Topic: ANN: BITMAIN has Tested Its 28nm Bitcoin Mining Chip BM1382 - page 102. (Read 143992 times)
and what about upgrades for S1? When are we going to know how much is that?
and just looking at the way its worded it may already be overclocked.. it looks like it can become more efficient if we underclock or undervolt the machine... ......"consuming less than 9.33 W on chip level at 0.75V. This means 0.59 J/GH on chip. If the core voltage is at 0.63V, the efficiency will be 0.40 J/GH on chip"
I will probably get the S1 upgrade kit instead 
Let's guess the HR of overclocked S3, 520GH/s for me 
My guess is 575GH/s Let's guess the HR of overclocked S3, 520GH/s for me ..."BM1382 has achieved 15.75 Gh/s in hash performance, consuming less than 9.33 W on chip level at 0.75V"...
Let's do some math...:
9.33W / 0.75V = 12.44Amps ...per chip...
16 chips per blade * 12.5Amps = 200Amps per blade ...
NEEDS hell-of -a-POWER Section on the blades...
I do NOT expect that BitMain will design S3 Power section with a lot of reserves...if any...
IMHO overclocking S3 will be a challenge...if possible at all...1-2-3% max...
Not even taking PSU in consideration...
ZiG
My goal is pencil mod undervol underclock getting 450gh at 300 watts which is .667 watts a gh vs .78 watts a gh.
If this can be done they will be really quiet and truly good spaceheaters. I will put I in each room for the winter my house will be warm and toasty. And difficulty could be at 110 vs the current 13.4 and they would still be worth running.
If there is that many chips 16 per hash board then how can 490 watts ever supply 200 amps @ 12v?
?Thats 200 Amps @0.75V not @12V
As noted, the initial calculations are wrong above. My 850W PS that has a single rail that supplies 70 amps at 12V will power 2 S3s if the 390 W spec from Bitmain is correct (which I'm sure it is). Amperage from the wall at 120 V will be less than 8 amps. A common 20 amp/120 V circuit will easily power 4 S3s.
Good catch:) Thanks
I will probably get the S1 upgrade kit instead Let's guess the HR of overclocked S3, 520GH/s for me
My guess is 575GH/s Let's guess the HR of overclocked S3, 520GH/s for me ..."BM1382 has achieved 15.75 Gh/s in hash performance, consuming less than 9.33 W on chip level at 0.75V"...
Let's do some math...:
9.33W / 0.75V = 12.44Amps ...per chip...
16 chips per blade * 12.5Amps = 200Amps per blade ...
NEEDS hell-of -a-POWER Section on the blades...
I do NOT expect that BitMain will design S3 Power section with a lot of reserves...if any...
IMHO overclocking S3 will be a challenge...if possible at all...1-2-3% max...
Not even taking PSU in consideration...
ZiG
My goal is pencil mod undervol underclock getting 450gh at 300 watts which is .667 watts a gh vs .78 watts a gh.
If this can be done they will be really quiet and truly good spaceheaters. I will put I in each room for the winter my house will be warm and toasty. And difficulty could be at 110 vs the current 13.4 and they would still be worth running.
If there is that many chips 16 per hash board then how can 490 watts ever supply 200 amps @ 12v?
?Thats 200 Amps @0.75V not @12V
I will probably get the S1 upgrade kit instead Let's guess the HR of overclocked S3, 520GH/s for me
My guess is 575GH/s Let's guess the HR of overclocked S3, 520GH/s for me ..."BM1382 has achieved 15.75 Gh/s in hash performance, consuming less than 9.33 W on chip level at 0.75V"...
Let's do some math...:
9.33W / 0.75V = 12.44Amps ...per chip...
16 chips per blade * 12.5Amps = 200Amps per blade ...
NEEDS hell-of -a-POWER Section on the blades...
I do NOT expect that BitMain will design S3 Power section with a lot of reserves...if any...
IMHO overclocking S3 will be a challenge...if possible at all...1-2-3% max...
Not even taking PSU in consideration...
ZiG
My goal is pencil mod undervol underclock getting 450gh at 300 watts which is .667 watts a gh vs .78 watts a gh.
If this can be done they will be really quiet and truly good spaceheaters. I will put I in each room for the winter my house will be warm and toasty. And difficulty could be at 110 vs the current 13.4 and they would still be worth running.
If there is that many chips 16 per hash board then how can 490 watts ever supply 200 amps @ 12v?
?Edit: Its obviously apparent that we are missing something here.. it cant possibly be that many amps...the s1 only pulls around 35 amps..
I don't understand all the love you guys show for Bitmain. Yes, they have delivered. No, their price has been reasonable, but not good for us miners.
Donno, my last few S1s that haven't reached breakeven will do so very soon (and that's with paying Swedish VAT on them) My first batch are sitting at around 20-25% profit in BTC (what more could you reasonably ask for?)
Rograz, now that Bitmain has become an established company due to the support of miners and what appears to be a good business plan. I think it would be reasonable to ask for 35-45% RoI for miners(market dependent and before VAT).
I suspect based off of the miners I have received, and posts I have read of other forum users that Bitmain doesn't just stress test(QC) units before they ship but actually uses them until they ship, then ship them as 'new' at a 20-50%(decreasing as the hash rate increases) markup on their manufacturing cost. Thus netting them miner and markup(double return) of 60-80%(guesstimate). Granted this is just speculation(like most of this thread) but it is by no means a crazy leap of logic.
I also don't begrudge them this business model, I think its a smart way to RoI, however I would like to see slightly more on our(miners) end of things.
Well at times they offer coupons and the s-1's were a really good buy some of the time. The hardest thing for them is pricing correctly. I never felt the s-2 was worth buying so I never purchased one. I purchased 16 s-1's 3 I overpaid looking back with
hindsight but 13 of them were a fair price.
Based on rk-box at $599 these would be killer at $599. All they need to do is price match the rk-box since it appears that this is a better piece of gear. Have to wait and see.
Agreed Philip, also they have to price the S3 in BTC not fiat like they did the S2. Basically price it the same way they did the S1, with the price cuts here and there, and they'll have another winner on their hands.
I don't understand all the love you guys show for Bitmain. Yes, they have delivered. No, their price has been reasonable, but not good for us miners.
Donno, my last few S1s that haven't reached breakeven will do so very soon (and that's with paying Swedish VAT on them) My first batch are sitting at around 20-25% profit in BTC (what more could you reasonably ask for?)
Rograz, now that Bitmain has become an established company due to the support of miners and what appears to be a good business plan. I think it would be reasonable to ask for 35-45% RoI for miners(market dependent and before VAT).
I suspect based off of the miners I have received, and posts I have read of other forum users that Bitmain doesn't just stress test(QC) units before they ship but actually uses them until they ship, then ship them as 'new' at a 20-50%(decreasing as the hash rate increases) markup on their manufacturing cost. Thus netting them miner and markup(double return) of 60-80%(guesstimate). Granted this is just speculation(like most of this thread) but it is by no means a crazy leap of logic.
I also don't begrudge them this business model, I think its a smart way to RoI, however I would like to see slightly more on our(miners) end of things.
Well at times they offer coupons and the s-1's were a really good buy some of the time. The hardest thing for them is pricing correctly. I never felt the s-2 was worth buying so I never purchased one. I purchased 16 s-1's 3 I overpaid looking back with
hindsight but 13 of them were a fair price.
Based on rk-box at $599 these would be killer at $599. All they need to do is price match the rk-box since it appears that this is a better piece of gear. Have to wait and see.
I don't understand all the love you guys show for Bitmain. Yes, they have delivered. No, their price has been reasonable, but not good for us miners.
Donno, my last few S1s that haven't reached breakeven will do so very soon (and that's with paying Swedish VAT on them) My first batch are sitting at around 20-25% profit in BTC (what more could you reasonably ask for?)
Rograz, now that Bitmain has become an established company due to the support of miners and what appears to be a good business plan. I think it would be reasonable to ask for 35-45% RoI for miners(market dependent and before VAT).
I suspect based off of the miners I have received, and posts I have read of other forum users that Bitmain doesn't just stress test(QC) units before they ship but actually uses them until they ship, then ship them as 'new' at a 20-50%(decreasing as the hash rate increases) markup on their manufacturing cost. Thus netting them miner and markup(double return) of 60-80%(guesstimate). Granted this is just speculation(like most of this thread) but it is by no means a crazy leap of logic.
I also don't begrudge them this business model, I think its a smart way to RoI, however I would like to see slightly more on our(miners) end of things.
I will probably get the S1 upgrade kit instead Let's guess the HR of overclocked S3, 520GH/s for me
My guess is 575GH/s Let's guess the HR of overclocked S3, 520GH/s for me ..."BM1382 has achieved 15.75 Gh/s in hash performance, consuming less than 9.33 W on chip level at 0.75V"...
Let's do some math...:
9.33W / 0.75V = 12.44Amps ...per chip...
16 chips per blade * 12.5Amps = 200Amps per blade ...
NEEDS hell-of -a-POWER Section on the blades...
I do NOT expect that BitMain will design S3 Power section with a lot of reserves...if any...
IMHO overclocking S3 will be a challenge...if possible at all...1-2-3% max...
Not even taking PSU in consideration...
ZiG
My goal is pencil mod undervol underclock getting 450gh at 300 watts which is .667 watts a gh vs .78 watts a gh.
If this can be done they will be really quiet and truly good spaceheaters. I will put I in each room for the winter my house will be warm and toasty. And difficulty could be at 110 vs the current 13.4 and they would still be worth running.
I don't understand all the love you guys show for Bitmain. Yes, they have delivered. No, their price has been reasonable, but not good for us miners. Most likely, they are going to price S3 at over 1 BTC, I'd say around 1.6 BTC or so. Why would they undercut their S2 models by selling S3 at 1 BTC or lower? It makes no sense. Don't be a fanboi.
They are abandoning the S2 model.. that's why. Going to a new chip.
S2 was mistake. If you compare s1(how they were build) to s2 you can clearly see that s2 was just done in rush. They realise that it's not even near of s1 so they abandoned that. upgrade kit probably will be a new hashing cards with new chip.
I don't understand all the love you guys show for Bitmain. Yes, they have delivered. No, their price has been reasonable, but not good for us miners. Most likely, they are going to price S3 at over 1 BTC, I'd say around 1.6 BTC or so. Why would they undercut their S2 models by selling S3 at 1 BTC or lower? It makes no sense. Don't be a fanboi.
They are abandoning the S2 model.. that's why. Going to a new chip.
I will probably get the S1 upgrade kit instead Let's guess the HR of overclocked S3, 520GH/s for me
My guess is 575GH/s Let's guess the HR of overclocked S3, 520GH/s for me 575gh is gonna use quite a bit of power lol
Not much more then my S1s running at 200
Im using a CX750 corsair running my overclocked S1 @ 200gh. surely it's not that high. Im disappointed that S3 is not based on S2, such as psu included
Not me. That big heavy box and PSU (which apparently wasn't all that great anyway) def. added to the shipping cost. If the S3's will run on my existing PSU fleet that's okayfine with me. I don't know about you guys but I have already spent over a thousand dollars on PSUs and am happy to have an opportunity to increase my hashrate without necessarily having to buy more PSUs.
Your results may vary
Happy Mining
What's up with the obsession?
I play with my antminers and it's fun to optimize the overclocking, maybe soon undervolt and so on. But I'm at a loss will probably never break even.
Instead of losing money on the S3 I will spend more money on dinners, holidays etc.
Investing in ASICs is for those who have
1) no customs/vat
and
2) cheap ~$0.05/kWh electricity
I play with my antminers and it's fun to optimize the overclocking, maybe soon undervolt and so on. But I'm at a loss will probably never break even.
Instead of losing money on the S3 I will spend more money on dinners, holidays etc.
Investing in ASICs is for those who have
1) no customs/vat
and
2) cheap ~$0.05/kWh electricity
I will probably get the S1 upgrade kit instead Let's guess the HR of overclocked S3, 520GH/s for me
My guess is 575GH/s Let's guess the HR of overclocked S3, 520GH/s for me ..."BM1382 has achieved 15.75 Gh/s in hash performance, consuming less than 9.33 W on chip level at 0.75V"...
Let's do some math...:
9.33W / 0.75V = 12.44Amps ...per chip...
16 chips per blade * 12.5Amps = 200Amps per blade ...
NEEDS hell-of -a-POWER Section on the blades...
I do NOT expect that BitMain will design S3 Power section with a lot of reserves...if any...
IMHO overclocking S3 will be a challenge...if possible at all...1-2-3% max...
Not even taking PSU in consideration...
ZiG
12.5 amps per chip isn't right. It's 9.33 watt for all 16 chips.
Nope, ZiG is correct:
BM1382 has achieved 15.75 Gh/s in hash performance, consuming less than 9.33 W on chip level at 0.75V. This means 0.59 J/GH on chip.
I will probably get the S1 upgrade kit instead Let's guess the HR of overclocked S3, 520GH/s for me
My guess is 575GH/s Let's guess the HR of overclocked S3, 520GH/s for me ..."BM1382 has achieved 15.75 Gh/s in hash performance, consuming less than 9.33 W on chip level at 0.75V"...
Let's do some math...:
9.33W / 0.75V = 12.44Amps ...per chip...
16 chips per blade * 12.5Amps = 200Amps per blade ...
NEEDS hell-of -a-POWER Section on the blades...
I do NOT expect that BitMain will design S3 Power section with a lot of reserves...if any...
IMHO overclocking S3 will be a challenge...if possible at all...1-2-3% max...
Not even taking PSU in consideration...
ZiG
S3 won't be hard underclocked/undervolted like S2, imho it has a lot of hidden potential in overclock that will be discovered by users. (Since bitmain has a large farm of miners is more interested to maximize energy efficiency, at the expense of a minor hashrate).
I will probably get the S1 upgrade kit instead Let's guess the HR of overclocked S3, 520GH/s for me
My guess is 575GH/s Let's guess the HR of overclocked S3, 520GH/s for me ..."BM1382 has achieved 15.75 Gh/s in hash performance, consuming less than 9.33 W on chip level at 0.75V"...
Let's do some math...:
9.33W / 0.75V = 12.44Amps ...per chip...
16 chips per blade * 12.5Amps = 200Amps per blade ...
NEEDS hell-of -a-POWER Section on the blades...
I do NOT expect that BitMain will design S3 Power section with a lot of reserves...if any...
IMHO overclocking S3 will be a challenge...if possible at all...1-2-3% max...
Not even taking PSU in consideration...
ZiG
I agree, i don't believe there will be much overclocking potential in the S3. What i think we will see with the S3 is undervolting potential, and when the price falls we may see it comparable in $/GH and W/GH to the SP10 and SP30
I will probably get the S1 upgrade kit instead Let's guess the HR of overclocked S3, 520GH/s for me
My guess is 575GH/s Let's guess the HR of overclocked S3, 520GH/s for me ..."BM1382 has achieved 15.75 Gh/s in hash performance, consuming less than 9.33 W on chip level at 0.75V"...
Let's do some math...:
9.33W / 0.75V = 12.44Amps ...per chip...
16 chips per blade * 12.5Amps = 200Amps per blade ...
NEEDS hell-of -a-POWER Section on the blades...
I do NOT expect that BitMain will design S3 Power section with a lot of reserves...if any...
IMHO overclocking S3 will be a challenge...if possible at all...1-2-3% max...
Not even taking PSU in consideration...
ZiG
I'm no math wiz or electrical engineer. Im going on S1 terms of 10% if you can OC this sucker 10% and still get 1w/gh that's pretty damn good
I will probably get the S1 upgrade kit instead Let's guess the HR of overclocked S3, 520GH/s for me
My guess is 575GH/s Let's guess the HR of overclocked S3, 520GH/s for me 575gh is gonna use quite a bit of power lol
Not much more then my S1s running at 200
Im using a CX750 corsair running my overclocked S1 @ 200gh. surely it's not that high. Im disappointed that S3 is not based on S2, such as psu included
I will probably get the S1 upgrade kit instead Let's guess the HR of overclocked S3, 520GH/s for me
My guess is 575GH/s Let's guess the HR of overclocked S3, 520GH/s for me ..."BM1382 has achieved 15.75 Gh/s in hash performance, consuming less than 9.33 W on chip level at 0.75V"...
Let's do some math...:
9.33W / 0.75V = 12.44Amps ...per chip...
16 chips per blade * 12.5Amps = 200Amps per blade ...
NEEDS hell-of -a-POWER Section on the blades...
I do NOT expect that BitMain will design S3 Power section with a lot of reserves...if any...
IMHO overclocking S3 will be a challenge...if possible at all...1-2-3% max...
Not even taking PSU in consideration...
ZiG
Jump to: