I'm upgrading to new hardware and have to make room. So I have 2 S3s from batch 3 for sale.

The price is 0.4 BTC each, you provide shipping. Power supplies not included.

Both S3s are solid performers. I kept notes regarding tuning and performance and can provide them if you're interested.

If you want an escrow, we can use OgNasty. Shipping in original boxes from Bitmain.

I can ship as early as Monday 12/1.

US Sales only please.

Thanks for getting back to me. Since my SP20 will be at a data center where we have fixed power costs, I don't care about how much power the unit takes or how loud the fans are. I can see that your situation at home is different, so under clocking is interesting for you.

But I think I'll take your advice and play it conservative for now with 250.

On my batch 3 S3s the cables from my CX750M PSUs have never been warm to the touch. This is with one PSU per S3 and all 4 connectors. I'm overclocking both S3s, but only to a frequency of 237.5 because that was the sweet spot. I don't know what the S3 draw is at that frequency, but I some people have reported it at around 400W at the wall.

I've searched the web and the Corsair site for details on what the CX750M cables are supposed to handle and haven't found any information.

I'm interested in replacing two S3s with a SP20. The SP20 would be deployed at a data center where we have 208 volt power and active cooling. Power consumption, cooling, and fan noise are not a concern.

My hope is to reuse my two CX750M PS units that are currently driving the S3s. My question is how far can I safely over clock an SP20 with these PS units?

The Spondoolies site (which references the CX750M in their testing) recommends a 20% power reserve. So theoretically 1200 watts out of 1500 should be usable.

In an earlier post, someone indicated that overclocking to 1.8 THz consumed 1200 watts. But I don't know if that was at the wall or on the PSUs.

I've looked at the Spondoolies documentation, but the only information I could find was how to use their GUI to specify power parameters for overclocking. The information is pretty basic and sparse where it comes to actually tuning a unit for performance. Note that this was for the SP30; I couldn't find the SP20 User Guide on their site, but I assume that the configuration GUI is largely the same across all their products.

I'd be interested in any real world experience that people have had with their SP20s that would help me estimate how far I can safely push it with two CX750M PSUs.

Thanks in advance for any suggestions.

I'm interested in this deal too. But I will not do a pre-order, and if that means no free shipping after units actually materialize, then it will be too risky for me from an ROI perspective.

With a 3% difficulty increase, free electricity, and an estimated 10% return increase for using a well performing multipool, 54 days to delivery, and no power supply cost (I already have a couple), ROI is almost 140 days out. Add another $150 and the ROI is something like 300 days out.

If units actually ship and the free shipping is still available, then I'll take this risk. In fact, I'd probably go for the 200 MHz unit. Otherwise I'll pass.

This price is close, but not quite there...

If you don't include shipping or power supplies, and you have free power, then at a 10% difficulty you're looking at about 180 days for break even. If you're willing to gamble that the difficulty might stagnate at around 5%, then you're still looking at 120 days for ROI.

Now if these came down to around 1.5 BTC (shipped), then we'd be in the target area for a 90 day ROI at 10% difficulty. I would buy then because I've already got the power supplies and free power.

It looks like a nice box, but at even at the reduced price it's just a non-starter.

In order to get to a break even point of 90 days, the price would have to fall to $600. That's as calculated with free power, not including shipping or PS costs, and a 3% difficulty increase. That's close enough that I might take the risk, if the price ever falls that low.

I'm interested in new hardware, but until BTC or LTC goes way up, or hardware manufacturers dramatically cut their profits, I'm just going to keep the miners I have and stay on the sidelines.

After little less than 80 days, my B3 S3s have broken even. They've mined as much BTC as they cost (factoring in the refund). This is with free power.

There really is no deal like this out there now. Everything is too expensive.

Bitmain, I'm sure that you're not moving units like you were before. I understand that you need to make money. And clearly BTC's price isn't helping.

But if you want to move units, you have to cut prices. If the S4 had a chance of a 90 day ROI, I'd buy today. And I'm sure I'm not the only one sitting on the sidelines waiting for a reasonable deal to come around.

For grins I priced out a pair of B10 S3+'s shipped to California. The price is 1.492 BTC.

Wow. I guess I got a deal for 1.45 delivered for batch 3, minus the 0.11165 BTC refund for under performance.

So effectively I paid 1.33835 BTC for two S3's over 3 months ago.

I've been mining the B3 S3's for 76 days and am less than 2 days from ROI.

I guess those were the days...

These units are just too expensive. With free power, no shipping costs, and a generous 3% difficulty increase, these have no chance of ROI. The calculation I did showed these still in the red 1206 days from now, which is effectively the same as infinity.

Now if the Terminator was more like $700 (assuming that shipping isn't too expensive), then they start to make sense with an ROI around the 90 day mark.

Of course the 10 unit minimum (assuming it applies to both models) also means that these would be outside the reach of most home miners without a group buy.

Nope. No mining necessary.

If you believe BTC will spike to $1200, then buy it now at $376.76. No need to risk hardware purchases or electricity costs.

Hasn't the myth of mixing currency speculation (i.e. wishful thinking) and mining been so thoroughly debunked that it's beyond reasonable doubt?

Without a coupon, and with US shipping, and with free power, and 15% difficulty, the S4 is still squarely negative ROI after a year.

Unless difficulty stagnates, these will lose BTC.

It's really too bad. If it wasn't for the price, the S4 would be the perfect next step for me in mining hardware. It's rack mountable and within the remaining amp budget that I've got at a colo facility. But at nearly 4 BTC shipped, it's insanely expensive.

Instead I've bought some BTC and will hold onto it.

Maybe the stars will align and BTC's price will recover in time for the next S4 batch announcement, but I'm not holding my breath.

It's really too bad that there's nothing out there like the deal we got with the S3. I bought into batch 3, and will ROI in a couple of weeks (with free power).

PoolWaffle, it looks like unconverted script is piling up. I'm seeing almost 3x unconverted verses earned.

Maybe something is stuck?

What? No songs this time?

With the current exchange rate of 411.8 USD, shipping costs to California is 0.391. That brings the grand total to 3.912.

I went through the process (without buying of course) to find out what the total would be.

That's just crazy.

+1

Personal attacks aside, this is the truth:

If the hardware costs more in BTC than you can mine in a reasonable time, don't buy it.

If you think the BTC is going to go up, then buy BTC.

Mixing currency speculation into your ROI calculations is just wishful thinking...

I did a calculation with a generous 15% difficulty and 2 TH/s output. At 100 days, an S4 will produce a little more than 2 BTC. This is before power costs (which in my case is free).

These units have to come down to under $1k, or BTC has to rally to something like $725 (getting the price per unit down to about 2 BTC) before these would make any sense.

Since neither event is likely to happen, I'm staying clear of the S4.

Everyone seems to assume that Bitmain already has S4s on hand and that they're mining with them now. While that would make business sense in an ideal mining manufacturer's world, I doubt that's actually happening.

Do you remember the S3's where they had to restate the performance because they only got the boards in hand days before shipping? Now, they could have been lying about that, but something tells me that that was accurate. They only got boards delivered at the last minute before assembly. It was a surprise to them that the boards underperformed once they actually tested them. To their credit, they issued the option of a partial refund or coupon.

I realize that none of us have any real visibility into Bitmain's business. But my gut tells me is that they run things very close to the bone. They have no S4s in hand yet. They haven't put thousands of them into service waiting for bitcoin's price to recover. Towards the end of September they will get boards delivered and build out the S4 systems. Hopefully once they test them, the S4 boards actually deliver the estimated performance.

The delay in the S4 sales details probably is not related to the drop in bitcoin price. That part is just bad timing. I think what's going on is just the usual delays that happen with businesses that run with limited resources and small margins. But you can bet they're not sitting on top of a warehouse of S4 units mining the crap out of them. They haven't built them yet.

All that having been said, I'm still interested in the S4s. The S3s turned out better than I expected. If they come up with a decent price for the S4 that makes sense, I'll buy. If not, I'll wait.

Sorry, I meant the S3. Too many S's