I think the issue is that someone who was linked to that scam now works for them. I'm just sharing my experience, and I've only dealt with the owner and the guy that runs the data center, not Matt Carson.

I don't recall Oregon Mines itself ever being involved in any scam, and they haven't scammed me. Quite the contrary, my SP20 chugs along, and I just have to pay my monthly invoice.

I guess I'm the odd man out here, but I've been hosting an SP20 there since March with no problems whatsoever.

Please put me on the list for 1. Thanks!

On the flip side, if price is stagnant or decreasing slightly, and as a result difficulty remains stagnant or even decreases some, that miner may run profitably for a long while, and you may end up with 4 bitcoins at $150 by the time the miner isn't profitable. In this case you've essentially purchased 4 bitcoins for $100 each and you started with $400 and ended with $600. But if you had bought 2 bitcoins instead you'd  have spent $400 and ended with $300.

Obviously there's other variables at play, but mining could be a way to protect against a drop in price at the expense of some upside if the price increases. 

That is a bit high ... Li must have gotten a great discount on the S5's back in January. Anyone lucky enough to have bought in back then can even sell them back for a little more after 8 months of mining at a profit. Not bad. Anyone considering mining with Li should do so in confidence, and I'd jump right back in if the miner cost were lower, or the electricity cost were a couple cents lower.

I'm guessing Bitmain isn't offering much of a discount this go-around. Maybe that will change going forward - bitcoin sales from the website seem a bit slow for this batch, around 550 or so in the first three days, which is about 2.7 Ph/s, when it seems that about 20 Ph/s went online this difficulty period despite the price drop late last month.

Right, but I'm in the tropics and the lowest the temp goes this time of year at night is 27C, daytime temps at 33C. With a little AC my S5s have not been in ambient temps above 35C. Without AC it would easily go above 40C during the day. I'm wondering what would happen above those temps. For an S5 with stock fans the board readings would probably still be in the mid-60s. I've run an S5 with quiet fans with temps in the low-mid 70s for months without any impact on performance or efficiency (edit - I'm not recommending this, btw, I'm sure running it that hot carries a higher risk of malfunction).

I'm curious to see how the S7 would handle higher ambient temps given the dense chip arrangement.

Shipping cost to the states isn't bad at least - I get $34.24 for one, $57.82 for two. I'm on the fence on these... a coupon might push me over

Looks like no coupons for S5 buyers - oh well, if anyone has an extra that won't be used I'll happily make use of it.   My username on bitmain is the same as here.

It's too bad - this chip could be used to make a great home miner. A 60 chip design like the S5 with the improvements they've made in cooling could probably run at about 1600-1700 gh/s somewhere between 0.6v and 0.66v and probably pull less than 400 watts from the wall. Cooling a miner like that would not be loud at all with good fans. That's not too far away from the S3 in terms of heat generated. And they'd probably sell easily for around BTC3.

The S5 has a limit of 35C. These are ambient temps, but I can see a data center getting higher than 40C and certainly higher than 35C without AC.

I'm sure there's many S5's that have been running at ambient temps above 35C in the past couple months. Any performance issues?

We don't know where voltage is set ... but the spec sheet lists 32.5 gh/s per chip at 0.66v. If that's where they're set, with 162 chips that comes to 5.265 th/s. This is pure speculation but that would seem like a logical limit for overclocking. But I don't fully understand (if I do at all) the differing effects that voltage and clock rate have on hashrate... hopefully someone that does can let us know if this is a reasonable assumption.

Does "Batch one scheduled shipping date is in late September of 21~30. If shipping out 10 days later than the scheduled one, it won't be considered as delay" mean that it would not be considered a delay if they are shipped as late as October 10?

And unfortunately unlike the S5 I don't think that we'll be able to replace the fans with quieter ones. 1200+ watts of heat have to be dissipated from a similar amount of space as an S5, which pulled 590 watts.

This is not a home miner by any means.

I'll take 0.51-0.75.

Side note - I was looking at Litecoin diff. and price performance around the recent halving (and prior to it).

It looks like the price settled in around $1.40-50 in April-May after a long decline, with difficulty fairly stable, bouncing around 40k. Then things started going crazy as the halving was approaching, and price and diff. increased.

Now, price is (somewhat) stable around $2.80-90, and difficulty is dropping, right back to around 40k (a little below). If this equilibrium holds, that will mean a wash for long term miners as the price doubled and difficulty remained roughly the same. Obviously bitcoin is a different animal, but interesting nevertheless.

Yep, folks have been using this thread to trade existing hosted units as far as I can remember. I wonder if Li could set up a little OTC exchange on the mining host website where we can buy/sell hosted miners from each other. There seems to be a decent secondary market, and that could boost sales. 

Probably, but I'd be surprised if the jumps are significantly higher. If diff. increases by 3% the rest of the year, that would be an increase of roughly 90PH between now and 2016.

By comparison, the network hashrate is about 100 PH higher since the start of the year (average diff. jump of under 2%), which is when Bitmain began sales of the S5. And a good chunk of that came from Bitfury and other manufacturers.

The slow growth this year also means that a significant amount of the network hashrate is likely comprised of hardware that will be at least 2 generations behind (S3 and earlier). Assuming a stable exchange rate, a significant amount of S3-gen. hardware will be forced offline if difficulty increases too fast, just as hardware >1 w/gh went offline over the past 8-10 months or so. This too will be a limiting factor on growth, and S3-gen. hardware is likely still a large chunk of the hashrate out there.

What's more likely is instability in difficulty going forward - I wouldn't be surprised to see, for example, an increase of close to 10% followed by a decrease of 6-7%. (edit: actually, I'd be a little surprised - a 10% increase means 40 PH went online in a 2 week period, allocating 32 PH of that to Bitmain (leaving the 2% 'normal' growth), that would mean 16,000 hypothetical 2TH S7's shipped in a single difficulty period, variance aside).

The exchange rate is the most important variable with respect to difficulty right now, I think, and all this assumes a constant exchange rate.

While I told them I'd be happy to review a test unit for them, I can tell you that this alone raises red flags for me.

There's no reason to sell an S5 for $199. The market price is twice that.

As far as I know, yes, that is where most of the estimates come from.

It may very well be variance. As always, wait a few days to confirm an actual large increase in hashrate. There wasn't an 80-100PH increase on July 6, for example. Sometimes everyone gets lucky around the same time.  http://www.coinwarz.com/network-hashrate-charts/bitcoin-network-hashrate-chart

Aside from electricity costs, timing is very important. A home miner with reasonable electricity costs that bought S5s or SP20s in Jan. 2015 has done very well, as the resale value of those miners has actually increased since then, while they've made decent revenue over cost.

Someone that bought the three SP20 deal for $1195 with free shipping, for example, can probably get that much back even on eBay after fees right now, and will have essentially had free equipment to run for six months. Of course, that is obvious in hindsight.

The best thing to do is closely watch difficulty changes, stay up to date on what manufacturers are up to, and monitor hashrate for new, unannounced farms that may be popping up (or new hardware being tested - it's pretty easy to spot when Bitmain started running their S5s if you study it close enough).

Basically, mining can be profitable. But you have to put some work into it, not just learning how to run the equipment (that's he easy part), but also studying the market. Luck helps.