ethminer-0.9.41-genoil-1.1

[Rookie41]

I'm having a bit of trouble finding info on how to fix my problem.

from the first post:

"4. Can I mine ETH with my 2GB GPU?
2GB should be sufficient until Ethereum switches to PoS."

I can mine fine on my RX 470 8GB.
My RX 470 with 4GB refuses to mine cause it has insufficient memory?

Using ethminer-0.9.41-genoil-1.1.7 on Win 10 with Radeon-Crimson-16.7.3-Win10-64Bit.

I tried looking for parameters but I'm having a hard time finding any information on how to set it all up.

On my working 8GB card I use something like this:
" ethminer -SP 2 -G -S pooladdress:port -O wallet.user:pass "

Any idea what parameters to use to mine with my 4GB card ? Or any link with all available parameters? The mining files come with no help file or readme

[1715182152]

1715182152

[1715182152]

1715182152

[PovertyByte]

Well I just realized what a fucked up bitch these nvidia voltage regulations are.

Underclocking the GPU core increases the power that you need because the driver feeds the chip a higher voltage. Therefore its useful with the current genoil miner to increase the gpu core speed to save some power( ), because the mining gpu speed is anyway locked at 1620 (atleast for my card) and reduce the power limit to 50%.

Core Clock -400 -> Voltage 1.043 V -> 77% TDP = 146 W
Core Clock +150 -> Voltage 0,76 V -> 54% TDP = 102 W

Have fun, it's really that retarded..

i noticed this also, but only if you play with the core, without touching the power limit, otherwise it is forced to drop since the PL limit the current

p.s. i was able to reach 30MH with +600 on the core and around 100w only



those 27MH as a bottom bar, is because i'm using the pc, otherwise it would be 29+ stable with 30 peak

i believe you can easilòy do 80-90w 30MH or even lower, which would put the 1070 as the best card for etheruem not even the nano can compete

I have the power limit set to 51 and the memory clock at 561 doing the same hash rate but more stable. No performance difference between 561 and 600. I'm probably sticking with this setting for the rest of the summer until it gets colder or when we are finally given access to the core voltage settings.

I have noticed that raising the core clock up a bit gives it more stability. Although mine is offset -200 mine is a factory overclocked card so that may be a factor

[Amph]

Well I just realized what a fucked up bitch these nvidia voltage regulations are.

Underclocking the GPU core increases the power that you need because the driver feeds the chip a higher voltage. Therefore its useful with the current genoil miner to increase the gpu core speed to save some power( ), because the mining gpu speed is anyway locked at 1620 (atleast for my card) and reduce the power limit to 50%.

Core Clock -400 -> Voltage 1.043 V -> 77% TDP = 146 W
Core Clock +150 -> Voltage 0,76 V -> 54% TDP = 102 W

Have fun, it's really that retarded..

i noticed this also, but only if you play with the core, without touching the power limit, otherwise it is forced to drop since the PL limit the current

p.s. i was able to reach 30MH with +600 on the core and around 100w only



those 27MH as a bottom bar, is because i'm using the pc, otherwise it would be 29+ stable with 30 peak

i believe you can easilòy do 80-90w 30MH or even lower, which would put the 1070 as the best card for etheruem not even the nano can compete

I have the power limit set to 51 and the memory clock at 561 doing the same hash rate but more stable. No performance difference between 561 and 600. I'm probably sticking with this setting for the rest of the summer until it gets colder or when we are finally given access to the core voltage settings.

I have noticed that raising the core clock up a bit gives it more stability. Although mine is offset -200 mine is a factory overclocked card so that may be a factor

mine is not stable probably because is the first gpu attached to a 1440p monitor

[PovertyByte]

Well I just realized what a fucked up bitch these nvidia voltage regulations are.

Underclocking the GPU core increases the power that you need because the driver feeds the chip a higher voltage. Therefore its useful with the current genoil miner to increase the gpu core speed to save some power( ), because the mining gpu speed is anyway locked at 1620 (atleast for my card) and reduce the power limit to 50%.

Core Clock -400 -> Voltage 1.043 V -> 77% TDP = 146 W
Core Clock +150 -> Voltage 0,76 V -> 54% TDP = 102 W

Have fun, it's really that retarded..

i noticed this also, but only if you play with the core, without touching the power limit, otherwise it is forced to drop since the PL limit the current

p.s. i was able to reach 30MH with +600 on the core and around 100w only

https://s9.postimg.org/cgtnaerqn/Untitled.png

those 27MH as a bottom bar, is because i'm using the pc, otherwise it would be 29+ stable with 30 peak

i believe you can easilòy do 80-90w 30MH or even lower, which would put the 1070 as the best card for etheruem not even the nano can compete

I have the power limit set to 51 and the memory clock at 561 doing the same hash rate but more stable. No performance difference between 561 and 600. I'm probably sticking with this setting for the rest of the summer until it gets colder or when we are finally given access to the core voltage settings.

I have noticed that raising the core clock up a bit gives it more stability. Although mine is offset -200 mine is a factory overclocked card so that may be a factor

mine is not stable probably because is the first gpu attached to a 1440p monitor

Reminds me. I saw a post you made in another thread about turning sleep mode off on the monitor. I can't find my reply in my post history, I might not have bumped that.

What does sleep mode matter on the monitor? And would lowering the screen resolution matter? I'm on a 1080p. And the -200 offset for the core clock brings me to about 1544 mhz which is higher than yours. So it is possible the core clock may be the source of stability although it may not be worth the power draw.

[Amph]

Well I just realized what a fucked up bitch these nvidia voltage regulations are.

Underclocking the GPU core increases the power that you need because the driver feeds the chip a higher voltage. Therefore its useful with the current genoil miner to increase the gpu core speed to save some power( ), because the mining gpu speed is anyway locked at 1620 (atleast for my card) and reduce the power limit to 50%.

Core Clock -400 -> Voltage 1.043 V -> 77% TDP = 146 W
Core Clock +150 -> Voltage 0,76 V -> 54% TDP = 102 W

Have fun, it's really that retarded..

i noticed this also, but only if you play with the core, without touching the power limit, otherwise it is forced to drop since the PL limit the current

p.s. i was able to reach 30MH with +600 on the core and around 100w only

https://s9.postimg.org/cgtnaerqn/Untitled.png

those 27MH as a bottom bar, is because i'm using the pc, otherwise it would be 29+ stable with 30 peak

i believe you can easilòy do 80-90w 30MH or even lower, which would put the 1070 as the best card for etheruem not even the nano can compete

I have the power limit set to 51 and the memory clock at 561 doing the same hash rate but more stable. No performance difference between 561 and 600. I'm probably sticking with this setting for the rest of the summer until it gets colder or when we are finally given access to the core voltage settings.

I have noticed that raising the core clock up a bit gives it more stability. Although mine is offset -200 mine is a factory overclocked card so that may be a factor

mine is not stable probably because is the first gpu attached to a 1440p monitor

Reminds me. I saw a post you made in another thread about turning sleep mode off on the monitor. I can't find my reply in my post history, I might not have bumped that.

What does sleep mode matter on the monitor? And would lowering the screen resolution matter? I'm on a 1080p. And the -200 offset for the core clock brings me to about 1544 mhz which is higher than yours. So it is possible the core clock may be the source of stability although it may not be worth the power draw.

so you gpu clock begin from 1700, this is the same as me with boost

the sleep was more about the pc not the monitor, i just added the monitor, just to make it clear what i meant

[PovertyByte]

i noticed this also, but only if you play with the core, without touching the power limit, otherwise it is forced to drop since the PL limit the current

p.s. i was able to reach 30MH with +600 on the core and around 100w only



those 27MH as a bottom bar, is because i'm using the pc, otherwise it would be 29+ stable with 30 peak

i believe you can easily do 80-90w 30MH or even lower, which would put the 1070 as the best card for etheruem not even the nano can compete

Reminds me. I saw a post you made in another thread about turning sleep mode off on the monitor. I can't find my reply in my post history, I might not have bumped that.

What does sleep mode matter on the monitor? And would lowering the screen resolution matter? I'm on a 1080p. And the -200 offset for the core clock brings me to about 1544 mhz which is higher than yours. So it is possible the core clock may be the source of stability although it may not be worth the power draw.

so you gpu clock begin from 1700, this is the same as me with boost

the sleep was more about the pc not the monitor, i just added the monitor, just to make it clear what i meant

I don't know which GTX 1070 you have to compare mine with, but I do know that when I clock mine to the same settings my sensors read the same memory clock of 4404 but my core clock is about 1493-1506. And this is with my power limit at 50 instead of 55 which means my core clock should be falling lower. There is a clear relationship between the core clock sensor graph and the hash rates I am seeing.

I've observed that setting the core clock up from -200 can give more stability assuming the power limit isn't too low. And when the power limit is too low I am observing the core clock fluctuating a lot. When the core clock fluctuates so does the hash rate.

[localhostage]

Using Genoil 1.1.7 from (110)

Code:

Found suitable OpenCL device [GeForce GTX 1080 ] with 8589934592 bytes of GPU memory
  i  13:33:15|main  Connecting to stratum server us1.ethermine.org:4444
  i  13:33:15|stratum  Connected to stratum server us1.ethermine.org : 4444
  i  13:33:15|stratum  Starting farm
  i  13:33:16|stratum  Subscribed to stratum server
  X  13:33:16|stratum  Read response failed: End of file
  i  13:33:16|stratum  Reconnecting in 3 seconds...
  i  13:33:19|stratum  Connecting to stratum server us1.ethermine.org:4444
  i  13:33:19|stratum  Connected to stratum server us1.ethermine.org : 4444
  i  13:33:19|stratum  Subscribed to stratum server
  X  13:33:19|stratum  Read response failed: End of file
  i  13:33:19|stratum  Reconnecting in 3 seconds...

I've tried SP 0, 1, 2 sadly keep getting same message for eu1 and us1 servers on ethermine.org.

Am I doing something wrong? My batch file looks like so:

Code:

setx GPU_FORCE_64BIT_PTR 0
setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_SINGLE_ALLOC_PERCENT 100
ethminer.exe -SP 0 --farm-recheck 200 -G -S us1.ethermine.org:4444 -FS eu1.ethermine.org:4444 -O <my address>.1080

Thanks.


*EDIT* Nevermind, figured it out, my address had extra dots that apparently breaks things. I was using something like:    0xABC123.1080.FOO   removing the extra . got things working.

[giagge]

I have a problem .

Config:

ethminer-0.9.41-genoil-1.0.7

1 x Nvidia gtx 1070
Windows 10 x64 redstone ( 1607 update )
Driver 369.09 win 10 x64

Overclock +0mhz core +400mhz GDDR5

29mhs

I mining for 1 hour , but after i have 0.00 mhs and buffer memory out .

Use this .bat:

ethminer -U --opencl-platform 1 -G -S ele.suprnova.cc:5533 -O my user:my pass

.

[Lafu]

I have a problem .

Config:

ethminer-0.9.41-genoil-1.0.7

1 x Nvidia gtx 1070
Windows 10 x64 redstone ( 1607 update )
Driver 369.09 win 10 x64

Overclock +0mhz core +400mhz GDDR5

29mhs

I mining for 1 hour , but after i have 0.00 mhs and buffer memory out .

Use this .bat:

ethminer -U --opencl-platform 1 -G -S ele.suprnova.cc:5533 -O my user:my pass

.

have you tryed port 3333

and use this one https://github.com/nicehash/cpp-ethereum/releases

[giagge]

I have a problem .

Config:

ethminer-0.9.41-genoil-1.0.7

1 x Nvidia gtx 1070
Windows 10 x64 redstone ( 1607 update )
Driver 369.09 win 10 x64

Overclock +0mhz core +400mhz GDDR5

29mhs

I mining for 1 hour , but after i have 0.00 mhs and buffer memory out .

Use this .bat:

ethminer -U --opencl-platform 1 -G -S ele.suprnova.cc:5533 -O my user:my pass

.

have you tryed port 3333

and use this one https://github.com/nicehash/cpp-ethereum/releases

THANKS!!!

Work fine now! .

[induktor]

Hello
Question, what hashrate should I expect from a GTX 1070 stock speed (no overclock) ?
i am getting 25MH in OpenGL (cuda does not work yet, some issues with the driver) I read in the recent posts that they are getting 30MH with overclock so i'm not that far off i think, but better to ask
thanks

[carlo_0000]

I'm having a bit of trouble finding info on how to fix my problem.

from the first post:

"4. Can I mine ETH with my 2GB GPU?
2GB should be sufficient until Ethereum switches to PoS."

I can mine fine on my RX 470 8GB.
My RX 470 with 4GB refuses to mine cause it has insufficient memory?

Using ethminer-0.9.41-genoil-1.1.7 on Win 10 with Radeon-Crimson-16.7.3-Win10-64Bit.

I tried looking for parameters but I'm having a hard time finding any information on how to set it all up.

On my working 8GB card I use something like this:
" ethminer -SP 2 -G -S pooladdress:port -O wallet.user:pass "

Any idea what parameters to use to mine with my 4GB card ? Or any link with all available parameters? The mining files come with no help file or readme

did you try with   --opencl-platform 1
i have one computer i have to add that option otherwise it s saying  not enough memory

when it s working can you tell us what speed you get with r9 470 ?
and test an older driver 15.x something (actually no need to install that driver but copy the dll's to the ether miner directory
it s somewhere here in the topic, with that driver it s faster on 2 and 3 series (but don't know if it works with 4)

[Spiffy_1]

Since I've updated to windows anniversary, I've notice my 1060 farm acting strangely using genoil miner.  It taxes my cpu 100% and grinds the system to a halt until it seems a buffer clears, then scrolls as normal hashing happily away at 85 Mhash, then 100% cpu again, and craters the system.  I've had to switch to claymore since he fixed the bug that puts cpu at maximum.  Please look into this.  I dislike being forced to donate.

[Lafu]

I'm having a bit of trouble finding info on how to fix my problem.

from the first post:

"4. Can I mine ETH with my 2GB GPU?
2GB should be sufficient until Ethereum switches to PoS."

I can mine fine on my RX 470 8GB.
My RX 470 with 4GB refuses to mine cause it has insufficient memory?

Using ethminer-0.9.41-genoil-1.1.7 on Win 10 with Radeon-Crimson-16.7.3-Win10-64Bit.

I tried looking for parameters but I'm having a hard time finding any information on how to set it all up.

On my working 8GB card I use something like this:
" ethminer -SP 2 -G -S pooladdress:port -O wallet.user:pass "

Any idea what parameters to use to mine with my 4GB card ? Or any link with all available parameters? The mining files come with no help file or readme

did you try with   --opencl-platform 1
i have one computer i have to add that option otherwise it s saying  not enough memory

when it s working can you tell us what speed you get with r9 470 ?
and test an older driver 15.x something (actually no need to install that driver but copy the dll's to the ether miner directory
it s somewhere here in the topic, with that driver it s faster on 2 and 3 series (but don't know if it works with 4)

Use this https://github.com/nicehash/cpp-ethereum/releases

setx GPU_FORCE_64BIT_PTR 0
setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_SINGLE_ALLOC_PERCENT 100


ethminer -G  -S  pooladdress:port -O wallet.user:pass

[dolby]

Any idea what parameters to use one gpu from 2 available, I tried --device 0 but it does not work.

[ethminr]

I tried to build the source on windows. What is wrong?

Code:

cmake -G "Visual Studio 12 2013 Win64" D:/cpp-ethereum-master
-- FindWindowsSDK: Detected Visual Studio 2012 or newer, not using the _xp toolset variant: including SDK versions that drop XP support in search!
 - WindowsSDK dirs: C:/Program Files (x86)/Windows Kits/8.1;C:/Program Files (x86)/Microsoft SDKs/Windows/v8.0A;C:/Program Files (x86)/Windows Kits/8.0;C:/Program Files (x86)/Microsoft SDKs/Windows/v8.1A;C:/Program Files (x86)/Microsoft SDKs/Windows/v7.1A
-- ctest path: C:/Program Files (x86)/CMake/bin/ctest.exe
CMake Error at cmake/FindPackageHandleStandardArgs.cmake:138 (message):
  Could NOT find CryptoPP (missing: CRYPTOPP_INCLUDE_DIR CRYPTOPP_LIBRARY)
  (Required is at least version "5.6.2")
Call Stack (most recent call first):
  cmake/FindPackageHandleStandardArgs.cmake:374 (_FPHSA_FAILURE_MESSAGE)
  cmake/FindCryptoPP.cmake:107 (FIND_PACKAGE_HANDLE_STANDARD_ARGS)
  cmake/EthDependencies.cmake:44 (find_package)
  CMakeLists.txt:246 (include)


-- Configuring incomplete, errors occurred!
See also "D:/cpp-ethereum-master/CMakeFiles/CMakeOutput.log".

Run cpp-ethereum/extdep/getstuff.bat first

Re: AMD/Linux:
I don't have Linux binaries. People with Linux build stuff themselves, right?

ok so I if I am right, than we need to compile the source in Visual Studio. But I have an error can't open pthread.lib. I downloaded this lib, but where I need to put it?

That's weird. I heard that before but I don't have the issue. --edit-- oh edit I did have it once, but not anymore ow...-- It's a cmake problem that apparently surfaces on some platforms/systems. The best you can do is to manually renove the reference from the projects that try to link it. Because you don't need it

I have the same problem. @Genoil what do you mean by "to manually renove the reference from the projects that try to link it"

Do you mean removing manually what is causing this cmake/FindPackageHandleStandardArgs.cmake:374 (_FPHSA_FAILURE_MESSAGE)?

[adamvp]

Any idea what parameters to use one gpu from 2 available, I tried --device 0 but it does not work.

--opencl-device

[mwgame1]

Since I've updated to windows anniversary, I've notice my 1060 farm acting strangely using genoil miner.  It taxes my cpu 100% and grinds the system to a halt until it seems a buffer clears, then scrolls as normal hashing happily away at 85 Mhash, then 100% cpu again, and craters the system.  I've had to switch to claymore since he fixed the bug that puts cpu at maximum.  Please look into this.  I dislike being forced to donate.

what gtx 1060 u using, iam buying WINDFORCE OC 6GB tomorrow, do u recommend it ? does it get 18mhs on eth ?

[reb0rn21]

Since I've updated to windows anniversary, I've notice my 1060 farm acting strangely using genoil miner.  It taxes my cpu 100% and grinds the system to a halt until it seems a buffer clears, then scrolls as normal hashing happily away at 85 Mhash, then 100% cpu again, and craters the system.  I've had to switch to claymore since he fixed the bug that puts cpu at maximum.  Please look into this.  I dislike being forced to donate.

what gtx 1060 u using, iam buying WINDFORCE OC 6GB tomorrow, do u recommend it ? does it get 18mhs on eth ?

First batch of cards all come with samsung memory that can go at 2350Mhz or so, there 1070 could give you 33Mhs, 1060 duno 22-23Mhz

if you get the micron memory it will give you 5-10% less

even the ppl who own same card can not help you because all or most vendors switched to micron later (MSI, Gigabyte are the one i know that they switched to micron with 1070 and card bought a week ago) so you will never know

[mwgame1]

Since I've updated to windows anniversary, I've notice my 1060 farm acting strangely using genoil miner.  It taxes my cpu 100% and grinds the system to a halt until it seems a buffer clears, then scrolls as normal hashing happily away at 85 Mhash, then 100% cpu again, and craters the system.  I've had to switch to claymore since he fixed the bug that puts cpu at maximum.  Please look into this.  I dislike being forced to donate.

what gtx 1060 u using, iam buying WINDFORCE OC 6GB tomorrow, do u recommend it ? does it get 18mhs on eth ?

First batch of cards all come with samsung memory that can go at 2350Mhz or so, there 1070 could give you 33Mhs, 1060 duno 22-23Mhz

if you get the micron memory it will give you 5-10% less

even the ppl who own same card can not help you because all or most vendors switched to micron later (MSI, Gigabyte are the one i know that they switched to micron with 1070 and card bought a week ago) so you will never know

alright, thx for the information