I got some NMC yesterday. I don't think there are any stats for it though. At least i can't find them.

Is there a risk of bricking the machine doing this, or will the controller still work (allowing further edits or restoring the old values) but the hasher not work?

I don't see the relevance of this example to a mining deal that is being offered right now.

Right now -- not in September or November, but right now -- it costs around 0.2 BTC per share to get in on this deal. If you want to conclude that is a good deal you have to believe you will get >0.2 BTC per share out, otherwise you can just buy the 0.2 BTC right now and put it in a drawer. That will be a better investment than any mining deal offered right now that gets <0.2 BTC out.

No not at all. The resale value contributes to the 0.2 BTC (or more) you need to get out. As long as that happens, you are better off with the mining deal. If it doesn't happen, you are better off having put the 0.2 BTC in a drawer instead.

Conceptually it isn't that complicated. What is complicated is coming up with an accurate prediction of difficulty before and during the useful life of the equipment.


In that case your ROI in terms of BTC is positive (0.3 > 0.217). You haven't identified some other path to profit.

47 seconds wow!

Fingers crossed that one doesn't orphan. Would be such a shame.

BitMinter should be fine. Eligius is definitely fine (I've been using it). I've had trouble trying to get them to work with p2pool-based pools. I get about 30% rejects (although I'm not actually sure how much that is hurting my earnings). Changing share difficulty doesn't seem to help. 

It seemed low because the autoswitching pools are paying out a lot more value. Please try that for a few days at least.  I know how to calculate electricity usage.

I think you have to consider depreciation though. My 3-year-old 6870s seem to be worth close to half what I paid for them. A 3-year old BFL single will almost certainly be a paperweight, and it might be a paperweight after one year. 

But if power is the limiting factor staying away from GPUs probably does make sense.

$200/month seems low to me. Are you using a switching pool (multipool/hascows/middlecoin)? They've been running around twice the output value of just mining LTC recently. The latter two pay out in BTC so you don't need to worry about a bunch of altcoin dust.

That's pretty extreme.  How about expanding the current GPU capacity a bit? If nothing else they seem to hold their value. I have some 6870s that I've been running for 3 years and not only are they still mining but they still have a bit of resale value! Who knows maybe a good primecoin (or some other coin) GPU miner will come to light and even if (and more importantly when) scrypt moves to ASICs, the GPUs can mine that, or otherwise liquidate them.

That's the problem. If you want them to run quietly you need the room to be cool. The fans are temperature controlled.

Scrypt hardware at this point is GPUs. The key is choosing the right ones and operating them efficiently. Scrypt ASICs are promised but vaporware so far.

Agree. There are some good deals on 280x stuff right now, when you can find it in stock. With smart system builds (I have the utmost confidence in OgNasty here) these have pretty good ROI with scrypt.

There may be an issue with scrypt ASIC's theoretically on the way, but GPUs look to pay for themselves in a month or so, which is sooner than any scrypt ASICs are likely to actually exist IMO, and even then they should still have decent resale value as gamer cards.

I'm doing some of this in my own operation, but I don't have near 75 MH. Wow.

Can we get some Christmas luck in this pool please:

86
46
96
59
252 (w00t?)
46
16
75

Applause!

It makes a difference to mining because you can always just hold the BTC and wait for them to go up.

Case in point this group buy currently costs around 0.2 BTC per share. If you just hold onto those 0.2 BTC you might do very well if the price of BTC goes up. Assuming the price does go up, to do better by sending it off to buy hardware you need to somehow eventually end up with more than 0.2 BTC at some point in the future.

One last comment on the debt and then I will largely shut up, unless there are new points to be addressed.

The lenders made an investment on terms that they would be paid back from "NASTY MINING's 25% share of NastyFans' donation distributions" (for example see https://bitcointalk.org/index.php?topic=258099.0). They made a somewhat risky investment with a potential return of 5% over a few months (i.e. 21.5% annualized) if the donations were enough to pay back the loan that fast (in fact some of the lenders have been repaid). This was a very generous return and it stands to reason there was potential down side.

In fact the donations have slowed and payback has taken longer, but along the way everything has been done as promised when the loan was offered and accepted by all parties involved.

I don't feel we have any obligation, ethical or otherwise, to redirect our donations in order to pay back the lenders faster than the deal originally called for. If we stick with the 25% agreed terms then NASTY FANS is not particularly at risk. The most we will ever have to pay to service this debt, regardless of what happens to BTC prices, fiat, etc. is 25% of donations. That is quite reasonable, responsible, and not particularly risky -- again in line with a 30% debt to equity ratio.

Is there a way to undervolt/underclock these devices?

If we get another 10 bagger, we will all be very happy, both with respect to NASTY FANS and, I suspect, our other activities (in many cases for people in this group). You are looking (myopically I would argue) at the debt as one small piece of a much bigger picture.


I think its entirely plausible the seat price will do very well. We have incoming capacity that is several times the existing capacity. If BTC does another 10x in the near future that capacity will produce millions of dollars worth of BTC. Yes the difficulty would eventually adjust to the price rise but that takes time. At present ASIC orders are already months behind production capacity. No matter how much people may want to mine more, it just isn't possible.

At some point the prediction of 100% increase in difficulty per month is going to be wrong. That is a certainty. Whether that happens by March I don't know (and neither, I suspect, does anyone else).

That (and really only that) is the key question here. If you plug in 50% difficulty growth per month these come out wildly ahead.  If you stick with 100% they come out behind.

I agree the hosting costs are pretty reasonable.  At 1500W that's $162 just for electricity at 0.15 kWh.

I didn't say silent but they are quieter than many PCs (thought not the quietest) that I've built. I expected louder. Honestly given the amount of air flow needed to cool ~400W, I don't really think they could be made much quieter (again recalling my experience building PCs, using "silent" fans, etc.), and I expected worse from the comments. Since they are temperature controlled, the issue may be the temperature of the room. No idea where you are located but if it is a warm climate that could explain it. Winter here now (and cool inside).