|
MN isn't THAT much colder than where I'm at in IA, and has similar wind levels.
6-8 months perhaps if he's in the northern part, and some of that only part of the day.
Now I'm trying to figure out how his electric company generates the power cheap enough to SELL it at that rate.
Only thing I can think of in THAT area would be nuclear....
|
|
|
|
BTW double conversion UPS often have less protection than power strips. Anyone can read its specification numbers. Most do not. How many joules does that double conversion UPS claim to absorb? Notice a tiny number.
you assert that any surge seen on input of such UPS coming through AC->DC->AC transformation survives and comes out in original form?? No, but it's possible with a big enough surge to overload the circuitry and pass part of the surge along, or for the voltage to just jump the isolation. 20k volts isn't exactly HIGH for a lighting strike. BTW, the APC unit linked in the message above this isn't double conversion, but while a double conversion UPS (don't think APC makes ANY of those) can probably soak a bigger surge without passing it, even double conversion has LIMITS to how big a surge it can soak. |
|
|
|
OK, just made one last RoI calculation estimate, and it appears there is no way I'll ever get RoI on a S7.
(my move plans took a bit of a hit, looking like next summer now instead of the end of Nov I was originally aiming for).
.....
.....
S7 is a vastly better product, but unless you have VERY cheap electric you won't RoI an S7 unless bitcoin price jumps a LOT sometime in the next 6-8 months, or difficulty increased flatten back out to even LESS than the 1.6% they averaged during the first half of this year about immediately.
This is not true. It is. Do the bloody calculation. Don't forget to add in cost of power supply(ies), don't ignore difficulty increase (I assumed 3% but that seems to be VERY OPTIMISTIC based on diff increases since Bitmain started the "Used S5 sales" that appear to have been driven by them replacing S5 hash power with early production S7s for Hashnest), and figure the 6.7 cent per KWH I used as my incrimental cost for electric, which is pretty bloody cheap in and of itself. I haven't seen S7s on DHGate yet, but I'm sure Bitmain will add them to the store shortly if they haven't done so already.
S7 coupons
why are you selling them when you got them for free
Why sell them? Because at least this way I get SOMETHING out of them. I wasn't going to be able to put the $$$$/Bitcoin together to get an S7 before they expired anyway, and I didn't get them "for free" anyway, I got them because I bought a couple of those overpriced "Used S5" units largely BECAUSE I was figuring on getting at least some sort of payback via coupon for the next gen unit per Bitcoin's announcement about coupons. |
|
|
|
And what do you think about the BitFury 16mn chip? It will be for small customers or only for large mines? Or is only smoke and they have nothing yet?
They announced tapeout - but on a node that's as new as 16nm it'll be months before they get in into production, potentially a year - and even then they'll only be selling it to industrial customers and PERHAPS in that crazy light bulb concept they've demonstrated. |
|
|
|
|
Based on recent diff increases, and current price of Bitcoin, the S7 will NEVER RoI at 6.7 Cents/KWH.
Even a 3% average diff increase kills it's chance to EVER RoI, it remains profitable after the halfing by a few $ a MONTH but it's still a few HUNDRED underwater in July 2016.
It might RoI at 5 cent / KWH or less, but even then it'll probably need a bitcoin price jump or a serious DROP in the rate of diff increase to do so.
IMO wait, unless you have VERY cheap or free electric - and hope competition from Lketc (and possibly others) gets the price to drop a lot.
Significantly better in 6 months is VERY iffy, unless you're a MBP scale LARGE industrial company of the size that Bitfury will deal with.
|
|
|
|
What is the best bitcoin miners
The only LEGITIMATE "best miner" offered to the public at this time is the S7 for Bitcoin and other SHA256 coins. This should change by sometime in December when Lktec (probably using Innosilicon A3 chips) offers their announced 5TH unit. It it probably that a few other companies have equally efficient equipment in production, but Bitmain and Bitfury don't sell to the public any more, Spondoolies is apparently also no longer planning to sell units to the public, and Avalon has made no announcement about anything competative (though they could have something in the works but not announced). KnC was good on performance, WHEN they worked, but the reliability of the company and of their miners was generally very poor to outright crap. There have only ever been 2 chips that could "switch" from SHA256 to Scrypt (technically, they were both designed to mine BOTH at the same time) - the Gridseed GC3355, and the SFards SF3301. Sfards was formed from a merger of Gridseed and Wiibox, so the SF3301 is in reality a second generation of the GC3355. NEITHER of those chips were particularly efficient at SHA, both were fair efficient at Scrypt when introduced but the introduction price was VERY high in both cases - and as the SF100 miner based on the SF3301 is pretty close to brand new the price is STILL way too high to be viable. Gridseed/SFards has a good track record on chip design but a VERY POOR record on board-level design - and based on the very few reports I've seen of SF100 units in the field, they have some serious issues with "not working reliably". The other factor to keep in mind is that you'll find more shares and get more rewards from mining Scrypt coins, but ALL of them are worth a lot less than Bitcoin - commonly around 1% OR LESS of a Bitcoin (LTC is a hair over 1% right now but dropping pretty fast since it's pre-halfing runup). Innosilicon has announced tapeout of both an A3 (SHA256) and an A4 (Scrypt) chip design, but no hard specs on either yet. Possibly in production before the end of the year, but NOT YET AVAILABLE. Welcome to the shakeout year of 2015, when the bubble collapse killed some of the legit companies and most of the attempted scammers..... |
|
|
|
|
The S4+ was NOT an S6.
WRONG CHIP GENERATION.
Dogie, why do you keep repeating that claim?
YOU should know better.
altcoin, there never was a S6 except PERHAPS as a prototype.
It's VERY unlikely that Bitmain ever deployed a significant number of them to Hashnest, or they would probably have offered them for sale as well.
It was skipped, or arguably the S5+ was the "S6" but Bitmain chose to name it differently due to the different form factor.
|
|
|
|
|
If you have free electric, or just want a semi-expensive toy to play with, they might be worth it.
Otherwise, they're so outdated and inefficient you'd have to get one for FREE to ever ROI it - and even then the cost of the power supply would sink the chance of ROI.
|
|
|
|
I hear tell he's working on negotiating lower power rates.
Saw that, think I might be misremembering but I thought he said he was in Kentucky? Might be thinking of the wrong person though. Namecoin is SHA256 and can be merge-mined, no point in mining it specifically. The concept is at least interesting. Monero is, from what little I've seen of it, worthless and in fact DOES appear to be somewhat scammy. Never even heard OF the other two. Cryptocoin does NOT need more altcoins, there are way too many already and most have ZERO long term lasting value. From what I'm seeing I doubt that more 4 cryptocoins will EVER achieve significant name recognition or long term value, and 3 of THOSE 4 are looking iffy.
crackable with a brute force attack
Why does that make me think of Deep Crack - and the probability it could have it's performance matched by an array of GPU-based machines for under $10k today? |
|
|
|
all my PSU's have:
"Heavy-duty protections, including OVP (Over Voltage Protection), UVP (Under Voltage Protection), OCP (Over Current Protection), OPP (Over Power Protection), and SCP (Short Circuit Protection)"
None of which provide any protection vs. INPUT power surges, those are all on the OUTPUT of the power supply protecting it from abnormal OUTPUT conditions and offering some protection to the gear connected to the PS for power from some failures OF the power supply. Personally, I generally go for APC or Square D equipment (all my UPSs happen to be APC, all my breaker panels for decades have been Square D) but there's plenty of good surge protector makers out there. DO keep in mind that some forms of surge protection DEGRADE every time they have to soak a hit - and some hits are small enough your equipment never notices, but it still does some degradation of the protector. Multi-stage surge protection tends to reduce this issue a lot by using one or more early stage(s) that don't degrade (but offers less protection) to limit the surge input to the stages that DO degrade, but even those aren't immune to the issue. |
|
|
|
|
The BF4500 is a *2* generation old product at this point, no way you'll ever get ROI on it unless you literally have FREE electric.
I suspect it's no longer in production, and that the only buyable ones are used - BitFury seems to have dropped all retail sales, dunno why they still list that and the BF3500 on their web site.
S7 is a vastly better product, but unless you have VERY cheap electric you won't RoI an S7 unless bitcoin price jumps a LOT sometime in the next 6-8 months, or difficulty increased flatten back out to even LESS than the 1.6% they averaged during the first half of this year about immediately.
|
|
|
|
|
OK, just made one last RoI calculation estimate, and it appears there is no way I'll ever get RoI on a S7.
(my move plans took a bit of a hit, looking like next summer now instead of the end of Nov I was originally aiming for).
I have 2 $100 coupons available, .1 BTC each or best offer.
|
|
|
|
Both difficulty and global hash rate are rising... So, i think there are some new miners somewhere... and if Bitfury (or anyone) releases a x20 times more efficient miner, the actual miners will be obsolet on next day...  It takes TIME to build and deploy new hardware. Won't happen in a month, might happen in 3-5 *IF* the efficiency gain was actually in the 5x they're CLAIMING over the now current generation gear - and even then some of the current gen might still be profitable for a while for some situations, like "wintertime space heating" or "free electric" like even the original ASIC gear is now. I suspect the last couple of diff increases have been fed mostly by (1) Bitmain making S7s to swap in place for their own farm S5s, then (2) those S5's being sold as "used" and being brought online, more than the fairly FEW sales of the S5+ they had. I find in interesting that the "available" PICMAC contracts are still at 30k, even with quite a few thousand having been sold.... Next: there are 3 hashboards at 400W each making the total for ~1210W. Not quite right. The hashboards eat less than that, then the PS (which is quoted as 93% efficiency) adds a little to the load. *IF* you are using a PS with the same efficiency, then your calculation would be correct and the final current draw from a 110V circuit would be in the ballpark you cite (exact amount would depend on the voltage at the input to YOUR powersupply, 117 VAC is a NOMINAL voltage it can commonly range 110-120 volts and sometimes down to 100ish). In any event, any combination of Gold or better efficiency power supplies running a single S7 and nothing else should easily stay under 12A total current draw. Unless your ambient tempertature is VERY high, you'll have plenty of reserve on a common 15 Amp US "117 VAC" outlet - I forget offhand but I think the National Electric Code cites current capasities at something like 50c before requiring derating, but would have to go back and reread it to be sure.
When S7 release it has more than 12 months BEP (include tax)
Actually, it's BEP per your stated method of calculation of the S7 at release was a little over 8 months, NOT 1 year. I came up with an estimate that it would be 4 BTC before release based on a 250 day BEP, but that was assuming 2 hash boards 2 strings and therefore a little less than half the TH at a little less than half the power at a BTC price when I made the estimate of a hair more than the release date - given the actual figures, it should have been released at more like 8.5 BTC for a 250 day BEP. (Bitmain appears to assume appx. 10 cent / KWH on their version of your BEP calculation). I have ZERO CLUE where you are comming up with that "1 year" figure for it, that's not even close given your base assumptions on the calculation unless your electric rate used is quite a bit HIGHER than 10 cents / KWH.
fits the economic theory-which monopolist ever reduced prices?
Standard Oil did quite a few times in many circumstances during it's "monopolistic" days.
I am surprised that there is no second chinese company at par with BM
Lktec actually announced THEIR upcomming 5TH miner before the S7 or the BM1385 was announced. On the other hand, it seems that a lot of the folks that were part of the last 2 gens of ASIC competition have decided to quit selling to end users. In some cases like KnC, NO BIG LOSS. In others, sad but not suprising as Bitcoin Mining has hit the big times and greatly gone industrial since then. Welcome to the situation of the "family farmer" over the course of this past century, but like most "tech" related thing the generations are a LOT shorter and time goes by a lot faster. If you want to compete with the bit industrial farms, you have to be SMART and you have to play the angles, not just assume you can keep going like Granddad did and still make a living with inefficient methodology. |
|
|
|
I'm not sure but I think that's the same RSP-320-12 I mentioned as one of my original 2 links, just from Jameco instead of from TRC. (if it's not, then I have changed my mind since then on the "optimal" one as it's the one in my booklinks.) Mouser also carries them, I have noticed, but for a couple bucks more, and at least 1 seller on Amazon in the same price range. I dunno if they have enough isolation to use them in series, DO let us know how that works out. |
|
|
|
|
Hint. Even when working with mature process nodes (which 14/16nm is NOT) it usually takes months to go from "tapeout" to "full production" on a chip.
I still don't expect to see repeated 5% diff increases even with the new stuff comming out, but I'm not using 2% as my "diff increase for ROI calculation" estimate for the next few months.
I DO suspect we're going to see a major DROP in diff shortly after the halfing, though I also suspect it'll take a few months for the "non-profitable" machine operators to realise that they're loosing money and to drop those machines out of the hashrate even after such a sudden drop.
*goes back to shopping for adjustable EFFICIENT power supplies to give my S5s a fighting chance at RoI*
|
|
|
|
The Rocks Compute Cluster is used primarily to generate 4K/8K/16K DH keys and large Mersenne primes.
GIMPS? Or something else? Something else, an in-house written DH Key generation routine which uses as it's seed a large Mersenne Prime. These large primes are no where near as large as the primes GIMPS is working on, but still fall into the definition of large. I'm guessing it's the DH key that's soaking all the computation power, as all Mersenne primes up to the 44'th one have been calculated and verified, along with 4 more Mersenne primes past that but not all of the "inbetween" candidates after the 44'th prime have been verified. I can only think of one person/group that is likely to be putting $4k/month into the GIMPS effort - Dr. Cooper of one of the Missouri universities - and that's WITH the backing of the university administration to run GIMPS on most of the computers at that university. At the kind of power level you use, why aren't you in a "cheap power" area? Seems like dropping your power costs from 12/kwh to under 5/kwh would save you a TON of money. |
|
|
|
I maybe plan to buy one PACMiC V3. Is it sure to be profitable ? How much will I earn, because on this point it seems your not all OK.
At this point, it appeast that a V3 contract should be profitable if bought in the next few months. It gets VERY iffy if you buy one appx 4 months or less before the halfing (exact timing depends on price of bitcoin between now and then, and difficulty increases). There is NO "absolute guarentee", which is true for pretty much any investment. On the other hand, a PACMIC v3 RIGHT NOW looks pretty solid. |
|
|
|
|
Having just recrunched the numbers on the S7, I'm also in "wait" mode - I can't RoI the damned thing under any likely circumstances, even with 6.7 cent electric (and it was looking VERY iffy at 5 cent electric for those in low-cost electric areas and even IF they get theirs by the earliest ship date Bitmain has mentioned).
|
|
|
|
I agree that Bitmain seems to be making it more difficult to undervolt than they could.
Side effect of their "keep it cheap" string design, doesn't seem to be deliberate. The old "pencil mod" was about changing an on-board regulator that doesn't exist on their current designs - and was STILL far inferior to Spondoolies methodology. Seem pretty deliberate to me when the hardware design support voltage control but the option get locked in newer firmware, even though if you dig enough you can unlock it again. Just where in the hardware for the S5 is "voltage control" supported? |
|
|
|
|
Old chips, not very efficient by current standards. Unless you are just wanting to play around with them, or have VERY cheap electric, not worth messing with any more.
|
|
|
|
|
Show Posts
