Per the "avalon 6 vs S7" thread recent posts, they are a hair worse efficiency than the S7, 3.65 TH vs. 4.86/4.66 (depending on batch etc), per my calculations should soak about 1100 watts at the wall with a 90% power supply, and the only price posted to date is higher per TH than the S7.

 Unless that price drops a lot, I plan to ignore them - my standard RoI estimate shows them barely managing RoI (they hit RoI in very late June 2015 per my estimate), if I don't include shipping or cost of power supply, at 3 cent/KWH electric - which I don't have access to yet.

 That DOES assume they actually start shipping on November 10, and 3 days to actually get shipped (which is what I commonly see for overseas "overnight" delivery).


 The S7 on that same calculation and same terms achieves RoI about a month sooner - *IF* you can get one ordered today (they've been "out of stock" most of the last couple weeks).

Ouch.

 $1300 PLUS shipping.

 This thing is bloody near a non-starter, it's priced a bit WORSE than the S7 per TH for slightly lower efficiency, if that's going be Avalon's own price.
 Hopefully there's a good chunck of "dealer markup" in that price, it's totally "no chance of RoI" as is unless your electric is VERY cheap.

 (does quick visit to bitcoinwisdom, does standard RoI calculation)

 Yep, it MIGHT achieve RoI. Barely. At 3cents/KWH. NOT including power supply cost or shipping costs.


 So my answer to the original question remains "I won't chose either one as they're both too expensive".

 Hydrofluid has a tendancy in the long term to leak, and I suspect the lubricant most "fluid" fans use has issues at the RPM level with foaming and such.

 Ball bearing overall is much superior to any hydrofluid-type sleeve bearing design in fan usage if both are well done designs.


 Some miners have lasted longer than 2 years. Depends on your electric rate, there are places that the S3 is still profitable and might be some modded S1s out there churning away at a profit.



 All of the Bitmain miners I have or have seen pictures of use Delta fans, except the C1. That focused flow fin set is a patented design feature IIRC.

 Delta fans are noisy?

 Well, the high end ones certainly aren't quiet, though none of the 120mm models I've seen on miners can compare to the old 80mm 80CFM screamer - thank goodness!


 There are VERY few companies that make fans that can push anywhere near as much air as the higher-end Deltas though, and they ALL are comparable on noise level - like the YSTech that the SP20 used, it only seemed quieter due to a better mount and default lower % setting than the comparable Delta models.


 I used to prefer Rotron over Delta, but Rotron got cheap over the last 20 years - their metal-frame ball bearing fans were bloody near unbreakable, and I've still GOT some that are 20+ years old in working gear with 150k+ hours on some of them.

220 uses half the current for the same power vs 110.

Many 220 circuits are designed for higher amperage and a LOT more power than typical US 15amp circuits though - 20 amps 220 circuits for clothes driers are pretty common, for example.
 I don't know of ANY non-commercial dishwasher that uses 220 - they're NOT that high a power draw.
 Some high-end window air conditioners use 220.
 Some homes will have 1 or 2 220 outlets in a garage/workshop area, typically intended for use on a small welder or a largish air compressor, but those are usually installed in custom designs not by default, or were a later add-on by the owner.

 Many electric heating units of the non-portable variety use 220, as does ANY electrical "central heat" unit I am aware of and most electric hot water heaters (those are generally hard-wired dedicated circuits though, not on outlets).

 Should be a way for Bitmain to bypass 1 or 2 chips on the circuit board per string, might need a small amount of glue logic in addition to required jumpers at worst, so they can run a couple chips less per string, run 2 boards instead of 3, and get 3.5TH out of the existing board design.

 Or they could just aim at 3.2 TH with 2 of the existing boards.


 I suspect the November 10 date is related to when Avalon plans to release the 6 - perhaps Bitmain feels the need to preempt upcomming competiton?

In most cases, electricity supplied to US homes is 220v split-phase - and gets split down to 110v at the main breaker panel for most circuits IN the home, usually with a couple of 220v circuits for large appliances like electric driers.

 It's possible to set up 220v circuits out of the panel, but you need someone competant at wiring to do so (most home owners and apartment dwellers would have to hire a professional, and most apartment landlords are NOT willing to allow modifications like that to their apartments).

 The difference in efficiency is noticeable but not huge - typically 1-2% ballpark. There is ZERO difference in the rate the electric company charges.


 SP was focused but not exclusive on units intended for data center use - the SP20 was the only real exception - but they have gone "exclusive" since their merger deal.


 At this time, the only "home" units available (and that seems to be spotty the last couple weeks, Bitmain appears like it might be waiting on the next batch of chips from the foundery to start selling more units) are the S7 Antminers - but that appears to be about to change.
 Avalon has announced their new Avalon 6 unit due shortly (next month I think?).
 Lketc/BW.com announced a couple months ago a home-level miner due probably in December - though exact specs on THAT unit have been a bit spotty.

 Innosilicon likely has at least 2 miner designs in the works as well, one for SHA256 based on their S3 one for Scrypt based on their A4.

 SFARDS seems to have faded back into the woodwork, no news at all on when/if they plan to put more of their SF100 units on sale.
 My suspicion there is "bad board level reliability, took the unit back to the drawing board for a redesign".

The whole point of this thing seems to be to prove that 21 has actual working technology - the economics of the thing make ZERO sense, way too bloody expensive for what it provides.

 I wouldn't even count it as a novelty - too expensive - though some folks buy cars as a novelty 'cause they have more money than sense or more money than they CAN spend.

If the total costs, hardware reliability, and electric usage is the same, and if you have enough space, multiple miners is better - if you lose one or you lose a PS you don't lose EVERYTHING.

Best probabilities for quiet right now are the Avalon 4.1 and the Spondoolies SP20 - SP20 clocked at 1.1-1.2 can run the fan at 20% at almost any reasonable temp and is pretty quiet, I have read that the 4.1 was just quiet period though it's not as efficient as a S5 or an undervolted SP20 can get.

 Everything I've seen about the "next gen" miners (S7, Avalon 6, LK/BW B11) are that they will be at least as loud as an S5 and will probably NEED the full capabilities of their fan setup or very close to work.


 As far as RoI goes - even with the recent Bitcoin price surge, IMO the S7 has zero chance of achieving RoI before the halfing next July, and will be very close to unprofitable afterwards (probably still won't RoI even with that final trickle of profits for a very few months) unless your electric cost is VERY VERY VERY low.

 *IF* (and I class this is a huge if, picture those letters 10 meters high at least) Bitcoin pricing kept fairly close pace on increases to the diff increases of late, that would change - but I suspect the current price increase isn't going to last, as it has no new news backing it and no real reason to exist except POSSIBLY the Chinese economy falling into the pot.

Original announcement from Lketc (which is part-owner of BW and the probable source of their miners) had "December 2015" buried in it somewhere.

 General concensus is that the LK/BW miner will be using Innosilicon A3 chips (tapeout of which was mentioned a few months back), though a few of the BW announcements seem to be implying their own chip design.

 This is NO NEW NEWS, it's just an article talking about info that has been available here in at least one other thread that is months old.

 Delta fans are crap?

 You have a very interesting definition of "crap".

You didn't see that largish Polish mine offering to sell off old "scraps" of it's mines?

3.1-3.2 range would be trivial for them to build - do 2 hash boards instead of 3.

 They've got to make some SERIOUS move on getting the price down per TH to make any S7 viable though.

There will be no more blocks - but this isn't slated to happen 'till somewhere around 2140.

 In THEORY, the transaction fees should have gone up enough on average by them to make mining still profitable - but it's going to take a major change over those years in average fee level, or a major DROP in difficulty AND a few generations of "more efficient" hardware for that to happen at current average fee levels.

Right now, I would WAIT - and do the research.

 Should be some significant competition in hardware miners by the end of the year, should get some price drops going - optimally to the point that some "home" miners get down into a price range that RoI is achieveable on them with electric that costs more than 3c/KWH (and even the S7 is looking VERY IFFY right now for RoI at THAT very cheap electric rate).

There are very few SHA256 coins, and about the only ones with any value at all are BitCoin and Namecoin.

 Blake is NOT SHA256, I don't think Photon is either.


 Sure, it would be NICE to mine some of the cloins like Litecoin and Doge and Dash etc with SHA256 massive TH hardware - but they aren't SHA256 coins so you can't.

Building an altcoin GPU-based rig should be viable - though not cheap unless you couldn't care less about it being profitable.

 Dash (ex-Darkcoin) among others is still profitable with GPU-based mining, though the ROI is very very long. if you use Maxwell-based GPU cards like the 750ti or 960/970/980

 I must live in a sauna in the summertime - 38C is only about 94F and we see higher temps than that routinely here about every summer, and often with high humidity.

 I'm really looking forward to the MUCH drier climate in Central Washington when I get to do my planned move....

With the right mounting holes, this pod board would fit on a Gridseed 80 "blade" heatsink.
10cm is a bit narrower than that HS, the length is plenty short.

 (HINT!)