do.



 Yes, I'm aware that the efficiency of the R4 is nearly identical to the S9, but it COSTS more per TH to start with meaning it is harder to achieve ROI on.

I've looked seriously at coin laundromats as a place to convert into a farm.

 Power wouldn't be an issue - plenty of outlets, though you'd probably have to rewire them to split the circuits in the drier areas or use a ton of PUDs.
 Cooling - might be a bit marginal, but they USUALLY have some way to vent most of the hot air from the driers out of the building so it SHOULD be manageable.

 The real trick seems to be finding one that's out of business or going out - coin laundromats don't seem to die very often.



 The other question - are the driers Electric (which WOULD work) or GAS (which would NOT work well, as the power available would be a lot lower, washers don't use all that much power).

 I'm personally not that worried about calling it a bribe.
 Too many years spent in 3'd world and even some 2'nd world countries where bribery is the norm and often is perfectly legal.

 However, calling it a "reward program for participation in Folding@Home on Team Curecoin" would be a more accurate description that isn't offensive.

 In some fundamentals, it's very little different from the 10-year-old EVGA "Folding Bucks" program - you do folding work for their specific team, they give you rewards (though the EVGA program limits how much reward you CAN earn through their program per month and limits where you can use the rewards to spending it on EVGA gear where CureCoin doesn't limit you nearly as much on where/how you can spend the rewards you earn).


 In the case of FoldingCoin, the case CAN be made that it offers the potential to be a reward for being part of the larger project, since FoldingCoin isn't team specific (though the requirement to have an entire BLOODY LONG FLDC address as part of your username is a pain) - this "non-team-specific" is what makes it possible to do both CureCoin and FoldingCoin (or EVGA Bucks and FoldingCoin) to earn dual rewards for the same work.
 
 DUAL REWARDS is the only thing that gives Folding any chance to compete with Cryptocoin work in general on profitability for high-end NVidia cards (1070 and UP) over the last year and change I've been aware of CureCoin - CureCoin by itself has never been all that close, and FoldingCoin by itself was worth less than half of Curecoin per unit of folding work done 'till Poloniex delisted Curecoin - and even NOW Foldingcoin isn't as profitable as ZEC much less some of the MORE profitable options for upper range NVidia cards.

 Even the dual-rewards option isn't competative with cryptocoin mining on anything AMD at this time (the Fury line makes a LOT more on ZEC mining, or even on ETH mining, than it can manage via folding) - that might be different for VEGA, but I have some doubts on that score given the "leaked specs" I've seen so far for VEGA.




 Interesting fact I noticed yesterday - Team Curecoin as of yesterday was responsable for almost 20% of ALL Folding@Home work.
 I couldn't find a figure for the percentage done by FoldingCoin participants that were not on Team Curecoin, but it didn't look like more than 1-2 more percent.
 There is the question of "how much of that work would have still been done by the same folks if CureCoin and FoldingCoin didn't exist", but I suspect over half of it is specifically due to the existance of the 2 coins.

 The entirety of MY farm (currently good for about 6.8 MPPD, and will jump to over 8 MPPD once the parts for my next rig get here and I get the system built and up-and-runnning) would NOT be folding at this time or for pretty much any part of the last year, and most of it would never have folded (some of it wouldn't EXIST) if CureCoin and FoldingCoin didn't exist - and none of my contribution to the BOINC project would EVER have happened if GridCoin didn't exist.

 Team EVGA was responsable for about 10% more of the work done yesterday, but given commentary I've seen in the EVGA folding forum I suspect a majority of those folders would be folding even without the EVGA Bucks program (just with fewer and older cards since they wouldn't be able to afford to upgrade or add cards as often).

 Those "almost 20%" and "about 10%" figures were also valid for all work done in the past month, but I didn't see any way to check for the "last year" equivilent.

A lot of power companies also kick their rates up for 2-4 months during the summer, DUE to the load A/C puts on their systems.

 New "workshop" type space actually, and per the Landlord I'll be able to take possession of the new place by the end of the month.
 End of last month would have been a lot better though - we've had *2* heat waves so far this month and the new place has a LOT better ventilation options than the current place, in addition to the more-than-double power capacity.

 I went "pro" as a miner (and have made my entire living from it) for almost a year now, the low-cost electric access to give me the ABILITY to "go pro" so is why I moved from Iowa to Central Washington last summer.