antminer s6 4500gh/s 1000watts

[VirosaGITS]

Thanks Rich.  I should have took a look for myself at their website.  However, I really don't care for SPT anymore unless they change their ways.

There were rumor of them coming up with a smaller unit, but yeah this is insane. This is basically a small furnace. I guess its efficient in space management, but thats probably not what i would want to setup in a warehouse dedicated to ASIC mining either.

[1713303873]

1713303873

[1713303873]

1713303873

[1713303873]

1713303873

[wlefever]

Thanks Rich.  I should have took a look for myself at their website.  However, I really don't care for SPT anymore unless they change their ways.

There were rumor of them coming up with a smaller unit, but yeah this is insane. This is basically a small furnace. I guess its efficient in space management, but thats probably not what i would want to setup in a warehouse dedicated to ASIC mining either.

Just setting up my new SP50 I mean cracking furnace. This will be the next set up lol

[notlist3d]

Except that the 10 1600w PSU. are Serevr style hot pluggable and are built into the unit.



Rich

Thanks Rich.  I should have took a look for myself at their website.  However, I really don't care for SPT anymore unless they change their ways.

Were lucky that I don't think they have chips yet.  If they did imagine how fast they could make difficulty go up  (added onto high already). 

The SP50 is just scary specs with all that.  The good news is it might be slow on making they don't seem to be selling them to many, I think anyone getting one currently is under NDA or they don't have any sold.  So I'm guessing NDA.

[VirosaGITS]

Except that the 10 1600w PSU. are Serevr style hot pluggable and are built into the unit.

Rich

Thanks Rich.  I should have took a look for myself at their website.  However, I really don't care for SPT anymore unless they change their ways.

Were lucky that I don't think they have chips yet.  If they did imagine how fast they could make difficulty go up  (added onto high already). 

The SP50 is just scary specs with all that.  The good news is it might be slow on making they don't seem to be selling them to many, I think anyone getting one currently is under NDA or they don't have any sold.  So I'm guessing NDA.

Dunno, maybe they ARE the reason why the difficulty went super up. Maybe they shipped 100's of units to a private data center. Its not like it would take a lot of these, when 10 of these do over 1PH/s.

[RichBC]

Dunno, maybe they ARE the reason why the difficulty went super up. Maybe they shipped 100's of units to a private data center. Its not like it would take a lot of these, when 10 of these do over 1PH/s.

If the SP50 is alive then all the early units will be going to the 3MW BTCS facility. It's over 2 Months now from the "we are getting ready for deployment" video.

Rich

[Amph]

Except that the 10 1600w PSU. are Serevr style hot pluggable and are built into the unit.

Rich

Thanks Rich.  I should have took a look for myself at their website.  However, I really don't care for SPT anymore unless they change their ways.

Were lucky that I don't think they have chips yet.  If they did imagine how fast they could make difficulty go up  (added onto high already).  

The SP50 is just scary specs with all that.  The good news is it might be slow on making they don't seem to be selling them to many, I think anyone getting one currently is under NDA or they don't have any sold.  So I'm guessing NDA.

Dunno, maybe they ARE the reason why the difficulty went super up. Maybe they shipped 100's of units to a private data center. Its not like it would take a lot of these, when 10 of these do over 1PH/s.

well diff increased by a lot more than 1 peta, it was 10%+ more which mean 70 peta, i doubt a bunch of new asic can do this, more likely is a combination of many things

[VirosaGITS]

Except that the 10 1600w PSU. are Serevr style hot pluggable and are built into the unit.

Rich

Thanks Rich.  I should have took a look for myself at their website.  However, I really don't care for SPT anymore unless they change their ways.

Were lucky that I don't think they have chips yet.  If they did imagine how fast they could make difficulty go up  (added onto high already).  

The SP50 is just scary specs with all that.  The good news is it might be slow on making they don't seem to be selling them to many, I think anyone getting one currently is under NDA or they don't have any sold.  So I'm guessing NDA.

Dunno, maybe they ARE the reason why the difficulty went super up. Maybe they shipped 100's of units to a private data center. Its not like it would take a lot of these, when 10 of these do over 1PH/s.

well diff increased by a lot more than 1 peta, it was 10%+ more which mean 70 peta, i doubt a bunch of new asic can do this, more likely is a combination of many things

Point is, just a small MOQ of 50 for SP50 is over 5 PH/s. Now for a sizeable batch for a data center, we're looking at 100 PH/s, which is very much capable of singlehandly raising the diff by double digits by itself for a round or two. (There's bunch few mW Bitcoin mines already)

We're looking at 80PH/s per mere mW, now and thats still pretty small when we have the main companies onlining/working on 10-30mW datacenter.

From there, simple math: a single 10mW datacenter filled with current/prototype gen miners would be more by itself than the current network.

[dmwardjr]

Point is, just a small MOQ of 50 for SP50 is over 5 PH/s. Now for a sizeable batch for a data center, we're looking at 100 PH/s, which is very much capable of singlehandly raising the diff by double digits by itself for a round or two. (There's bunch few mW Bitcoin mines already)

We're looking at 80PH/s per mere mW, now and thats still pretty small when we have the main companies onlining/working on 10-30mW datacenter.

From there, simple math: a single 10mW datacenter filled with current/prototype gen miners would be more by itself than the current network.

After doing the math:

10,000,000 watts with 80% allowed usage [Especially, by insurance company for data center and NEC] should be operate 8,000,000 watts.  We need to allow for a few more items, such as, exhaust fans, etc...

We will deduct another 3% of that 8,000,000 watts for the other items used in the facility, which is 240,000 watts.  This leaves 7,760,000 watts remaining for rigs.

If all rigs were S7's batch 8 @ 1,293 watts each, this comes out to 6001 x S7's running 24/7 at allowed 80% demand of service.  The insurance company may limit them to 75% but I'm doing figures at 80 percent.

6001 x S7's at an average hash of 4,667 GH/s each [Just shy of 4,700 GH/s] = 28,006,667 GH/s (28,006 TH/s OR 28.006 PH/s).  This is far from the present network hash rate with 10 Megawatt facility.

I wanted to throw this up in the air for others to see.  You're welcome to do the figures for the SP50 if you like.  I just think it's important to keep in mind that a data center can only have 75 to 80 percent demand on their service when doing the calculations.  It will not quite be the full amount of wattage (power) the data center states.  Especially, if they are insured.

David

[VirosaGITS]

Point is, just a small MOQ of 50 for SP50 is over 5 PH/s. Now for a sizeable batch for a data center, we're looking at 100 PH/s, which is very much capable of singlehandly raising the diff by double digits by itself for a round or two. (There's bunch few mW Bitcoin mines already)

We're looking at 80PH/s per mere mW, now and thats still pretty small when we have the main companies onlining/working on 10-30mW datacenter.

From there, simple math: a single 10mW datacenter filled with current/prototype gen miners would be more by itself than the current network.

After doing the math:

10,000,000 watts with 80% allowed usage [Especially, by insurance company for data center and NEC] should be operate 8,000,000 watts.  We need to allow for a few more items, such as, exhaust fans, etc...

We will deduct another 3% of that 8,000,000 watts for the other items used in the facility, which is 240,000 watts.  This leaves 7,760,000 watts remaining for rigs.

If all rigs were S7's batch 8 @ 1,293 watts each, this comes out to 6001 x S7's running 24/7 at allowed 80% demand of service.  The insurance company may limit them to 75% but I'm doing figures at 80 percent.

6001 x S7's at an average hash of 4,667 GH/s each [Just shy of 4,700 GH/s] = 28,006,667 GH/s (28,006 TH/s OR 28.006 PH/s).  This is far from the present network hash rate with 10 Megawatt facility.

I wanted to throw this up in the air for others to see.  You're welcome to do the figures for the SP50 if you like.  I just think it's important to keep in mind that a data center can only have 75 to 80 percent demand on their service when doing the calculations.  It will not quite be the full amount of wattage (power) the data center states.  Especially, if they are insured.

David

Bah, the 1.2mW they listed was their consumption, not some rated number based on some regulation i havent heard of. But it does not really matter, as the math is for 1 mW consumed, if a 30mW only use 80% of 30mW then its going to use 24mW.

Then i'm not sure what you're going to be doing with the 3% you're adding, fans, maybe? There's no way a facility is going to be using 240kW of fans, but what if they dont need fans in the first place?

Anyways, the math for the SP50 was already done in the post you replied to, its about twice with S7, which sum up to 70-80PH/s. Albeit perhaps lower than actual 70PH/s if you start factoring in additional electric consumption overall. I can't say my numbers were conservative at all.

If you do simple quick dirty math of 10000 (kW) / 16(kW) you got how many SP50 you fit per mW. Then you multiply that by 110TH/s bam about 70PH/s.

Therefore a mine consuming 10mW would get close to 50% of the new total hashrate. Kinda crazy.

[dmwardjr]

Bah, the 1.2mW they listed was their consumption, not some rated number based on some regulation i havent heard of. But it does not really matter, as the math is for 1 mW consumed, if a 30mW only use 80% of 30mW then its going to use 24mW.

Then i'm not sure what you're going to be doing with the 3% you're adding, fans, maybe? There's no way a facility is going to be using 240kW of fans, but what if they dont need fans in the first place?

Anyways, the math for the SP50 was already done in the post you replied to, its about twice with S7, which sum up to 70-80PH/s. Albeit perhaps lower than actual 70PH/s if you start factoring in additional electric consumption overall. I can't say my numbers were conservative at all.

If you do simple quick dirty math of 10000 (kW) / 16(kW) you got how many SP50 you fit per mW. Then you multiply that by 110TH/s bam about 70PH/s.

Therefore a mine consuming 10mW would get close to 50% of the new total hashrate. Kinda crazy.

I totally agree with you about not using 240kW of exhaust fans.  IF they are located in a warmer climate, they would use AC for cooling.  IF they are in Iceland, Sweden, etc... they would not need it for AC.  The etc. was mainly for AC but would also include exhaust fans, lighting, etc...  It would probably be more like 100kW for AC and exhaust fans but it depends on the size of the facility.

BY THE WAY, BTCS has 3MW total at their facility in North Carolina with racks prepared for SP50's once they are rolled off the assembly line:  https://www.youtube.com/watch?v=DgHuyItPUfg

If a mining facility had 10MW available for mining [Typically reduced to 80 [IF not 75%] and had gear with similar efficiency and hash rates as that of the SP50, we would come up with the following figure [I will use 100kW for AC, fans, lighting, etc.:

80% of 10MW is 8MW minus 100kW for peripherals = 7,900,000 watts.

7,900,000 watts divided by 16,570 watts each = 476 SP50's

476 SP50's at 110 TH/s each = 53,360 TH/s OR 53.360 PH/s.  The present network hash rate [According to bitcoinwisdom.com/bitcoin/difficulty] is 666 PH/s.

Most every data center [Like BTCS] that tells you, for instance, they have a 3MW facility.  That's what they actually have, not what is available for mining.  Most any data center will tell you they are limited to 75 - 80 percent of their available power.  If their facility had 10MW available for mining, they would tell you.  If they just say their facility is 10MW, I would suspect they can only do 70 to 80 percent of that.  I could be wrong on this but BTCS, for instance, does not say specifically if their "facility" has 3MW available for mining or is it just their facility has 3MW service; of which, only 75 to 80 percent of it can be used within NEC guidelines without Main Service Breakers UL listed at 100% usage.

[VirosaGITS]

Bah, the 1.2mW they listed was their consumption, not some rated number based on some regulation i havent heard of. But it does not really matter, as the math is for 1 mW consumed, if a 30mW only use 80% of 30mW then its going to use 24mW.

Then i'm not sure what you're going to be doing with the 3% you're adding, fans, maybe? There's no way a facility is going to be using 240kW of fans, but what if they dont need fans in the first place?

Anyways, the math for the SP50 was already done in the post you replied to, its about twice with S7, which sum up to 70-80PH/s. Albeit perhaps lower than actual 70PH/s if you start factoring in additional electric consumption overall. I can't say my numbers were conservative at all.

If you do simple quick dirty math of 10000 (kW) / 16(kW) you got how many SP50 you fit per mW. Then you multiply that by 110TH/s bam about 70PH/s.

Therefore a mine consuming 10mW would get close to 50% of the new total hashrate. Kinda crazy.

I totally agree with you about not using 240kW of exhaust fans.  IF they are located in a warmer climate, they would use AC for cooling.  IF they are in Iceland, Sweden, etc... they would not need it for AC.  The etc. was mainly for AC but would also include exhaust fans, lighting, etc...  It would probably be more like 100kW for AC and exhaust fans but it depends on the size of the facility.

BY THE WAY, BTCS has 3MW total at their facility in North Carolina with racks prepared for SP50's once they are rolled off the assembly line:  https://www.youtube.com/watch?v=DgHuyItPUfg

If a mining facility had 10MW available for mining [Typically reduced to 80 [IF not 75%] and had gear with similar efficiency and hash rates as that of the SP50, we would come up with the following figure [I will use 100kW for AC, fans, lighting, etc.:

80% of 10MW is 8MW minus 100kW for peripherals = 7,900,000 watts.

7,900,000 watts divided by 16,570 watts each = 476 SP50's

476 SP50's at 110 TH/s each = 53,360 TH/s OR 53.360 PH/s.  The present network hash rate [According to bitcoinwisdom.com/bitcoin/difficulty] is 666 PH/s.

Most every data center [Like BTCS] that tells you, for instance, they have a 3MW facility.  That's what they actually have, not what is available for mining.  Most any data center will tell you they are limited to 75 - 80 percent of their available power.  If their facility had 10MW available for mining, they would tell you.  If they just say their facility is 10MW, I would suspect they can only do 70 to 80 percent of that.  I could be wrong on this but BTCS, for instance, does not say specifically if their "facility" has 3MW available for mining or is it just their facility has 3MW service; of which, only 75 to 80 percent of it can be used within NEC guidelines without Main Service Breakers UL listed at 100% usage.

Hmm. Well, it does not really matter. Not sure where your power for peripherals will go, unless you use millions of RasPi, i think you'll be fine.

And you're right, i added a 0 at the kW for some reason. It would be 1000/16.5=60.60*110 6.6whatever PH/s per mW.

And then Gridseed said they were going to roll out 0.06efficiency i believe? Then that would be about 15 PH/s per mW.

[dmwardjr]

Hmm. Well, it does not really matter. Not sure where your power for peripherals will go, unless you use millions of RasPi, i think you'll be fine.

And you're right, i added a 0 at the kW for some reason. It would be 1000/16.5=60.60*110 6.6whatever PH/s per mW.

And then Gridseed said they were going to roll out 0.06efficiency i believe? Then that would be about 15 PH/s per mW.

If many manufacturers come out with very power efficient rigs starting spring next year (around March), I can see the S7 becoming equivalent to the way an S5 is right now by block halving IF THE PRICE OF BITCOIN DOES NOT KEEP UP WITH THE DIFFICULTY.

The REAL test will be at block halving.  It will be an interesting last 3 months before block halving with the difficulty and price of bitcoin for sure.