ANTMINER S5: 1155GH(+OverClock Potential), In Stock $0.319/GH & 0.51W/GH

[dmwardjr]

If you're talking about 18,000W and 15 SP20s, why are you quoting the single unit price on the SP20? 15 SP20s cost $5500 shipped, which is a considerably better price.
Downclocked to the same speed as an S5, the SP20 is as efficient or moreso that an S5 (neither of mine hit .510 J/GH on a Platinum supply, BTW), and the price is close to the same.

30 SP20s will set you back $10,990 shipping included, while 30 S5s will cost $11,100 plus shipping (~$11,900 shipped to the US). If you have 30 coupons you could probably knock $1000 off that price, but that just makes it a wash.

Owning both, if I had 18,000W to play with and had to choose between the SP20 and S5 at the same price, power and hashrate I'd definitely go with the SP20. Either would work, but arranging that many would be nicer with the SP20 and it just feels like a much more rugged unit.

And when doing profit calcs, don't put the SP20 at 1.7 TH/s. Most people get about 1.6 TH/s out of it, almost no one gets 1.7 TH/s. Whereas the S5 is easily overclocked to 1.3 TH/s from virtually every unit.

Another excellent point to make and it's doing so while having more power efficiency than the SP20.

[MrTeal]

If you're talking about 18,000W and 15 SP20s, why are you quoting the single unit price on the SP20? 15 SP20s cost $5500 shipped, which is a considerably better price.
Downclocked to the same speed as an S5, the SP20 is as efficient or moreso that an S5 (neither of mine hit .510 J/GH on a Platinum supply, BTW), and the price is close to the same.

30 SP20s will set you back $10,990 shipping included, while 30 S5s will cost $11,100 plus shipping (~$11,900 shipped to the US). If you have 30 coupons you could probably knock $1000 off that price, but that just makes it a wash.

Owning both, if I had 18,000W to play with and had to choose between the SP20 and S5 at the same price, power and hashrate I'd definitely go with the SP20. Either would work, but arranging that many would be nicer with the SP20 and it just feels like a much more rugged unit.

And when doing profit calcs, don't put the SP20 at 1.7 TH/s. Most people get about 1.6 TH/s out of it, almost no one gets 1.7 TH/s. Whereas the S5 is easily overclocked to 1.3 TH/s from virtually every unit.

Yeah, mine won't get to 1700GH/s regardless of cooling. 1650 is fine, though I only run mine at ~1250 to maximize profits. Overclocking the S5 is a little more variable, as it somewhat depends on your PSU. The difference between 11.9 vs 12.3V in negligible on the SP20, but it actually has a decent impact on both overclocking room and efficiency on the S5.

What kind of power supplies are you using where you can adjust the 12v line?

I generally use server supplies, but even different models of ATX PSUs (or sometimes even different units in the same model) can have different "12V" outputs. The ATX spec is actually +11.40 V to +12.60 V. Especially with big crossloads (IE using only 12V and nothing from 5V or 3.3V) different PSUs can range anywhere from 11.8V under full load to 12.2V or so, even without purposefully adjusting them. If you have one feeding an S5 at 11.8V and one at 12.2V, you could see a 6-7% difference in power consumption just from the extra voltage.

[Tupsu]

If limited to 18,000 watts, one can obtain more hashing power with the S5 versus the SP20.

One single SP20 @ $479.00 without shipping [1,700 GH/s @ 1200 watts = .705 watts per GH]
One single S5 @ $370.00 without shipping [1,155 GH/s @ 590 watts = .510 watts per GH]

Let's say power limitations at your location is 18,000 watts:

18,000 watts / 1,200 watts = 15 SP20's x 1,700 GH/s = 25,500 GH/s (25.5 TH/s)
18,000 watts / 590 watts = 30 S5's x 1,155 GH/s = 34,650 GH/s (34.65 TH/s)

Without going any further with comparisons on ROI, one can easily conclude the S5 is easier to achieve ROI than an SP20.  This is based off the price of one single unit.  If one wishes to compare based on bulk pricing, they must use both bulk package prices [SPT's 15 x SP20's versus BITMAIN's 52 x S5's].  Even if one wishes to choose the bulk price of 15 x SP20's versus the price of one S5, it would still be more worth while to get the S5 at this point in time.  The main reason why it would be worth it IS TO KEEP FROM LIMITING THE AMOUNT OF HASHING POWER I CAN HAVE WITH THE LIMITED AMOUNT OF WATTS AVAILABLE TO MINE.

If limited to 18,000 watts you will have approximately 36% more hashing power with the S5.

Just saying...

This is not true. If  power is limited, buy more SP20

1x SP20  1000GH/s   0.45W/GHs  Can the S5  do this? No.

I am happy that I bought 2015-01-26 at night three S5 for 3.983 BTC   ( Today it costs 4.894 BTC)

But I bought later on the same day   3x  SP20  for  3.954 BTC , and it was a much better deal.  1BTC=303USD

 
3x SP20   3 TH/s   0.45W/GHs   total 1350W

3 x S5       3,45TH/s   0.51WGH/s    total   1750 W
450GH/s   more  costs  me every hour    extra 400W  

nearly 300 kw / 30EUR per month, But only produces 0.0122 x 0.45x 30=0.165BTC per month

[Tupsu]

You are using the bulk 15 unit price by SPT.  I'm sure if you PM Bitmain, they will come up with a discount bulk price for 30 units the same way they come up with a bulk price for 52 units AND offer free shipping or discounted shipping.  So, just because SPT has a quoted bulk price FOR 15 UNITS does not mean Bitmain will not do a bulk price for 15 or even 30 units.

See what I'm saying?

I do agree with you for the under clocking of the SP20 to the same efficiency as the S5. However, do not assume Bitmain will not give a discount on shipping or no shipping costs for 30 units. That's all I'm saying.  Just because SPT actually lists their bulk price with shipping costs ($0.00) for 15 units does not mean Bitmain would do similar with 15 or 30 units.

ANTMINER S5 BULK BATCH 4 1155GH/s x 52  
68.745 BTC
 ( 18247 USD )
Shipping Cost  7.529 BTC
 ( 1998.391 USD )
Total    76.274 BTC
 ( 20245.391 USD

...............................................................................
ANTMINER S5 BATCH 4 1155GH/s  x 52
72.358 BTC
 ( 19240 USD )
Shipping Cost 7.516 BTC
 ( 1998.391 USD )
Total  79.874 BTC
 ( 21238.391 USD )


Where do you see here discount from shipping costs or even  free shipping? Only in your dreams...

..
I do agree with you for the under clocking of the SP20 to the same efficiency as the S5

Even better. 1000GH/s      0.45W/GHs

[Syke]

nearly 300 kw / 30EUR per month, But only produces 0.0122 x 0.45x 30=0.165BTC per month

.165BTC = 38EUR, so the additional income more than makes up for the additional power draw.

[MrTeal]

You are using the bulk 15 unit price by SPT.  I'm sure if you PM Bitmain, they will come up with a discount bulk price for 30 units the same way they come up with a bulk price for 52 units AND offer free shipping or discounted shipping.  So, just because SPT has a quoted bulk price FOR 15 UNITS does not mean Bitmain will not do a bulk price for 15 or even 30 units.

See what I'm saying?

I do agree with you for the under clocking of the SP20 to the same efficiency as the S5.  However, do not assume Bitmain will not give a discount on shipping or no shipping costs for 30 units.  That's all I'm saying.  Just because SPT actually lists their bulk price with shipping costs ($0.00) for 15 units does not mean Bitmain would do similar with 15 or 30 units.

Yeah, just going by the 18kW limit posed by the question. Tupsu touched on this already, but even in qty 52 Bitmain is still charging $351 per unit, or $376.03 per unit shipped to the US. You'd only be able to use one coupon though, so if you had 30+ coupons you actually wouldn't save anything ordering 52. You might be able to swing a better deal than the single price, but I'm not sure Bitmain would give you a better price on 30 than they offer on 52.

[jermwerty]

This is not true. If  power is limited, buy more SP20

1x SP20  1000GH/s   0.45W/GHs  Can the S5  do this? No.

What you never want to ROI?  Seems to me unless you have expensive power SP20 should be ran hard (1.4-1.6TH) to make up for the initial investment.

I think the beauty of the S5 is its pretty much the SAME efficiency @ 1.1 or 1.3TH.  So run a S5 @ 1.3TH with 650W then compare that to 950W required to get an SP20 to say 1.4TH give or take.

The SP20 can be downclocked to be more efficient sure but I have both the S5 and SP20 and at this point I'm more impressed with the S5.  Simply because I see that roughly 0.5GH/W efficiency @ 1.3TH, which maximizes my return without losing 25-30% of the hashrate just to meet some arbitrary W/GH goal (which varies greatly depending on location overhead and power costs so YMMV of course, to each their own).

[Tupsu]

This is not true. If  power is limited, buy more SP20

1x SP20  1000GH/s   0.45W/GHs  Can the S5  do this? No.

What you never want to ROI?  Seems to me unless you have expensive power SP20 should be ran hard (1.4-1.6TH) to make up for the initial investment.

I think the beauty of the S5 is its pretty much the SAME efficiency @ 1.1 or 1.3TH.  So run a S5 @ 1.3TH with 650W then compare that to 950W required to get an SP20 to say 1.4TH give or take.

The SP20 can be downclocked to be more efficient sure but I have both the S5 and SP20 and at this point I'm more impressed with the S5.  Simply because I see that roughly 0.5GH/W efficiency @ 1.3TH, which maximizes my return without losing 25-30% of the hashrate just to meet some arbitrary W/GH goal (which varies greatly depending on location overhead and power costs so YMMV of course, to each their own).

I won this  SP20 miner. https://bitcointalk.org/index.php?topic=789369.0
cost me only 40USD VAT,  so I can use him to efficacy tests

[MrTeal]

This is not true. If  power is limited, buy more SP20

1x SP20  1000GH/s   0.45W/GHs  Can the S5  do this? No.

What you never want to ROI?  Seems to me unless you have expensive power SP20 should be ran hard (1.4-1.6TH) to make up for the initial investment.

I think the beauty of the S5 is its pretty much the SAME efficiency @ 1.1 or 1.3TH.  So run a S5 @ 1.3TH with 650W then compare that to 950W required to get an SP20 to say 1.4TH give or take.

The SP20 can be downclocked to be more efficient sure but I have both the S5 and SP20 and at this point I'm more impressed with the S5.  Simply because I see that roughly 0.5GH/W efficiency @ 1.3TH, which maximizes my return without losing 25-30% of the hashrate just to meet some arbitrary W/GH goal (which varies greatly depending on location overhead and power costs).

That's an interesting argument. I would say that's actually a point in favor of the SP20 actually. If you can get the same (or essentially the same) hashrate and efficiency around the 0.5-0.55 point, but the SP20 allows you to run faster if your hosting and power costs dictate that doing so would be more profitable, that's a plus in my books. If both units are the same cost I'd say the SP20 is the better choice at this juncture*.
The S5 won't get 1.3TH/s at 650W though, unless your unit is significantly better than mine are.

*In min qty 3. The SP20 is harder to justify at the single unit price + shipping.

[pekatete]

Just putting this out there, would you not get better efficiency if you run the S5 off 2 PSU's, e.g a 550W PSU for each blade, than run the whole rig off a single 750W/850W/1100W PSU?

[MrTeal]

Just putting this out there, would you not get better efficiency if you run the S5 off 2 PSU's, e.g a 550W for each blade than run the whole rig off a single 750W/850W/1100W PSU?

Depends. Generally within the same product line a 550W running at 50% load will be about as efficient as a 1100W running at the same 50% load. More practically, the 550W units would tend to be easier to find and cheaper, but often the cheap ones you find aren't as efficient or well built as the 1100W units. You then also have another full ATX supply to deal with, and if you have more than a couple of these it gets to be a pain. I actually thing the sweet spot is probably 750-850W. You can get decent well built gold PSUs in that range dirt cheap, running a 750W PSU at 550W (DC) won't really get it into the bad part of it's efficiency curve, and they're still super easy to source.

[pekatete]

Just putting this out there, would you not get better efficiency if you run the S5 off 2 PSU's, e.g a 550W for each blade than run the whole rig off a single 750W/850W/1100W PSU?

Depends. Generally within the same product line a 550W running at 50% load will be about as efficient as a 1100W running at the same 50% load. More practically, the 550W units would tend to be easier to find and cheaper, but often the cheap ones you find aren't as efficient or well built as the 1100W units. You then also have another full ATX supply to deal with, and if you have more than a couple of these it gets to be a pain. I actually thing the sweet spot is probably 750-850W. You can get decent well built gold PSUs in that range dirt cheap, running a 750W PSU at 550W (DC) won't really get it into the bad part of it's efficiency curve, and they're still super easy to source.

1. Throw into that mix the 550W PSU being a server one - smaller (yes), less efficient than a 1100W (no)
2. If we get to the sweet spot being between 750W-850W for the S5, then I can not see how either would be more efficient than 2 550W PSU's, assuming same product line.
3. Then there will be price, I know you can pick up a multi rail 750W PSU for around GBP 50.00 (could be cheaper depending), but you can also pick up a single rail 550W server PSU for GBP 10.00
I think it is both cheaper and more efficient to run an S5 that way .... I am just not sure how the draw at the wall would differ (and that's why I put it out here, been tumbling about in my head for a while now!)

[jermwerty]

The S5 won't get 1.3TH/s at 650W though, unless your unit is significantly better than mine are.

Hmm I get 650W (119V) on my KillAWatt running 1245GH on a 1000W LEPA Bronze supply with 375mhz.  I did briefly run 390/400mhz with a Gold supply and thought I was averaging about 1.3T/650W but maybe I'll go back and double check...

Thats what I don't get all the SP20 love here.  It takes 960W @ the wall for 1.4/1.45TH on my SP20s.  So I feel like 300W to gain 100/150GH is just a bad deal on the SP20.  Heck thats nearly 50% more power required!

I dunno I just haven't touched my S5s in two weeks, and they are solid.  The amount of messing-around I have done on the SP20E and the number of firmware updates/fluctuating GHs numbers, etc -- well lets just say it reminds me of the early Jups and all the tweaking that was required to get those big chips all happy.  

[dogie]

Just putting this out there, would you not get better efficiency if you run the S5 off 2 PSU's, e.g a 550W for each blade than run the whole rig off a single 750W/850W/1100W PSU?

Depends. Generally within the same product line a 550W running at 50% load will be about as efficient as a 1100W running at the same 50% load. More practically, the 550W units would tend to be easier to find and cheaper, but often the cheap ones you find aren't as efficient or well built as the 1100W units. You then also have another full ATX supply to deal with, and if you have more than a couple of these it gets to be a pain. I actually thing the sweet spot is probably 750-850W. You can get decent well built gold PSUs in that range dirt cheap, running a 750W PSU at 550W (DC) won't really get it into the bad part of it's efficiency curve, and they're still super easy to source.

Between 50-85% there is usually low single digit efficiency differences. Not enough to care about really.

[dogie]

seems they want to price in BTC but adjust the USD price manually. Not sure why they do tha

You've seen what happens when the price changes once while someone is mentally locked into buying, imagine what would happen if that price kept changing every 2 minutes?

this is economics 101, but seems to go over yours and BMT head: stable(r) price entices MORE buyers, unless buying is completely hopeless.
yes, you could have 10% fatter margins if you jerk up the price, but you would have 30-50% less customers.
Besides, if the price was moved up in response to BTC/$ ratio change, why it did not went down when BTC went from $310 to $250 this am?
Answer this, please.

That's not what I was talking about.

[HerbPean]

The S5 won't get 1.3TH/s at 650W though, unless your unit is significantly better than mine are.

Hmm I get 650W (119V) on my KillAWatt running 1245GH on a 1000W LEPA Bronze supply with 375mhz.  I did briefly run 390/400mhz with a Gold supply and thought I was averaging about 1.3T/650W but maybe I'll go back and double check...

Thats what I don't get all the SP20 love here.  It takes 960W @ the wall for 1.4/1.45TH on my SP20s.  So I feel like 300W to gain 100/150GH is just a bad deal on the SP20.  Heck thats nearly 50% more power required!

I dunno I just haven't touched my S5s in two weeks, and they are solid.  The amount of messing-around I have done on the SP20E and the number of firmware updates/fluctuating GHs numbers, etc -- well lets just say it reminds me of the early Jups and all the tweaking that was required to get those big chips all happy.  

Well I prefer to push my SP20 harder now and lower then later (diff jump). It always depends on so much factor ... for me electricity is cheap, so it's a no-brainer (~1100 watt for like 1560GH)

[Tupsu]

....
The S5 won't get 1.3TH/s at 650W though, unless your unit is significantly better than mine are.
...

My water-cooled S5 with selected miner blades works.with EVGA G2 1300  
1295W    2580GH/s   0.502W/GHs


I have to try a more powerful power supply unit.

[pekatete]

Between 50-85% there is usually low single digit efficiency differences. Not enough to care about really.

Do not feel obliged to comment. I already got a sensible response.

[pekatete]

My water-cooled S5 with selected miner blades works.with EVGA G2 1300  
1295W    2580GH/s   0.502W/GHs


I have to try a more powerful power supply unit.

What frequency are you running that thing at? Also, I cannot see you having more than 2 blades, but I may be wrong, how many blades are those?

2580GH/s .... ? Care to post a screenshot of the webUI?

[MrTeal]

Just putting this out there, would you not get better efficiency if you run the S5 off 2 PSU's, e.g a 550W for each blade than run the whole rig off a single 750W/850W/1100W PSU?

Depends. Generally within the same product line a 550W running at 50% load will be about as efficient as a 1100W running at the same 50% load. More practically, the 550W units would tend to be easier to find and cheaper, but often the cheap ones you find aren't as efficient or well built as the 1100W units. You then also have another full ATX supply to deal with, and if you have more than a couple of these it gets to be a pain. I actually thing the sweet spot is probably 750-850W. You can get decent well built gold PSUs in that range dirt cheap, running a 750W PSU at 550W (DC) won't really get it into the bad part of it's efficiency curve, and they're still super easy to source.

1. Throw into that mix the 550W PSU being a server one - smaller (yes), less efficient than a 1100W (no)
2. If we get to the sweet spot being between 750W-850W for the S5, then I can not see how either would be more efficient than 2 550W PSU's, assuming same product line.
3. Then there will be price, I know you can pick up a multi rail 750W PSU for around GBP 50.00 (could be cheaper depending), but you can also pick up a single rail 550W server PSU for GBP 10.00
I think it is both cheaper and more efficient to run an S5 that way .... I am just not sure how the draw at the wall would differ (and that's why I put it out here, been tumbling about in my head for a while now!)

1. Not sure, it depends on the server PSU. There's a lot of 550W PSUs on eBay that are pulls from the mid-2000s, and they often aren't necessarily any more efficient than an ATX PSU.
2. They wouldn't be more efficient, but they wouldn't be much less efficient. I saw the sweet spot because you then get away from having to use two PSUs with the corresponding extra volume and cable management, and the price per watt is still good. Once you go up over 1000W, the price per watt goes up dramatically.

For example, you can go get a Rosewill Capstone 750W gold supply at Newegg for $90 shipped in quantity, no rebates or other crap. It'll still do 90% efficiency at 550W (DC) output. You could pick up two 550W supplies and be right in the 50% part of the curve, but any two supplies you get for $90 will probably do worse than the 750W supply anyway since you'd really have to shop around or dick around with buying one at a time and using MIRs to get anything other than a Bronze PSU under $50.

3. Server PSUs are a who other matter. I wouldn't even bother with a 550W server PSU, since you can find 800W server PSUs for that price just as easily. You just need to be willing to do some extra work to get them working.