ANTMINER S3+ Discussion and Support Thread - page 506.

asnghost

sr. member

Activity: 258

Merit: 250

Quote from: oskuro on July 03, 2014, 02:10:16 AM

Just to know if this S3 ROI or not, made simple maths:

lets imagine i buy 50 S3.

50 S3 = 23900 GH/s without overclock
25 pairs x 1,48 BTC = 37 BTC = 23888.57 $

And for power costs, i calculate around 34 $ / S3 without overclock / month(costs 0.18 $kwh). So 34 x 50 = 1700 $ / month all together.

Right now, if you have 23.900 GH/s you earn / month --> 14083.18 $

So imagine:

14.083$(earnings) - 1.700$(power) = 12.383 $ / month.

We have difficulty increases, yeah, but right now, i see next difficulty will be 15998489889.0 / -4.9% [est.], is that correct??. Difficulty decrease?. Anyways, i think S3 ROI, but of course im just a noob here and its just my simple maths Grin

I dont have to buy PSU also, as i have 2 corsair CX750M from my S1´s

https://bitcoinwisdom.com/bitcoin/difficulty
Estimated Next Difficulty: 18,740,225,207 (+11.43%)
Adjust time: After 1464 Blocks, About 10.3 days

oskuro

sr. member

Activity: 294

Merit: 251

Just to know if this S3 ROI or not, made simple maths:

lets imagine i buy 50 S3.

50 S3 = 23900 GH/s without overclock
25 pairs x 1,48 BTC = 37 BTC = 23888.57 $

And for power costs, i calculate around 34 $ / S3 without overclock / month(costs 0.18 $kwh). So 34 x 50 = 1700 $ / month all together.

Right now, if you have 23.900 GH/s you earn / month --> 14083.18 $

So imagine:

14.083$(earnings) - 1.700$(power) = 12.383 $ / month.

We have difficulty increases, yeah, but right now, i see next difficulty will be 15998489889.0 / -4.9% [est.], is that correct??. Difficulty decrease?. Anyways, i think S3 ROI, but of course im just a noob here and its just my simple maths Grin

I dont have to buy PSU also, as i have 2 corsair CX750M from my S1´s

bones

sr. member

Activity: 248

Merit: 250

This was how you could power the s1, should work for s3, so long as wired right.
Guessing you would have to rewire though, not just PnP.

Quote from: ZiG on July 03, 2014, 12:28:18 AM

Quote from: phillipsjk on July 02, 2014, 11:43:34 PM

Quote from: ZiG on July 02, 2014, 11:24:36 PM

Quote from: phillipsjk on July 02, 2014, 10:26:55 PM

I am curious: is the cover made out of plastic or metal? It should save on the cost of a case (which is one reason I have been a little reluctant to get an S1 (still haven't payed off my old blade either). Cases block EMI, which helps prevent interference with Over-the-air television.

I was thinking these things just need a 500W power supply: I suspect (but have not tested) that you can plug the 4Pin CPU connectors into the spare ports for an extra safety margin.

CPU connector(s) are different pinouts...Don't even try...

ZiG

OK, you made me try to look it up:

Quote from: Tom's hardware

Caution: The eight-pin PCI Express Auxiliary Power Connector and the eight-pin EPS12V CPU Power Connector use similar Molex Mini-Fit Jr. connector housings. Although they are keyed differently, the keying can be overcome by sufficient force such that you can plug an EPS12V power connector into a graphics card, or a PCI Express power connector into a motherboard. Either of these scenarios results in +12 V being directly shorted to ground, potentially destroying the motherboard, graphics card, or power supply.

-Power Supply 101: A Reference Of Specifications

Forgot that CPU connectors are 8 pin now. Also was not aware that the 6 and 8 pin connectors had "sense" pins.
Further, I compared the PCI-E diagram to the 4 pin diagram on the wikipedia ATX page. The polarities are reversed if you line up the tabs.

Exactly...GOOD JOB, buddy... Grin

I simply did not want to lecture you...You did it yourself...

Cheers,

ZiG

EDIT...:

A guy in Habanero thread already did try this "successfully"... Grin

vulgartrendkill

hero member

Activity: 518

Merit: 500

Anyone know what procedure is for overclocking? I note from other threads about s2 that the overclocking is done by editing the config file while ssh in the miner. Will it be the same?

ZiG

sr. member

Activity: 406

Merit: 250

Quote from: phillipsjk on July 02, 2014, 11:43:34 PM

Quote from: ZiG on July 02, 2014, 11:24:36 PM

Quote from: phillipsjk on July 02, 2014, 10:26:55 PM

I am curious: is the cover made out of plastic or metal? It should save on the cost of a case (which is one reason I have been a little reluctant to get an S1 (still haven't payed off my old blade either). Cases block EMI, which helps prevent interference with Over-the-air television.

I was thinking these things just need a 500W power supply: I suspect (but have not tested) that you can plug the 4Pin CPU connectors into the spare ports for an extra safety margin.

CPU connector(s) are different pinouts...Don't even try...

ZiG

OK, you made me try to look it up:

Quote from: Tom's hardware

Caution: The eight-pin PCI Express Auxiliary Power Connector and the eight-pin EPS12V CPU Power Connector use similar Molex Mini-Fit Jr. connector housings. Although they are keyed differently, the keying can be overcome by sufficient force such that you can plug an EPS12V power connector into a graphics card, or a PCI Express power connector into a motherboard. Either of these scenarios results in +12 V being directly shorted to ground, potentially destroying the motherboard, graphics card, or power supply.

-Power Supply 101: A Reference Of Specifications

Forgot that CPU connectors are 8 pin now. Also was not aware that the 6 and 8 pin connectors had "sense" pins.
Further, I compared the PCI-E diagram to the 4 pin diagram on the wikipedia ATX page. The polarities are reversed if you line up the tabs.

Exactly...GOOD JOB, buddy... Grin

I simply did not want to lecture you...You did it yourself...

Cheers,

ZiG

EDIT...:

A guy in Habanero thread already did try this "successfully"... Grin

pvnamk19

full member

Activity: 178

Merit: 100

Quote from: Gogreen on July 02, 2014, 11:29:26 PM

8 days left with another difficulty change , oh my ?!

The first time from Jan 23 2013, we will enjoy a RED difficulty change or get trapped by ASICs manufacturers!?
Up to now: network hashrate: 114,206,200 GH/s vs 120,391,236 GH/s (last change).

phillipsjk

legendary

Activity: 1008

Merit: 1001

Let the chips fall where they may.

Quote from: ZiG on July 02, 2014, 11:24:36 PM

Quote from: phillipsjk on July 02, 2014, 10:26:55 PM

I am curious: is the cover made out of plastic or metal? It should save on the cost of a case (which is one reason I have been a little reluctant to get an S1 (still haven't payed off my old blade either). Cases block EMI, which helps prevent interference with Over-the-air television.

I was thinking these things just need a 500W power supply: I suspect (but have not tested) that you can plug the 4Pin CPU connectors into the spare ports for an extra safety margin.

CPU connector(s) are different pinouts...Don't even try...

ZiG

OK, you made me try to look it up:

Quote from: Tom's hardware

Caution: The eight-pin PCI Express Auxiliary Power Connector and the eight-pin EPS12V CPU Power Connector use similar Molex Mini-Fit Jr. connector housings. Although they are keyed differently, the keying can be overcome by sufficient force such that you can plug an EPS12V power connector into a graphics card, or a PCI Express power connector into a motherboard. Either of these scenarios results in +12 V being directly shorted to ground, potentially destroying the motherboard, graphics card, or power supply.

-Power Supply 101: A Reference Of Specifications

Forgot that CPU connectors are 8 pin now. Also was not aware that the 6 and 8 pin connectors had "sense" pins.
Further, I compared the PCI-E diagram to the 4 pin diagram on the wikipedia ATX page. The polarities are reversed if you line up the tabs.

Gogreen

sr. member

Activity: 392

Merit: 250

8 days left with another difficulty change , oh my ?!

ZiG

sr. member

Activity: 406

Merit: 250

Quote from: phillipsjk on July 02, 2014, 10:26:55 PM

I am curious: is the cover made out of plastic or metal? It should save on the cost of a case (which is one reason I have been a little reluctant to get an S1 (still haven't payed off my old blade either). Cases block EMI, which helps prevent interference with Over-the-air television.

I was thinking these things just need a 500W power supply: I suspect (but have not tested) that you can plug the 4Pin CPU connectors into the spare ports for an extra safety margin.

CPU connector(s) are different pinouts...Don't even try...

ZiG

phillipsjk

legendary

Activity: 1008

Merit: 1001

Let the chips fall where they may.

I am curious: is the cover made out of plastic or metal? It should save on the cost of a case (which is one reason I have been a little reluctant to get an S1 (still haven't payed off my old blade either). Cases block EMI, which helps prevent interference with Over-the-air television.

I was thinking these things just need a 500W power supply: ~~I suspect (but have not tested) that you can plug the 4Pin CPU connectors into the spare ports for an extra safety margin.~~ Edit: bad idea (as in flames shooting out of one of your expensive components if you manage to force it).

CryptoCrane

sr. member

Activity: 384

Merit: 256

Quote from: r1senfa17h on June 30, 2014, 07:06:47 AM

How many S3s will there be in the first batch?

Edit: Batch 1 units for sale on Amazon. Great if you only want to buy 1 or don't have enough bitcoin.

Hey, thanks for linking to our batch-1 S3s on Amazon! Our primary market are people who want bitcoin but don't feel comfortable with the exchanges and therefore don't already have enough bitcoin to purchase one directly from Bitmain. Cheesy

shaxs

full member

Activity: 168

Merit: 100

Quote from: RealMalatesta on July 01, 2014, 06:06:11 PM

Quote from: iglasses on July 01, 2014, 06:01:24 PM

Quote from: philipma1957 on July 01, 2014, 05:25:26 PM

Quote from: oskuro on July 01, 2014, 05:22:52 PM

Quote from: RealMalatesta on July 01, 2014, 05:19:10 PM

Quote from: data55 on July 01, 2014, 12:42:39 PM

Quote from: GrapeApe on July 01, 2014, 12:35:02 PM

Did I read correctly that it only needs 2 PCI-e cables connected to run as normal (not overclocked)?

Correct. 2 for regular, 4 for OC'ing.

What I do not understand is this sentence: "Don’t parallel connected different DC input from different PSU into the same Hashing board." So does this mean you need two PSU if you're overclocking?

I understand they mean dont use 2 PSU for same hashing board. Anyways, with 1 PSU u can connect 4 PCI-e so no problem

yeah this is most likely exactly what they mean.

but with 2 psu's each one doing one board each even thou it seems okay I will let some one else do it first.

Depending on how the shipping goes I might be the guinea pig here. My understanding is that the reason, the only reason, to use two PSU's is to overclock. If you are not planning to OC one PSU is all that you should need. In my case my PSU army is a mix of Corsair 750's and Thermaltake 850's so if I want to OC I am going to have to use the dual config.

I've still got a couple of these here:

https://bitcointalksearch.org/topic/server-psu-for-antminer-s1-step-by-step-guide-518013

So basically, I could use one for two S3s if not oc'd, correct?

I am in the same boat. I have a bunch of those too. However, since there are no longer screw terminals, I do not know how I will connect them to the S3. How do you plan on doing it?

visdude

legendary

Activity: 1081

Merit: 1001

Quote from: pumaro on July 02, 2014, 02:22:26 PM

you should get 4 850 watt psu's your psu's should only be running about 60% to stay in the efficiency range.

The Corsair CX750M (with 4 PCIe connectors) is I think the best deal around (especially with promo code discounts plus rebates) as a dedicated single PSU solution for an overclocked S3. Alternatively, two CX500/CX500M (this would be my power configuration if I decide to overclock since I already got a few of them) would also be a good buy which can be had for as low as $19.99 if timed with regular promo code discount offerings up to $20 plus $20 rebate at the Egg. In fact, I was able to get some of my CX500s just by applying the prepaid rebate cards from past purchases...and so on and so forth

. Such a solution would offer plenty of headroom (a total of 1000W) at a dirt cheap price and modularity (i.e. presumably, if one PSU goes out, only one blade would be out of commission with the other continuing to hash).

Current deals at the Egg:

CX750M - $20 rebate
CX500M - $15 promo code discount plus $20 rebate
CX500 - none

iglasses

legendary

Activity: 1148

Merit: 1000

Maybe you guys are joking and my sarcasm radar is out of tune but if you are serious I think that's a pretty wild accusation. If Guild, or any pool for that matter, was posting results that didn't jive with what was in the blockchain don't you think someone would figure it out? BTC has a lot of really smart, dedicated people and if that isn't enough it also has people with tens of thousands of dollars invested who pay attention to stuff like that.
JMHO.

ig

Threader

sr. member

Activity: 406

Merit: 250

Quote from: theonegilly on July 02, 2014, 03:40:29 PM

Quote from: pumaro on July 02, 2014, 03:20:45 PM

Quote from: dropt on July 02, 2014, 03:14:09 PM

Quote from: hurricandave on July 02, 2014, 03:09:22 PM

I see post after post where mathematicians calculate profit or loss based on speed and difficulty, but the biggest factor that will get in the way is LUCK, the POOL LUCK has been destroyed by manufacturers and the private farms. Bitmain is a contributor to this as well. The difficulty is rising in leaps and bounds based on network speed but most of the public pools combined hash rate is getting left in the dust! Oh except when Bitmain or one of the other Manufacturers decides to raid a pool by jumping in with a Ph/s or two for a couple of hours while "testing". Crunch the numbers all you want, till you start adding in the Variance and LUCK of your favorite pool, those calculators are meaningless.

Yes, I too have been raped hard by luck over the last three months, but this is independent of hashrate distribution. Slush has had a 1 month luck of 107% while BTCGuild has been at 91.9% even though BTCGuild is 2x-3x larger than Slush by HR.

Its all Luck some good some bad but it all averages out, although SLUSH and ELIGIUS seem to payout better than BTCGuild, maybe they are just inherently unlucky.

Maybe BTC Guild is actually running at 101.1% and 1 of every 10 blocks goes straight to the Owner.

Who know.

Sure I have been guilty of thinking this a few times during shitty luck days

theonegilly

member

Activity: 106

Merit: 10

Your Pool Your Way - Admin

Quote from: pumaro on July 02, 2014, 03:20:45 PM

Quote from: dropt on July 02, 2014, 03:14:09 PM

Quote from: hurricandave on July 02, 2014, 03:09:22 PM

I see post after post where mathematicians calculate profit or loss based on speed and difficulty, but the biggest factor that will get in the way is LUCK, the POOL LUCK has been destroyed by manufacturers and the private farms. Bitmain is a contributor to this as well. The difficulty is rising in leaps and bounds based on network speed but most of the public pools combined hash rate is getting left in the dust! Oh except when Bitmain or one of the other Manufacturers decides to raid a pool by jumping in with a Ph/s or two for a couple of hours while "testing". Crunch the numbers all you want, till you start adding in the Variance and LUCK of your favorite pool, those calculators are meaningless.

Yes, I too have been raped hard by luck over the last three months, but this is independent of hashrate distribution. Slush has had a 1 month luck of 107% while BTCGuild has been at 91.9% even though BTCGuild is 2x-3x larger than Slush by HR.

Its all Luck some good some bad but it all averages out, although SLUSH and ELIGIUS seem to payout better than BTCGuild, maybe they are just inherently unlucky.

Maybe BTC Guild is actually running at 101.1% and 1 of every 10 blocks goes straight to the Owner.

Who know.

pumaro

member

Activity: 78

Merit: 10

Quote from: dropt on July 02, 2014, 03:14:09 PM

Quote from: hurricandave on July 02, 2014, 03:09:22 PM

I see post after post where mathematicians calculate profit or loss based on speed and difficulty, but the biggest factor that will get in the way is LUCK, the POOL LUCK has been destroyed by manufacturers and the private farms. Bitmain is a contributor to this as well. The difficulty is rising in leaps and bounds based on network speed but most of the public pools combined hash rate is getting left in the dust! Oh except when Bitmain or one of the other Manufacturers decides to raid a pool by jumping in with a Ph/s or two for a couple of hours while "testing". Crunch the numbers all you want, till you start adding in the Variance and LUCK of your favorite pool, those calculators are meaningless.

Yes, I too have been raped hard by luck over the last three months, but this is independent of hashrate distribution. Slush has had a 1 month luck of 107% while BTCGuild has been at 91.9% even though BTCGuild is 2x-3x larger than Slush by HR.

Its all Luck some good some bad but it all averages out, although SLUSH and ELIGIUS seem to payout better than BTCGuild, maybe they are just inherently unlucky.

hurricandave

legendary

Activity: 966

Merit: 1003

Quote from: iglasses on July 02, 2014, 03:12:29 PM

Quote from: hurricandave on July 02, 2014, 03:09:22 PM

I see post after post where mathematicians calculate profit or loss based on speed and difficulty, but the biggest factor that will get in the way is LUCK, the POOL LUCK has been destroyed by manufacturers and the private farms. Bitmain is a contributor to this as well. The difficulty is rising in leaps and bounds based on network speed but most of the public pools combined hash rate is getting left in the dust! Oh except when Bitmain or one of the other Manufacturers decides to raid a pool by jumping in with a Ph/s or two for a couple of hours while "testing". Crunch the numbers all you want, till you start adding in the Variance and LUCK of your favorite pool, those calculators are meaningless.

Well now we have really jumped the shark. Unless we calculate LUCK the math is unreliable?
ugh.

The guys at bitminter just went over 24hrs without a block, where is that in your calculations. The pool returns are diminishing and always have been, but we are at a tipping point where the drops are nearing an exponential loss for the public if the manufacturers don't back off the network and get the gear distributed to public miners.

dropt

legendary

Activity: 1512

Merit: 1000

Quote from: hurricandave on July 02, 2014, 03:09:22 PM

I see post after post where mathematicians calculate profit or loss based on speed and difficulty, but the biggest factor that will get in the way is LUCK, the POOL LUCK has been destroyed by manufacturers and the private farms. Bitmain is a contributor to this as well. The difficulty is rising in leaps and bounds based on network speed but most of the public pools combined hash rate is getting left in the dust! Oh except when Bitmain or one of the other Manufacturers decides to raid a pool by jumping in with a Ph/s or two for a couple of hours while "testing". Crunch the numbers all you want, till you start adding in the Variance and LUCK of your favorite pool, those calculators are meaningless.

Yes, I too have been raped hard by luck over the last three months, but this is independent of hashrate distribution. Slush has had a 1 month luck of 107% while BTCGuild has been at 91.9% even though BTCGuild is 2x-3x larger than Slush by HR.

iglasses

legendary

Activity: 1148

Merit: 1000

Quote from: hurricandave on July 02, 2014, 03:09:22 PM

I see post after post where mathematicians calculate profit or loss based on speed and difficulty, but the biggest factor that will get in the way is LUCK, the POOL LUCK has been destroyed by manufacturers and the private farms. Bitmain is a contributor to this as well. The difficulty is rising in leaps and bounds based on network speed but most of the public pools combined hash rate is getting left in the dust! Oh except when Bitmain or one of the other Manufacturers decides to raid a pool by jumping in with a Ph/s or two for a couple of hours while "testing". Crunch the numbers all you want, till you start adding in the Variance and LUCK of your favorite pool, those calculators are meaningless.

Well now we have really jumped the shark. Unless we calculate LUCK the math is unreliable?
ugh.

Topic: ANTMINER S3+ Discussion and Support Thread - page 506. (Read 710258 times)