One of the other PS of the non-Seasonic maker on the Jameco page was also specced 10.8-13.something, but on IT'S data sheet it showed a "typical performance" page where the PS would actually adjust below 10V.

 The Seasonic unit might not "exceed" it's specs by that much though.



 On the other hand, it's a lot less of a risk taking one of these $60ish PS and digging into one to see if you can modify it minimally to extend the lower end of that adjustment range.

If you're talking powerpeg like the Alpha 3085 or the Swiftech 370, those things were EXPEN$$IVE for a reason - high cost to MAKE that style of HS, though they worked well. They don't work better than fins for crossflow cooling though, they were intended for updraft/downdraft specifically.

 If not, you'll probably need to explain what you mean.



 PCI-E hardware can use a LOT more than 175 watts. Look at ANY of the Radiion "x2" cards, typically in the 400+ watt range, for examples. Just have to use enough power connectors.

 Standard PCI-E does NOT use USB in any way shape or form.

I hope they don't waste the money on wifi. Not all that many folks that use it especially in a mining situation, as it's nowhere near as reliable as standard Ethernet (it's also slow, but that's not really a factor for a miner due to the low needed datarate).

 I doubt they'll go 3 boards or the 800+ watts a 3 board unit would need, that would have broken THEIR existing Hashnest setup and you can't reasonably run 2 such machines from those 1600 Watt PS they also sell and are likely using on their own farm hardware.
 For similar reasons *AND* lack of board space, I dount they'll put 3 strings on one board with the BM1385 - 54 chips would be a NIGHTMARE to keep the corner chips cool on their current board form factor.

 Price above $1000 is unlikely unless bitcoin pricing JUMPS soon.



 S7+/S8 (whichever they decide to go with for a "BIG" unit) will be a lot more TH - I'd guess ballpark 10.5 TH at 2400 watt ballpark in a S5+ form factor. It wouldn't make sense for their "new big miner" when they come up with it to be lower TH than the S5+


 S4+ would NOT have fit as an S6 - it was the BM1382 NOT the BM1384. S6 would have fit the S5+, though they may have decided that since it's not specifically a rack mount unit it wasn't really fitting for the S6 name.

https://bitcointalk.org/index.php?topic=1156770.0

 You're a couple days late.

IMO right now it's best to wait - we're just starting a "generation transition" and ALL of the older generation stuff is getting very iffy on RoI, especially with the Bitcoin Block Reward halving due probably next July (possibly late June if hashrates shoot up too much) unless you have very cheap electric or can get a used unit VERY inexpen$ively.

 SFards (Gridseed merged with WiiBox to form this company) recently released their SF100 dual miner (SHA256 + Scrypt) unit, but it's WAY overpriced even from SFards directly, and the availability has been "very small number of one batch sold, a few distributers got a very few and jacked the price to insane levels". It's got better efficiency than anything that came before, but not good ENOUGH to have a reasonable chance at RoI.

 Bitmain has announced their next-gen chip the BM1385, anticipation is that they'll have a S7 miner unit out "soon" (I'm guessing before end of September, but could be longer) in the 2.3TH/s at ballpark 550Watts concensus estimate of it's probably performance. Price is just as "up in the air" along with the release date, but educated guesses seem to be centering around the 3.5-4.5 BTC ballpark. Most folks feel that Bitmain already has the unit in production, and has been replacing their farm/Hashnest S5 units with new S7s.

 Lketc has announced a 5 TH at either 1KW or 1.3KW (depending on which GH/watt spec you believe), no other solid information though the concensus is that it will probably be using the "we've taped out" Innosilicon A3 chips and likely available by December.

 No announcement on it other than a comment that "chips have taped out" but it seems likely that Innosilicon will also release their own miners, A3-based for SHA256 and A4-based for Scrypt. They've been very light on info though, but that's nothing new for them before they actually have working chips/hardware in hand.

 Spoondolies got bought out or merged into some other company, no clue if they're going to be selling to end-users any more.

 KnC has pulled out of retail sales - not like their units were all that reliable anyway, no big loss there.

 Bitfury seems to have pulled entirely out of retail, and are talking about some silly "mining lightbulb" concept.

 Avalon - just released a mini miner based on their oldish "4" chip, but it's way overpriced and based on tech that's already a bit outdated. No clue if they're working on a next-gen chip at all, much less what and when.

 Alcheminer (OEM for the MAT and at least one other company) had a 256MH/s rated "Alcheminer", but they've sold out on those and no clue if they plan to do any other cryptomining design. Unit was similar in efficiency to the Innosilicon A2 based units, but quite a bit smaller in size than the 90/110 "Terminator" units.


 There WAS a group working on an X11 (+ X13 and X15 I think? can't remember on those two offhand) chip, but they had to halt development due to lack of fundage. They said they were going to look at other financing options, but no word on if they were successfull at that and it's been months.


 I might be missing something, but IIRC pretty much everyone else that's done ASIC for cryptomining in the past is history.

 And your TON of math, from what I saw, didn't even look at difficulty increases, and only looked at a very few power cost levels (10 cent power seems to be more common in the US than 8 cent and 12 cent COMBINED, for example).


 They've been "sold out" for most of the last week now, as are the "used S5" units, I suspect Bitmain might actually release the S7 in late August or sometime in September making the whole point finally moot.

Thought they both specced to go down to 10 or 10.2 volts, guess I misread it. I suspect they'll both get real close to 10V though, the "spec" adjustment range usually is noticeably narrower than the real range.

 Be nice if someone could find some similar PS with 90+ efficiency, though - as long as the price isn't a LOT higher.

 8-)


 I saw the SSE 4501, but Jameco wants somethign like twice as much for those. Almost certain you could get away with one of them per S5 even at MILD undervolt, but not as much margin and the price doesn't make it worth it unless you're VERY VERY space limited.

 I do like the 60mm fan on the Seasonics, should be quieter than the more common 40mm screechers....

The crash happened BEFORE Bitmain announced the BM1385.

 Most of us that have published estimates are estimating that the S7 will have a little less power draw than the S7, at a hair more than double the hash rate.
 I don't think anyone has published a serious estimate under 500 watts, the concensus seems to be 520-550 watt ballpark.

 Probably 1.7854, quite a few areas use the , as their default seperator.

Finally did some digging, and found 2 better PS options.

 Both are specified to adjust down to at least 10V, and will probably go somewhat before that (9.0v probably not but 9.5 perhaps).
 Both are fairly efficient (88% ballpark) - not gold but close, as good or better than running a 95% buck that might not handle the power load from a gold ATX PS and tossup with running a 95% buck from a platinum ATX.

https://e4btc.com/sea-sonic-pf-series-sse-3201pf-12
http://www.jameco.com/webapp/wcs/stores/servlet/Product_10001_10001_2172084_-1

 I wish I could find a US seller for the Seasonic.

 Might need 2 (1 per side) if you don't undervolt a LOT.
 Might get away with one for both sides of an S5 at minimum voltage setting on either of these.


 Jameco showed some lower-price lower-power varients on that second one, a couple of them were also specced at 88% efficient. 20 amp range would probably be plenty for one side at a time with enough of an undervolt.

Bitcoin difficulty has gotten so high it's really not practical at ALL to try to solo mine. Even with that S5+ (biggest single miner to date), your chance of finding a block is tiny - many many YEARS for a 50% chance to find ONE block even if difficulty stays FLAT that entire time (it won't).


 Might be different on some altcoins, though.

antpool, solo.antpool, and I forget what the third one was - p2pool.antpool or some such I think?

Don't bother playing proxy tricks IMO, it's simple and MORE RELIABLE to just set the pool(s) once then not worry about them again.

Difficulty increase at 3% is a bit high, but nicely conservative

This year to date has averaged 1.7%, but the last couple months have averaged a hair over 2%.

 Around 2019.

 8-O

 If NewEgg sold mining gear, it might be worth it to me. Amazon.... hasn't had much and it's been WAY overpriced lately, also not really worth it.

 Perhaps when my mom passes and if I end up with $$$$$.....

What happens if 2 folks pick the same price (like the existing "4" and "4.00" entries)?

 You're not factoring in the reward drops with the amount of BTC remaining on the contract. which drops the average rate of return on the PACMIC by appx. 2/3ds


 Antminer S3 undervolted CAN be profitable, if your cost of electric is low enough. Right now I'd guess around .12c/kwh or less is making money for a very undervolted S3, though probably need a fair bit lower yet to have a chance at RoI if you bought one now.



 Which pretty much means "pray for a bitcoin price jump" because bitcoin difficulty drops are uncommon, small, and so far always followed by an increase bigger than the drop was.



 A better estimate of annualised rate of return would be more like 8-9% at recent difficulty increases, with a fairly high probability of getting all of your principal back in 8-9 months (shortly BEFORE the halving). This is assuming an average difficulty increase of 2% (the last year has averaged 1.7%, but the last couple months have been a bit over 2% *PROBABLY due mostly to Bitmain deployment of S7s into Hashnest/their own farm and resale of the used S5s they're replacing*).

You might be able to superglue a broken fan blade back to the fan, and it might even work for a while, but it won't be balanced.
 Better than trying to run the fan with the blade completely gone though.
 I'd stick a fan grill between the fan and the S5 if I DID try that repair though, in case of rebreakage while running it SHOULD keep the blade from doing any significant damage inside the unit.

 DO NOT try epoxy. It's not strong enough, given the very small thickness of the blade.

 That might be the S8, if/when it's released. Definitely NOT the S7.

I didn't see it either when I looked for it the day after the original announcement.

 The limited number of voltages listed has some interesting implications - seems like we are definitely hitting the hard limits on 28nm (not a shock if it's full custom), though I'm a bit supprised that the max voltage listed is so low.

 I'm even MORE supprised that the "default" is now a mid-voltage point, as opposed to Bitmain's normal "highest voltage is the default".
 THAT makes me wonder if the S7 is going to turn out to be a 17 chip per string device, not an 18 as some of us have speculated - but upon reflection, no way they'd have published that "230 watts at the wall per TH" claim if so.