2 in one rig, 172 at -i5 for one, 157 at -i3 on the other (second card runs a lot hotter due to spacing issues).

 Both Gigabyte Windforce OC models running at stock clocks on stock BIOS.

 I did a bit of playing with the clocks on them, but it didn't seem to make a noticeable difference on the cool card and made the hot card VERY unstable.

 1800 watts ASSUMES 120 volts - common residential power delivery is more likely to be 115-117 volts at the outlet and I've seen as low as 110 in some areas NOT counting "brownouts".

 (Power companies don't normally do "deliberate" brownouts any more, as they tended to fry some industrial equipment then the industrial company would sue the power company for the damages - and when you have BIG INDUSTRIAL COMPANY suing they tend to be able to afford high-capability lawyers and WIN some of those lawsuits costing the power companies more than they saved by doing the deliberate brownout).
 
 Also, the outlet itself has a 15 amp maximum rating (assuming the common NEMA 5-15 type used almost everywhere in the US for residential circuits) so even if the CIRCUIT is 20 amps the outlet isn't.

 I generally aim to keep draw from a single outlet to 1200 watts or a BIT over at most, to leave leeway for voltage sag, but in most cases 1320 watts should meet the "80% continuous usage" derating guideline safely.

 PCI-E to multi barrel jack adapters were COMMON back in the Gridseed days, should still be around.

 3 barrels from one 6-pin PCI-E was very common, but I've seen some that did more than that intended for use with the orbs in Scrypt-only mode.

 Blower fan cards can get away with closer spacing, as they direct the hot air AWAY from the other cards.

 Fan-type would work better with more spacing.


 A fan blowing air at the intake-side of the cards along their axis helps in either case - and if you have enough rigs stacked vertically it makes for a great place to put a room-type fan to blow LOTS of air.

 If you attach a fan to the "metal bar" that should be set to suck air OUT, as that's NOT the "intake" side but is part of the exhaust, even with fan-type cards.

 That's not as fast as I'm seeing under Win 7 as I previously posted, much less "more fast".

 Win 10 is widely reported to have issues with some of the NVidia driver versions - seems to be VERY picky there - but with the "correct" driver version(s) is about the same as Win 7.

Is your GT 520 a 2GB card?

 If so, you're hitting the "DAG file got too big to fit on the card with the other data that also has to be there" issue that is affecting pretty much ALL 2GB cards at this point.

 This is a known issue, THERE IS NO LONG-TERM WORKAROUND, there is no way to avoid it, it's inherent to the design of ETH.


 Options are:

 1) Switch to Ethereum CLassic (ETC) or one of the other "ETH clones" that have a smaller DAG file and don't have this issue yet.
 2) Switch to XMR or ZEC or other such coins that don't have this type of issue at all.
 3) Retire the card
 4) Upgrade the card to a card with 3GB or more memory on the card.

 No, our "free time" is spent hanging out on these forums badgering the developers for more speed.

 8-P

 Depends on your cost of electric - it COULD be more profitable to run it at the more efficient undervolt setting.

 Still waiting on a LINUX version, since Windows is an active PAIN to deal with if you don't have 100% reliable power.

 It's accurate - but INCOMPLETE.

 That's the power the GPU chip ITSELF is eating - does not include the memory chips, the interface chips, conversion efficiency loss, etc.

 Doubt it, the memory on those isn't all THAT fast - they should be comparable to a little faster than the Nano/FuryX.

 Curecoin was profitable to mine at 3c/coin - it's just not *AS* profitable to mine the past month or so as ETH ETC or ZEC have been the past few weeks even on NVidia hardware that is optimal for folding.




The price jump the last 2 days has been nice, but it's still not enough to make my 1070s competative on Curecoin profitability to ETH. Right now that would take a jump to over 11k satoshi (8c per Curecoin or close, perhaps 7c after factoring in the tiny FoldingCoin value since that coin dumped out HARD this past month).


 IMO rather than that charity "funding a cloud folding site" they should buy up and burn (or buy up and long term hold) Curecoin to drive the price up and LET THE MARKET DECIDE TO MASS SHIFT TO FOLDING.

 Wouldn't take all that much - just make it competative with other options for profitability and the many already existing rigs would shift back over and start multiplying.

 I'm not in Ed O's league (yet or even all that close) but I could easily exceed 7 Million PPD on 24 hour average (and might be able to push with just my current 1070-based rigs if I had the incentive to do so.
 Unfortunately, I need the money for living expenses and electric so I have to chase the higher profitability coins.



 As far as quantum computing on cryptocoins - the don't work worth beans. Quantum computing  is excellent at figuring out MULTIPLE VARIABLE equations quickly, it is not good at all on "serial very simple computations" as all Crypto is based on.

 280x is 384 bit memory bus, RX 470 is 256.

100TH is 8x Antminer S9s (in most batches) - or well under *3* x $5000 even with power supplies and some left over to set up an acceptable cooling system.


 Sometimes you just have to build up to a goal - take your mining profits and roll them back into expansion.

Or to high an intensity setting - had the same issue on one R9 280x in my dual rig and had to drop intensity a bit to get it reliable.
 The card I had to drop runs noticeably warmer (space for the fans to draw air limits) than the other one.

 Think I'm getting 173/156 or some such on that rig, ended up at stock clocks for best hashrate on those stock cards.

I noticed a couple nights back I didn't have any of my 2GB cards in rigs I had gotten back online - and quickly fixed that.

 Both of my single-card HD 7870 rigs DID work fine prior to the most recent epoch - but they were enough more profitable on ZEC that I slid them over to that and haven't looked back yet.

 Given the 2GB issue appears to be hitting NOW (I'd neglected to consider the other stuff that has to go into the card's RAM) I might just leave them there.


 I've also got my dual R9 280x rig on ZEC, as it's profitability ratio on ZEC/ETH is pretty much identical to the HD 7870 rigs - but that rig CAN go back as those are 3GB cards.



 Been a busy week since I pulled the "in-wall electric heaters" and wired 220 outlets to their wiring to run some of my A2s on - I've actually got some POWER available again!

Notice that our hashrate made a HUGE jump when Curecoin bounced up from 2.7ish cents per to over 5 this summer - then dropped back DOWN bigtime (and appears to still be sliding a bit) when ZEC was released and suddenly ZEC ETH and ETC became quite a bit more profitable to mine on NVidia gear.

 I suspect a LOT of Curecoin folders does not "consider it as a charity".
 What's the point when you can just fold for another team if you're going to do it "for charity"?

 It's getting very marginal at best - I just realised a couple of nights back I didn't have ant of my rigs with 2GB cards back online yet - turns out my pair of 2GB HD 7870s can still run ETH but they're 60-80% more profitable on ZEC right now, so I'll probably just leave them pointed that way even when the profitability drops back to parity ballpark.
 Moving over to ETC might also be an option, it's generally very close on profitability to ETH.


 For comparison - Genoil 1.1.13 on my 1070 rigs with Afterburner set for +550 memory, -200 core, 75% power limit pull 28-29 Mh/s under Windows 7 with the 372.70 drivers. I've had nothing but ISSUES out of any of the 375 series, and NVidia has stated in at least one of their "known issues" on one of the 375 series drivers that they had problems with Folding@home - I suspect that issue likely ALSO affects mining.

Partly that, and partly at least a couple of local credit unions don't do wire transfers at ALL since ACH got widespread.

Keep in mind that most credit unions don't do "business" accounts - a business is a TON more likely to need to do a wire transfer than an individual.



 I also have never had anything but ISSUES when I've been with "real banks" - don't get me started on Wells "bloody FUBAR" Fargo (I told them to shove themselves over ripping me off with overdraft fees due to a deposit THEY put in the wrong account and WOULD NOT FIX almost 20 years ago, LONG before their current scandal but indicating how long that corporate culture has existed) or Bank of "shaft you" America.

 Credit unions are a LOT better at getting things right in the first place, and INFINITELY better at fixing them when they DO mess up, IME - and they don't want to gouge you anywhere near as bad with bloody "rip you off fees" like banks do (University of Iowa Credit Union was a rare exception to that last).

 I find a Credit Union does a MUCH better job of letting me do what I want with MY money, as they don't do everything in their power to shaft me over to turn MY money into THEIR money.



 I'll also point out that back when I DID have a business and had to deal with wire transfers, that they tend to be EXPEN$$$$$IVE.
 That "$25 surcharge to take Bitcoin" might not have cost anything extra at all compared to a wire transfer.

 There never was a 15.20 - AMD driver version naming convention has been "last 2 digits of year . month" of official release date for a driver version for some years now.

 The "pro" drivers seem to have partly broken or changed that, as they use a xx.xx.xx version format - I'm guessing the final pair is "day".