While I have no "dog in this fight", I will ask if there is a way to underclok/undervolt  the Neptune?  If so, have you folks considered it?

The benefits I can see are:

- It would reduce the stress on the one over-worked power connector for the Neptune, and possibly extend the life of the miner. Maybe prevent another melt-down problem?

- While it would no doubt reduce your income by reducing your hashrate, it might reduce your electricity costs even more. In other words increase your NET income by reducing expenses more than you reduce your gross income.

Best of luck.

The reason to "underclock/undervolt" the SP20 is to increase it's efficiency (i.e. GH per Watt), reduce it's heat output, and allow you to slow the fan (i.e. reduce noise). If your electricity cost is $.05/KWH (5 cents), and you don't care about heat or noise, then by all means run it as fast as your power supply will stand.

This also assumes you want to maximize your return or minimize your loss in terms of mining.

You better hope she doesn't read this Forum, or can't connect your ID to you!  

Make sure you buy hearing protection to go with those dozen or more S5's.

I am surprised that you would have 8KW or more of "free electricity".

So where has the hashrate gone that took Slush from 13+ TH to now a smidge under 8TH. That's 5TH that either left the network, or relocated to another pool.

What's up?

My late night update from bitcoinwisdom:

Bitcoin Difficulty:    49,692,386,355
Estimated Next Difficulty:    50,308,957,959 (+1.24%)
Adjust time:    After 581 Blocks, About 4.0 days
Hashrate(?):    351,673,066 GH/s
Block Generation Time(?):    
1 block: 9.8 minutes
3 blocks: 29.5 minutes
6 blocks: 59.0 minutes

Price $245

My poorly calibrated eyeball thinks that the last difficulty jump moved us to a pretty stable state, and what we are seeing now is the normal "variance". I think that we see drops, and then subsequent jumps due to some amount of miner relocation. Just watching the HaoBTC thread and videos make it clear that there is a fair amount of PH that's down for a few days and then "springs back to life" after it has found a new, lower cost home.

The game is all about efficiency and electricity cost. A fact which is probably obvious to all. 

My bet is that you would have a great deal of difficulty actually ordering 2500 S5's from Bitmain.

Save a big chunk of your $1M for a location somewhere with cheap electricity, and also power supplies and infrastructure. Just dividing $1,000,000 by a miner price won't get you anywhere.

That's unfortunate. You can always look at it as 3/4 alive, rather than 1/4 dead. Should run cooler and use less power. 

I wonder if at this late in the SP20 life cycle, Spondoolies will accept just a blade, and not require shipment of the whole miner and the like.

Depending on what Spondoolies says, and how adventurous you are, you might pull the 1/2 broken blade and examine it for obviously blown components and the like.

My fairly cynical view is that they'll try and charge what they think they can get. I don't expect that customer ROI or time frames will matter one bit. At the present time, they have only Bitmain as competition with the S5. Yes they have to worry a bit about the long fabled, completely unspecified and unpriced S7, but that's it. It's easy to adjust a price downwards later if needed. Unless and until Bitmain has a definite product with a delivery date, specifications, and a price, they can just wait. We are also assuming that they are free to produce at whatever rate they want. They may well have bought the number of chips for N miners, and then they would have to wait for another slot at TSMC (or whoever) for more, or pay more to get them "rushed".

I don't mean to be a Negative Nancy, but assuming SFARDS is in the business to make money, I'll bet their CFO thinks more like I do, than what folks might think of in terms of BTC price, difficulty and ROi.

Just my BTC.02.

While the quoted SP20 rate is 1700GH, it's really more like 1500GH as a regular, predictable, not terribly efficient, and loud miner. While I like the SP20, the advertised rate of 1700GH is more a marketing value, than a practical value.

To be completely fair, we have zero idea what the realities of the SFARD 100 are in terms of sound and reliability, beyond the quoted rate and power value. Until a trustworthy guy here has put one in his home and turned it on, with a Kilo-Watt meter I'll take the quoted figures with a grain of salt.

Do we even know if it requires a power supply, and if so what kind of connection (e.g. PCIE 6-wire)? It's also unclear if the 770W (BTC only) figure is measured "at the wall" or what.

Lot's of questions to answer, and verification to happen.

I guess I don't really know how to respond. Folks come to these forums all the time with claims of "free electricity", with dreams of riches in their head. I just wanted to make sure the original poster was aware of the risks associated with this. If that's "cancer", then I will happily move along and never comment here again.

Sorry to have clouded things here with a dose of reality.

Why not have a single power Supply for each new miner. I don't see the big upside of a single 1500W power supply, compared to two equal quality 750W supplies, or perhaps larger.

I'd be willing to bet that SFARDS will think that they should get a "premium price" based on the dual-miner capability, compared to a 1700GH Bitcoin-only miner. While you may be correct in your LTC profitability assessment, I doubt that SFARDS will price it that way.

Assuming it's accurate, this looks way better than the previous dual-miner gear when actually used that way,

Here in the US, the bar for firing somebody is pretty low. Proof generally isn't required. Plan on losing your mining hardware as well, unless you want to claim it, then it's owner is no longer in doubt, is it? In many companies office supply "theft" is better tolerated at a small scan than Bitcoin mining.

Besides the excellent question that phillipma1957 raised, I would also ask: WHY mining?

Mining is the slowest, and possibly most expensive way to acquire Bitcoins. It is NOT away to get rich quick, or with minimal risk.

Late night update for those folks that check-in here:

Bitcoin Difficulty:    49,692,386,355
Estimated Next Difficulty:    49,806,032,346 (+0.23%)
Adjust time:    After 734 Blocks, About 5.2 days
Hashrate(?):    340,748,975 GH/s
Block Generation Time(?):    
1 block: 10.1 minutes
3 blocks: 30.3 minutes
6 blocks: 1.0 hours

Price: $245

@CrazyGuy: I assume that you have not tinkered with the pot on you sample stick? You are using it the way it came "out of the box"?

Great work, and good pictures!

What's kinda odd about this is that in general, there a huge number of buyer protections on Ebay. If somebody buys one and it doesn't get delivered, assuming they use PayPal, they shouldn't lose a dime if it doesn't get delivered. Ebay doesn't do "pre orders" and they they discourage/forbid Bitcoin as a form of payment. I don't understand how the seller expects to serious scam anybody.

Looks to me like the original KB1150 power supply had a non-standard 8-pin connector that probably has a subset of the wires from an ATX 24-pin connector.

If the old KB1150 isn't totally dead, you might be able to tinker with the wires on the 8-pin connector using the ATX color code as a guide on what might be where. . If the KB1150 is truly dead, you might open it up and see what wires go to which pins on the 8-pin connector.

Is the manufacturer still around that supplied the KB1150? They might be able to supply pinout info.

While I think it will be positive, I don't think it will be as bad as 5%. With little room available, I'll take:

+0.76% to +1.00%

Now I just have to hope that HaoBTC takes their time in finishing their move to Kangding.