Ethereum mining still profitable?

[coinyard]

It is a well known fact that smi tool shows way too low numbers.

actually not they are right i then measured with wattmeter, total for the system was 520w with cpu running 25%, 2 additional fan(120mm) on the motherboard, the motherboard itself, the psu+the ssd and the ram(2x4)

cpu alone is 15w at 25%-30% peak, ram are like 3w, the two fan are like other 2-3w, the ssd, is another 2-3w the psu is like 10w, and the motherboard is at least 30w probably more

this is a total of 70w, 520-70 = 450/3(i'm running 3 970) = 150w each, now from the web it seems that a motherboard can consume more than 30w in overloaded, plus add that each g1 gaming has 3 fan and not two

you need then to remove another 20w at least, so 140w each is the best number you can get

Did a test run with latest 368.22 drivers (W7 x64) and I must say I am really surprised it actually works now, numbers are on par with my power meter.

I still need 175-195w to 22MH though, 140w can be done with 19-20MH. Tested on coinotron stratum, recheck value 4000.

My undervoltaged 390 can do 27MHz with 160 W. If you undervolt the 970, can you do better?

[1715361941]

1715361941

[1715361941]

1715361941

[1715361941]

1715361941

[1715361941]

1715361941

[Amph]

It is a well known fact that smi tool shows way too low numbers.

actually not they are right i then measured with wattmeter, total for the system was 520w with cpu running 25%, 2 additional fan(120mm) on the motherboard, the motherboard itself, the psu+the ssd and the ram(2x4)

cpu alone is 15w at 25%-30% peak, ram are like 3w, the two fan are like other 2-3w, the ssd, is another 2-3w the psu is like 10w, and the motherboard is at least 30w probably more

this is a total of 70w, 520-70 = 450/3(i'm running 3 970) = 150w each, now from the web it seems that a motherboard can consume more than 30w in overloaded, plus add that each g1 gaming has 3 fan and not two

you need then to remove another 20w at least, so 140w each is the best number you can get

Did a test run with latest 368.22 drivers (W7 x64) and I must say I am really surprised it actually works now, numbers are on par with my power meter.

I still need 175-195w to 22MH though, 140w can be done with 19-20MH. Tested on coinotron stratum, recheck value 4000.

My undervoltaged 390 can do 27MHz with 160 W. If you undervolt the 970, can you do better?

you can not compare like this, because different card have different consumption

i mean your card probably have two fan only and less component than a g1 gaming

i just tested and i can do 20MH with only 124w, now from those 124 you need to remove the fact that g1 gaming have 3 fan and a more power hungry pcb, because of the bigger heatsink

[Posum578]

It is a well known fact that smi tool shows way too low numbers.

actually not they are right i then measured with wattmeter, total for the system was 520w with cpu running 25%, 2 additional fan(120mm) on the motherboard, the motherboard itself, the psu+the ssd and the ram(2x4)

cpu alone is 15w at 25%-30% peak, ram are like 3w, the two fan are like other 2-3w, the ssd, is another 2-3w the psu is like 10w, and the motherboard is at least 30w probably more

this is a total of 70w, 520-70 = 450/3(i'm running 3 970) = 150w each, now from the web it seems that a motherboard can consume more than 30w in overloaded, plus add that each g1 gaming has 3 fan and not two

you need then to remove another 20w at least, so 140w each is the best number you can get

Did a test run with latest 368.22 drivers (W7 x64) and I must say I am really surprised it actually works now, numbers are on par with my power meter.

I still need 175-195w to 22MH though, 140w can be done with 19-20MH. Tested on coinotron stratum, recheck value 4000.

My undervoltaged 390 can do 27MHz with 160 W. If you undervolt the 970, can you do better?

you can not compare like this, because different card have different consumption

i mean you card probably have two fan onyl and less component than a g1 gaming

i just tested and i can do 20MH with only 124w, now from those 124 you need to remove the fact that g1 gaming have 3 fan and a more power hungry pcb, because of the bigger heatsink

So basically, it is better to buy the cheap non "gaming" cards with less components and undervolt and get better efficiency.

[Amph]

It is a well known fact that smi tool shows way too low numbers.

actually not they are right i then measured with wattmeter, total for the system was 520w with cpu running 25%, 2 additional fan(120mm) on the motherboard, the motherboard itself, the psu+the ssd and the ram(2x4)

cpu alone is 15w at 25%-30% peak, ram are like 3w, the two fan are like other 2-3w, the ssd, is another 2-3w the psu is like 10w, and the motherboard is at least 30w probably more

this is a total of 70w, 520-70 = 450/3(i'm running 3 970) = 150w each, now from the web it seems that a motherboard can consume more than 30w in overloaded, plus add that each g1 gaming has 3 fan and not two

you need then to remove another 20w at least, so 140w each is the best number you can get

Did a test run with latest 368.22 drivers (W7 x64) and I must say I am really surprised it actually works now, numbers are on par with my power meter.

I still need 175-195w to 22MH though, 140w can be done with 19-20MH. Tested on coinotron stratum, recheck value 4000.

My undervoltaged 390 can do 27MHz with 160 W. If you undervolt the 970, can you do better?

you can not compare like this, because different card have different consumption

i mean you card probably have two fan onyl and less component than a g1 gaming

i just tested and i can do 20MH with only 124w, now from those 124 you need to remove the fact that g1 gaming have 3 fan and a more power hungry pcb, because of the bigger heatsink

So basically, it is better to buy the cheap non "gaming" cards with less components and undervolt and get better efficiency.

yeah like the 970 mini itx, 1 fan only smaller pcb and heatsink, will consume much less while performing similar, but with less potential oc

same for 380 mini itx, and nano, that's why the nano is the best right now

[macwika]

I think nano is different. The chips are selected specially so that the power consumption is very low.

[antantti]

Did a test run with latest 368.22 drivers (W7 x64) and I must say I am really surprised it actually works now, numbers are on par with my power meter.

I still need 175-195w to 22MH though, 140w can be done with 19-20MH. Tested on coinotron stratum, recheck value 4000.

My undervoltaged 390 can do 27MHz with 160 W. If you undervolt the 970, can you do better?

You can't undervolt 970, you just lower the limit you let GPU "spend" power and nvidiamagic does it all for you.

Let me guess, -200mV and 1000 core with that 390? I am running mine 30MH+0.6GH with 260w...

[antantti]

you can not compare like this, because different card have different consumption

i mean your card probably have two fan only and less component than a g1 gaming

i just tested and i can do 20MH with only 124w, now from those 124 you need to remove the fact that g1 gaming have 3 fan and a more power hungry pcb, because of the bigger heatsink

With same type of memory, core&mem speed they all have more or less the same power consumption.

Bigger than 6+6pin power connectors just let you use more power and bigger heatsinks and more fans give tools for getting rid of that increased heat.

980 G1 is just as efficient at 20MH as Strix 970.

[Amph]

you can not compare like this, because different card have different consumption

i mean your card probably have two fan only and less component than a g1 gaming

i just tested and i can do 20MH with only 124w, now from those 124 you need to remove the fact that g1 gaming have 3 fan and a more power hungry pcb, because of the bigger heatsink

With same type of memory, core&mem speed they all have more or less the same power consumption.

Bigger than 6+6pin power connectors just let you use more power and bigger heatsinks and more fans give tools for getting rid of that increased heat.

980 G1 is just as efficient at 20MH as Strix 970.

that "more or less" can make the difference here, since we are talking about a very small difference, about 10w per card

this is what i'm talking about

[antantti]

you can not compare like this, because different card have different consumption

i mean your card probably have two fan only and less component than a g1 gaming

i just tested and i can do 20MH with only 124w, now from those 124 you need to remove the fact that g1 gaming have 3 fan and a more power hungry pcb, because of the bigger heatsink

With same type of memory, core&mem speed they all have more or less the same power consumption.

Bigger than 6+6pin power connectors just let you use more power and bigger heatsinks and more fans give tools for getting rid of that increased heat.

980 G1 is just as efficient at 20MH as Strix 970.

that "more or less" can make the difference here, since we are talking about a very small difference, about 10w per card

this is what i'm talking about

If you pay $0.1/ Kwh your mining rev/ profit would be (according to whattomine):

GTX970 20MH/120w $2.62/ 2.33 per day
AMD390 30MH/230w $3.93/ 3.38 per day

I just can't see how 10w could be a gamechanger here.

[mrehanabbasi]

Which would be more profitable: ETH cloud mining or using my own hardware?

[Marvell1]

It is a well known fact that smi tool shows way too low numbers.

actually not they are right i then measured with wattmeter, total for the system was 520w with cpu running 25%, 2 additional fan(120mm) on the motherboard, the motherboard itself, the psu+the ssd and the ram(2x4)

cpu alone is 15w at 25%-30% peak, ram are like 3w, the two fan are like other 2-3w, the ssd, is another 2-3w the psu is like 10w, and the motherboard is at least 30w probably more

this is a total of 70w, 520-70 = 450/3(i'm running 3 970) = 150w each, now from the web it seems that a motherboard can consume more than 30w in overloaded, plus add that each g1 gaming has 3 fan and not two

you need then to remove another 20w at least, so 140w each is the best number you can get

Did a test run with latest 368.22 drivers (W7 x64) and I must say I am really surprised it actually works now, numbers are on par with my power meter.

I still need 175-195w to 22MH though, 140w can be done with 19-20MH. Tested on coinotron stratum, recheck value 4000.

My undervoltaged 390 can do 27MHz with 160 W. If you undervolt the 970, can you do better?

BS man i have a ton of 390s -120mv undervolt with claymore dual miner they still pull 240 watts, if i was solo mining eth i would gain at most 50 watts thats still 190 watts

[Amph]

you can not compare like this, because different card have different consumption

i mean your card probably have two fan only and less component than a g1 gaming

i just tested and i can do 20MH with only 124w, now from those 124 you need to remove the fact that g1 gaming have 3 fan and a more power hungry pcb, because of the bigger heatsink

With same type of memory, core&mem speed they all have more or less the same power consumption.

Bigger than 6+6pin power connectors just let you use more power and bigger heatsinks and more fans give tools for getting rid of that increased heat.

980 G1 is just as efficient at 20MH as Strix 970.

that "more or less" can make the difference here, since we are talking about a very small difference, about 10w per card

this is what i'm talking about

If you pay $0.1/ Kwh your mining rev/ profit would be (according to whattomine):

GTX970 20MH/120w $2.62/ 2.33 per day
AMD390 30MH/230w $3.93/ 3.38 per day

I just can't see how 10w could be a gamechanger here.

30MH at 230w isn't better than 20MH at 120w, 0.13 vs 0.16 ratio

30MH is only 50% more than 20MH, but 230w isn't 50% more than 120w, but almost 2x more, so it's worse

[Marvell1]

you can not compare like this, because different card have different consumption

i mean your card probably have two fan only and less component than a g1 gaming

i just tested and i can do 20MH with only 124w, now from those 124 you need to remove the fact that g1 gaming have 3 fan and a more power hungry pcb, because of the bigger heatsink

With same type of memory, core&mem speed they all have more or less the same power consumption.

Bigger than 6+6pin power connectors just let you use more power and bigger heatsinks and more fans give tools for getting rid of that increased heat.

980 G1 is just as efficient at 20MH as Strix 970.

that "more or less" can make the difference here, since we are talking about a very small difference, about 10w per card

this is what i'm talking about

If you pay $0.1/ Kwh your mining rev/ profit would be (according to whattomine):

GTX970 20MH/120w $2.62/ 2.33 per day
AMD390 30MH/230w $3.93/ 3.38 per day

I just can't see how 10w could be a gamechanger here.

30MH at 230w isn't better than 20MH at 120w, 0.13 vs 0.16 ratio

30MH is only 50% more than 20%, but 230 isn't 50% more than 120w, but 2x more, so it's worse

No GTX 970 uses only 120 watts my god stop with the missinformation, bad info like this lead me to buy 18 gtx 970s of different brands and guess what i found ?

they all pull on average 150 to 170 watts to get 20mhs. If you enable SMI you start seeing wattage in the 190s-200s to get 22 watts.

Belive my im slowing trying to get rid of ALL my gtx 970s since they cant even dual mine and use this much power dont belive these 120 watts clowns i can show pics

if needed with all the cards hooked up to my PDU , six cards i would see 4-5amps which is close to 1200 watts


-People are not stupid btw  they are buying 390s for Eth mining for gtx 970s those in the know whats up, theres a reason why on ethermining AMD cards are recomeded hands down.

[Amph]

you can not compare like this, because different card have different consumption

i mean your card probably have two fan only and less component than a g1 gaming

i just tested and i can do 20MH with only 124w, now from those 124 you need to remove the fact that g1 gaming have 3 fan and a more power hungry pcb, because of the bigger heatsink

With same type of memory, core&mem speed they all have more or less the same power consumption.

Bigger than 6+6pin power connectors just let you use more power and bigger heatsinks and more fans give tools for getting rid of that increased heat.

980 G1 is just as efficient at 20MH as Strix 970.

that "more or less" can make the difference here, since we are talking about a very small difference, about 10w per card

this is what i'm talking about

If you pay $0.1/ Kwh your mining rev/ profit would be (according to whattomine):

GTX970 20MH/120w $2.62/ 2.33 per day
AMD390 30MH/230w $3.93/ 3.38 per day

I just can't see how 10w could be a gamechanger here.

30MH at 230w isn't better than 20MH at 120w, 0.13 vs 0.16 ratio

30MH is only 50% more than 20%, but 230 isn't 50% more than 120w, but 2x more, so it's worse

No GTX 970 uses only 120 watts my god stop with the missinformation, bad info like this lead me to buy 18 gtx 970s of different brands and guess what i found ?

they all pull on average 150 to 170 watts to get 20mhs. If you enable SMI you start seeing wattage in the 190s-200s to get 22 watts.

Belive my im slowing trying to get rid of ALL my gtx 970s since they cant even dual mine and use this much power dont belive these 120 watts clowns i can show pics

if needed with all the cards hooked up to my PDU , six cards i would see 4-5amps which is close to 1200 watts


-People are not stupid btw  they are buying 390s for Eth mining for gtx 970s those in the know whats up, theres a reason why on ethermining AMD cards are recomeded hands down.

misinformation you say?



before you say that smi it's not accurate, i have tested with wattmeter and it report the same

[antantti]

you can not compare like this, because different card have different consumption

i mean your card probably have two fan only and less component than a g1 gaming

i just tested and i can do 20MH with only 124w, now from those 124 you need to remove the fact that g1 gaming have 3 fan and a more power hungry pcb, because of the bigger heatsink

With same type of memory, core&mem speed they all have more or less the same power consumption.

Bigger than 6+6pin power connectors just let you use more power and bigger heatsinks and more fans give tools for getting rid of that increased heat.

980 G1 is just as efficient at 20MH as Strix 970.

that "more or less" can make the difference here, since we are talking about a very small difference, about 10w per card

this is what i'm talking about

If you pay $0.1/ Kwh your mining rev/ profit would be (according to whattomine):

GTX970 20MH/120w $2.62/ 2.33 per day
AMD390 30MH/230w $3.93/ 3.38 per day

I just can't see how 10w could be a gamechanger here.

30MH at 230w isn't better than 20MH at 120w, 0.13 vs 0.16 ratio

30MH is only 50% more than 20MH, but 230w isn't 50% more than 120w, but almost 2x more, so it's worse

What I am trying to say is that, this thread being about eth mining profitability, 390 wins 970 hands down 3.38-2.33. Even when I used numbers that favor 970. Only a fool would buy 970 over 390 to eth mining.

Hash/ watt is maybe worth another thread. My 390's kill my 970's on that competition too. My 970's need more than 120w to 20MH and 390's less than 230w to 30MH.

[Opquar]

you can not compare like this, because different card have different consumption

i mean your card probably have two fan only and less component than a g1 gaming

i just tested and i can do 20MH with only 124w, now from those 124 you need to remove the fact that g1 gaming have 3 fan and a more power hungry pcb, because of the bigger heatsink

With same type of memory, core&mem speed they all have more or less the same power consumption.

Bigger than 6+6pin power connectors just let you use more power and bigger heatsinks and more fans give tools for getting rid of that increased heat.

980 G1 is just as efficient at 20MH as Strix 970.

that "more or less" can make the difference here, since we are talking about a very small difference, about 10w per card

this is what i'm talking about

If you pay $0.1/ Kwh your mining rev/ profit would be (according to whattomine):

GTX970 20MH/120w $2.62/ 2.33 per day
AMD390 30MH/230w $3.93/ 3.38 per day

I just can't see how 10w could be a gamechanger here.

30MH at 230w isn't better than 20MH at 120w, 0.13 vs 0.16 ratio

30MH is only 50% more than 20MH, but 230w isn't 50% more than 120w, but almost 2x more, so it's worse

that is right It depends on the price of Ethereum. If the price drops 50%, then it might be better to use 970.

What I am trying to say is that, this thread being about eth mining profitability, 390 wins 970 hands down 3.38-2.33. Even when I used numbers that favor 970. Only a fool would buy 970 over 390 to eth mining.

Hash/ watt is maybe worth another thread. My 390's kill my 970's on that competition too. My 970's need more than 120w to 20MH and 390's less than 230w to 30MH.

[antantti]

No GTX 970 uses only 120 watts my god stop with the missinformation, bad info like this lead me to buy 18 gtx 970s of different brands and guess what i found ?

they all pull on average 150 to 170 watts to get 20mhs. If you enable SMI you start seeing wattage in the 190s-200s to get 22 watts.

Belive my im slowing trying to get rid of ALL my gtx 970s since they cant even dual mine and use this much power dont belive these 120 watts clowns i can show pics

if needed with all the cards hooked up to my PDU , six cards i would see 4-5amps which is close to 1200 watts


-People are not stupid btw  they are buying 390s for Eth mining for gtx 970s those in the know whats up, theres a reason why on ethermining AMD cards are recomeded hands down.

I haven't yet seen 970 or 980 that can do 20MH with 120watts but 18MH is really easy.

Would I choose 970 for eth mining over amd? No.

[Amph]

you can not compare like this, because different card have different consumption

i mean your card probably have two fan only and less component than a g1 gaming

i just tested and i can do 20MH with only 124w, now from those 124 you need to remove the fact that g1 gaming have 3 fan and a more power hungry pcb, because of the bigger heatsink

With same type of memory, core&mem speed they all have more or less the same power consumption.

Bigger than 6+6pin power connectors just let you use more power and bigger heatsinks and more fans give tools for getting rid of that increased heat.

980 G1 is just as efficient at 20MH as Strix 970.

that "more or less" can make the difference here, since we are talking about a very small difference, about 10w per card

this is what i'm talking about

If you pay $0.1/ Kwh your mining rev/ profit would be (according to whattomine):

GTX970 20MH/120w $2.62/ 2.33 per day
AMD390 30MH/230w $3.93/ 3.38 per day

I just can't see how 10w could be a gamechanger here.

30MH at 230w isn't better than 20MH at 120w, 0.13 vs 0.16 ratio

30MH is only 50% more than 20MH, but 230w isn't 50% more than 120w, but almost 2x more, so it's worse

What I am trying to say is that, this thread being about eth mining profitability, 390 wins 970 hands down 3.38-2.33. Even when I used numbers that favor 970. Only a fool would buy 970 over 390 to eth mining.

Hash/ watt is maybe worth another thread. My 390's kill my 970's on that competition too. My 970's need more than 120w to 20MH and 390's less than 230w to 30MH.

well not my case then, as you can see my 970 can do 20MH with 125w, this is better than a 390 or at least equal, so i don't agree on "only a fool would buy a 970 instead of 390"

i'm talking about g1 gaming, maybe other gpu have low performace because of poor compoent, poor memeory or stuff like that

[antantti]

you can not compare like this, because different card have different consumption

i mean your card probably have two fan only and less component than a g1 gaming

i just tested and i can do 20MH with only 124w, now from those 124 you need to remove the fact that g1 gaming have 3 fan and a more power hungry pcb, because of the bigger heatsink

With same type of memory, core&mem speed they all have more or less the same power consumption.

Bigger than 6+6pin power connectors just let you use more power and bigger heatsinks and more fans give tools for getting rid of that increased heat.

980 G1 is just as efficient at 20MH as Strix 970.

that "more or less" can make the difference here, since we are talking about a very small difference, about 10w per card

this is what i'm talking about

If you pay $0.1/ Kwh your mining rev/ profit would be (according to whattomine):

GTX970 20MH/120w $2.62/ 2.33 per day
AMD390 30MH/230w $3.93/ 3.38 per day

I just can't see how 10w could be a gamechanger here.

30MH at 230w isn't better than 20MH at 120w, 0.13 vs 0.16 ratio

30MH is only 50% more than 20MH, but 230w isn't 50% more than 120w, but almost 2x more, so it's worse

What I am trying to say is that, this thread being about eth mining profitability, 390 wins 970 hands down 3.38-2.33. Even when I used numbers that favor 970. Only a fool would buy 970 over 390 to eth mining.

Hash/ watt is maybe worth another thread. My 390's kill my 970's on that competition too. My 970's need more than 120w to 20MH and 390's less than 230w to 30MH.

well not my case then, as you can see my 970 can do 20MH with 125w, this is better than a 390 or at least equal, so i don't agree on "only a fool would buy a 970 instead of 390"

i'm talking about g1 gaming, maybe other gpu have low performace because of poor compoent, poor memeory or stuff like that

Look at my example again.

If 970 used ZERO watts it would earn only $2.62/ day. 390 profits $3.38 after 230 watts expenses.

Only a fool would buy a 970 instead of 390 for ethereum mining.

[Amph]

you can not compare like this, because different card have different consumption

i mean your card probably have two fan only and less component than a g1 gaming

i just tested and i can do 20MH with only 124w, now from those 124 you need to remove the fact that g1 gaming have 3 fan and a more power hungry pcb, because of the bigger heatsink

With same type of memory, core&mem speed they all have more or less the same power consumption.

Bigger than 6+6pin power connectors just let you use more power and bigger heatsinks and more fans give tools for getting rid of that increased heat.

980 G1 is just as efficient at 20MH as Strix 970.

that "more or less" can make the difference here, since we are talking about a very small difference, about 10w per card

this is what i'm talking about

If you pay $0.1/ Kwh your mining rev/ profit would be (according to whattomine):

GTX970 20MH/120w $2.62/ 2.33 per day
AMD390 30MH/230w $3.93/ 3.38 per day

I just can't see how 10w could be a gamechanger here.

30MH at 230w isn't better than 20MH at 120w, 0.13 vs 0.16 ratio

30MH is only 50% more than 20MH, but 230w isn't 50% more than 120w, but almost 2x more, so it's worse

What I am trying to say is that, this thread being about eth mining profitability, 390 wins 970 hands down 3.38-2.33. Even when I used numbers that favor 970. Only a fool would buy 970 over 390 to eth mining.

Hash/ watt is maybe worth another thread. My 390's kill my 970's on that competition too. My 970's need more than 120w to 20MH and 390's less than 230w to 30MH.

well not my case then, as you can see my 970 can do 20MH with 125w, this is better than a 390 or at least equal, so i don't agree on "only a fool would buy a 970 instead of 390"

i'm talking about g1 gaming, maybe other gpu have low performace because of poor compoent, poor memeory or stuff like that

Look at my example again.

If 970 used ZERO watts it would earn only $2.62/ day. 390 profits $3.38 after 230 watts expenses.

Only a fool would buy a 970 instead of 390 for ethereum mining.

you are not understading what you are posting

it's obvious that more hash equal to more profit, but this does not mean that the 390 is better than a 970

you need to do thing in the right way, by calculating with the same hash, so put in the calculator this

3x970 at 120w each, 60MH total, vs 2x390 at 230w each, 60MH total, this with your example...