You're complaining about not being able to get ahold of a card that HAS NOT EVEN BEEN RELEASED YET?

 IRRITATING.

Litecoin is profitable NOW with the older A2 units.

 The A4 will be profitable immediately, but will take some time to achieve ROI - and if enough of them are sold they will eventually make the older A2/Alcheminer and eventually the Titan miners unprofitable at anything less then VERY VERY cheap or free electric.

 It will NOT be a $60/day machine - I've got a bit over 450MH mining Litecoin right now, and I pull in more like $20/day (bit higher this past week with the out-of-nowhere BIG diff drop coupled with the Litecoin price surge of the last month and some). That's GROSS income, not net after electric use - the NET on an A4 will be higher than on my existing A2 farm due to the much lower power use.


 Get used to machines needing a YEAR to ROI - the days of 3 month ROI on new gear is LONG GONE for the most part, and even managing 3 month ROI on USED gear isn't real likely.


 Based on what I'm estimating the cost of a single A4 to be to the "partners", I'm figuring on Litecoin diff doubling over the course of the rest of this year - NOT factoring the "where the heck did it come from" massive DROP of the last week. The main kicker in that estimate is "how many chips will they get out of the foundry how fast" coupled with "how much money are non-founders willing to invest into Scrypt gear".

 S9 at current pricing I'm not sure WILL ROI - many variables involved, and the current price is high even for an "early adopter" premium price.

 My grandpa still had a full head of hair in his 90s. Some of us are lucky and don't GET "male balding pattern" to a significant degree if at all.

Definitely 14.04, as most of the miners have pre-compiled versions for 14.04 floating around SOMEWHERE.

 ignore 16.x - they deliberately broke compatability in 16.x with the fglrx drivers, deciding to go only support the "open source" stuff - which does NOT support OpenCL worth beans IF at all yet and flat out does NOT work with mining software.

 There's a bug in 14.04.4 that can cause issues if you use "mixed" GPU environments, but if you don't try to mix-and-match it's not an issue, and it CAN be worked around even if you do - it's on their "fix for 14.04.5" list, and didn't happen in 14.04.3 for some reason.

The first presumption is that the new RX 480 will actually be availaibly in quantity - track record of the GTX 1080 is making that look kinda iffy right now.

 The second presumption is that folks will be able to get RX 480 at close to it's list price - I suspect that will not happen for a month or two, "new card gouge on price" is an old tradition even when the availability is pretty good.


 On the other hand, unless you can get a current card pretty cheap, it MIGHT make some sense to "wait and see" what actually happens on availability, pricing, and how well the RX 480 can mine in fact as opposed to the current "speculation only" situation.


 S9 though is DEFINITELY a "wait for price to come down and competiton to show up" situation IMO - the current price is a ripoff, though I'm sure Bitmain intends to sell as many units at the current crazy-high pricing as they can while they can to help pay for the development costs on the new chip.

 I'm NOT saying I BLAME them for the pricing, as there's good business logic behind "charge as much as the customers will bear while we can", just that I'm not willing to pay that large of an early adopter premium myself on a miner that will probably be profitable for years.

 I'll take Kano's word over yours.
 His trust factor is quite a bit higher overall on technical issues.

IME trying to mine and play games on the same machine leads to lots of laggyness, many crashes, and other issues.

 Mining programs on windows tend to NOT be "well behaved" and hog all resources they CAN grab.

 Mixing AMD and Nvidia right now is not generally a good idea - the drivers don't tend to play well together, though it CAN be done.

 It's not about the stream processors, or the Nano would be an easy 50MH card as my R9 290s do 30.

 It seems to be a lot more about memory bandwidth, and per Genoil possibly TLB limitations even more.


 Also, on the power consumption - does ANY recent card mining ETH get close to it's TDP?


 IMO wait for cards in hand and REAL stats from testing/tweeking them before trying to claim which is better. 8-P


 As a general rule less power = runs cooler, as long as the fan/HS setup is equally capable or very close.
 Runs cooler USUALLY also means more reliable long-term, but there are a lot of variables in that - the R9 290 is NOT a cool running card but they're engineered to be tanks despite that, per many reports.

Is this an actual 40 watt miner? I remember seeing a pic somewhere with a wattmeter showing 53 (which is STILL impressive, just not AS impressive).

5.5mm x 2.5mm barrel - same as the Gridseed 80 blades?

There's only one GOOD reason for mixing AMD and NVidia in a single rig.

 DirectX 12.


 There are other "niche" reasons sometimes to do so, but if you know about one of those niches you don't need the hand-holding.

 9-)

You have to use a 64 bit OS to mine Ethereum at all.

 Win 10 seems to be very sensitive to driver version on mining - I've seen quite a few "stick with 15.12 Catalyst" comments for Win10.

LINUX is generally easier to work with in many ways - no worries about dummy plugs for one - but the tools are a bit more primitive for overclocking and fan control.

 Nobody has STATED that the RX 480 will mine 30 MH - we're just speculating and estimating based on the published specs.

 IMO more likely it will be a 22-25MH card - but at the PRICE and at the probable power consumption that's a very nice place to be.

 After all, my 290s are legitimate 30MH cards - but they also eat upwards of 200 watts even WITH modified The Stilt bios on them (probably closer to 250) - so while they were inexpensive (sub-$200 from the Newegg "refurbs" sale a bit back) and have a fairly good chance to achieve ROI just from ETH mining, they're not exactly cool running nor highly efficient.

290x will mine ETH quite nicely, abet not exactly at a low power consumption level.

My "big rig" is triple R9 290s Sapphire Reference cards, with one of The Stilt modified BIOS on them to drop the power draw and let them run a lot cooler - still doing some fine-tuning on it to maximise the hashrate, but they're pulling right around 30 MH/s each right now running qtminer (I've NEVER been able to get on bloody github, so trying one of the "more efficent" miners like Claymore's or Genoil's just isn't an option for me).

You just hit one nail on the head - centralization.

 The other one - and for the BIG farms, this matters a lot - is pool fees.
 I strongly suspect that very few big farms run through some pool not their own, and probably quite a few of them run their own P2Pool nodes, in order to maximise profit.


 In my specific case, every time I've even TRIED to use AntPool, it's been nothing but issues that I've never been able to figure out. By the time I get back into BitCoin mining at all, I intend to have a solid P2Pool node set up to point my miners at (already working on getting that set up for LiteCoin, which should be good practice - had it set up for X11 at one point, back when I mined X11).

The best time to buy some Bitcoin in the last year was ... last summer or early fall when it was still in the $200s.

 Right now - hard to tell if it's going to keep climbing or not, much less how far.

 I never used it on my SP20E. It's NOT needed.

 NO.

 An ASIC can only mine the specific algorythm (or algorythmS in the case of the Gridseed/SFARDS dual-algorythm chips) it was made for.

 RX 480 is listed as appx. 8/5 of the memory speed of the R9 390, as I recall.
 Should make up partly for the narrower memory bus, but probably not entirely.

 It's not really AIMED at being an R9 390 replacement though for most usage - more of a R9 280/280x/380/380x range card with some pretensions of arguing with an R9 290/390 on stuff that isn't memory-limited (It'll probably be faster on RC5-72 work than my 290s, but that's a rare "don't care at all about the memory" extreme example).


 It's still GCN supporting OpenCL, should work with current software once the drivers are updated to support the card - might not be "optimised" but should still work.

Just remembered, there are 2GB 750 TI varients - THOSE can mine Ethereum, but they don't get good hashrate (3-4 MH ballpark) so they're only going to make perhaps 50 cents/day at most gross - when you factor in the power usage of the entire system, they might be profitable but not very much so. They're also going to run into the "DAG file grew too large for any 2GB card" issue that will hit ballpark end-of-year timeframe, but for right now ETH might be the best coin to mine on such a card.

 X11 diff has gone WAY up over the last couple months, you'd have to have very very cheap or free electric to be able to mine it now on any GPU. Apparently SOMEONE is finally shipping significant quantities of ASIC gear for X11.

Litecoin has been profitable with some ASIC if you bought them at the right time - the early Gridseed stuff was so overpriced it may have never made ROI, but once the blades got down to $800 they were able to achieve ROI if you had fairly low cost electric. A2/Titan/Alcheminers were profitable if you got them early enough but late enough you didn't pay $8000+ for them.

 For GPU mining, X11 was profitable 'till fairly shortly after that first ASIC for it showed up, if youre electric wasn't high and you had an efficient mining setup - that covers well into the timeframe when Ethereum BECAME significantly profitable.

 I'm sure there were other options I'm not aware of.

EACH BOARD of an S7 (or S9 now) eats close to 500 watts.

 The probability of Bitmain making a 400W or less miner at this point is .... I won't say zero, they could in theory do a long-overdue update to the U3 - but essentially zero.

 Just isn't worth it to them to re-engineer their boards to cut them down to 2 strings just for a few folks that want a small miner.


 In theory, they COULD put out a 1-board "quietish" miner - but I dunno if THAT would be worth it to them.


 1KW is NOT going to be quiet, that's a LOT of heat to dissipate.



 3'd party folks are another story - if any of the manufacturers will actually get a significant number of chips sold to a non-hugh-farm type operation.