Pages:
Author

Topic: lolMiner 1.64: Fastest Kaspa miner in market - page 24. (Read 155326 times)

jr. member
Activity: 85
Merit: 2
guys third gpu shows 0, any ideas? thanks

That is GPU 0 that is showing 0.00 hash.
Probably need to watch miner output when it starts to see error message why its not hashing.

Is gpu0 a 4gb rx580?  Also is your driver anywhere near current?

Windows uses some of the memory of the first video card for some reason (linux does not suffer this).

Post up what the miner is showing when spinning up and any error message it is showing about gpu0.
Also your command line.


1) no its 8gb
2) driver is crimson beta old version (but it worked 1 week ago with no problems)
3) miner stil works with no error message
4) command line? can you explain what this means and how to post it

p.s i have updated to new driver but still the same.. and then downgraded again

thanks for your reply
full member
Activity: 1264
Merit: 138
guys third gpu shows 0, any ideas? thanks

That is GPU 0 that is showing 0.00 hash.
Probably need to watch miner output when it starts to see error message why its not hashing.

Is gpu0 a 4gb rx580?  Also is your driver anywhere near current?

Windows uses some of the memory of the first video card for some reason (linux does not suffer this).

Post up what the miner is showing when spinning up and any error message it is showing about gpu0.
Also your command line.


jr. member
Activity: 85
Merit: 2
guys third gpu shows 0, any ideas? thanks




jr. member
Activity: 152
Merit: 3
Quote
Are you testing with 3060 rigs? Only for the driver choose   Smiley Curiosity


No, i have only 3060Ti, 3070 and 3080 ^_^ sorry



Really nice rigs... I found that lastest drivers are better in case a Riser fail... What I found is that, lastest drivers can ommit better waiting for the riser information. at least in old rigs where the risers are not in best situation...


well I have last generation risers with 8 capacitor + the small coil above 150 (often 100).
I find that I have better performance with the nvidia 460.39 drivers

I redid the tests, I really have a bad hashrate / share / watt with lolminer, I consume more for less sharing
member
Activity: 639
Merit: 19
Quote
Are you testing with 3060 rigs? Only for the driver choose   Smiley Curiosity


No, i have only 3060Ti, 3070 and 3080 ^_^ sorry



Really nice rigs... I found that lastest drivers are better in case a Riser fail... What I found is that, lastest drivers can ommit better waiting for the riser information. at least in old rigs where the risers are not in best situation...
jr. member
Activity: 152
Merit: 3
Quote
Are you testing with 3060 rigs? Only for the driver choose   Smiley Curiosity


No, i have only 3060Ti, 3070 and 3080 ^_^ sorry

member
Activity: 639
Merit: 19
Testing Silicion Lottery...

3080 : 96Mhs at 201W with Memory Temp 78C

Same Model : Without pushing it 205W and 82C Memory Temp... (I haven't push it up yet)


I tested the last version lolminer, and it's a shame because under ETC and ETH we consume more WATT than other software, I have uploaded this internally, I hope you will find how optimized

How can consume more watts?? are you using Fixed Core or Power Limit? It hasn't been any update in ETHASH only for the 3060...

Yes i try 24H and lolminer more +-100w

Exemple: lolminer 3827w, other miner 3780w
Lolminer: 4014W
Today add cards other miner: 3914W

i verified on the wall + linky (an electricity company meter)

it is not a big consumption either but just sometimes a little tour in the code to optimize it

:*

i use locked core in hiveos


Really interesting... in my case comparative with CUDA 1.31 for example in 3070 and 3080 are 2-3W less with some others miners.... It is true, I use low Core with low Memory and I can reduce Watts and stress lot less the GPUs.. About the new card is nearly I  lolMiner 190W and 134W in the other miner... I think that's to much difference?

I rollback driver nvidia last version to 460.39 and I will redo the test between lolminer and another that I use


Are you testing with 3060 rigs? Only for the driver choose   Smiley Curiosity
jr. member
Activity: 152
Merit: 3
Testing Silicion Lottery...

3080 : 96Mhs at 201W with Memory Temp 78C

Same Model : Without pushing it 205W and 82C Memory Temp... (I haven't push it up yet)


I tested the last version lolminer, and it's a shame because under ETC and ETH we consume more WATT than other software, I have uploaded this internally, I hope you will find how optimized

How can consume more watts?? are you using Fixed Core or Power Limit? It hasn't been any update in ETHASH only for the 3060...

Yes i try 24H and lolminer more +-100w

Exemple: lolminer 3827w, other miner 3780w
Lolminer: 4014W
Today add cards other miner: 3914W

i verified on the wall + linky (an electricity company meter)

it is not a big consumption either but just sometimes a little tour in the code to optimize it

:*

i use locked core in hiveos


Really interesting... in my case comparative with CUDA 1.31 for example in 3070 and 3080 are 2-3W less with some others miners.... It is true, I use low Core with low Memory and I can reduce Watts and stress lot less the GPUs.. About the new card is nearly I  lolMiner 190W and 134W in the other miner... I think that's to much difference?

I rollback driver nvidia last version to 460.39 and I will redo the test between lolminer and another that I use
member
Activity: 639
Merit: 19
Testing Silicion Lottery...

3080 : 96Mhs at 201W with Memory Temp 78C

Same Model : Without pushing it 205W and 82C Memory Temp... (I haven't push it up yet)



I tested the last version lolminer, and it's a shame because under ETC and ETH we consume more WATT than other software, I have uploaded this internally, I hope you will find how optimized

How can consume more watts?? are you using Fixed Core or Power Limit? It hasn't been any update in ETHASH only for the 3060...

Yes i try 24H and lolminer more +-100w

Exemple: lolminer 3827w, other miner 3780w
Lolminer: 4014W
Today add cards other miner: 3914W

i verified on the wall + linky (an electricity company meter)

it is not a big consumption either but just sometimes a little tour in the code to optimize it

:*

i use locked core in hiveos


Really interesting... in my case comparative with CUDA 1.31 for example in 3070 and 3080 are 2-3W less with some others miners.... It is true, I use low Core with low Memory and I can reduce Watts and stress lot less the GPUs.. About the new card is nearly I  lolMiner 190W and 134W in the other miner... I think that's to much difference?
jr. member
Activity: 152
Merit: 3
Testing Silicion Lottery...

3080 : 96Mhs at 201W with Memory Temp 78C

Same Model : Without pushing it 205W and 82C Memory Temp... (I haven't push it up yet)



I tested the last version lolminer, and it's a shame because under ETC and ETH we consume more WATT than other software, I have uploaded this internally, I hope you will find how optimized

How can consume more watts?? are you using Fixed Core or Power Limit? It hasn't been any update in ETHASH only for the 3060...

Yes i try 24H and lolminer more +-100w

Exemple: lolminer 3827w, other miner 3780w
Lolminer: 4014W
Today add cards other miner: 3914W

i verified on the wall + linky (an electricity company meter)

it is not a big consumption either but just sometimes a little tour in the code to optimize it

:*

i use locked core in hiveos
member
Activity: 639
Merit: 19
Testing Silicion Lottery...

3080 : 96Mhs at 201W with Memory Temp 78C

Same Model : Without pushing it 205W and 82C Memory Temp... (I haven't push it up yet)



I tested the last version lolminer, and it's a shame because under ETC and ETH we consume more WATT than other software, I have uploaded this internally, I hope you will find how optimized

How can consume more watts?? are you using Fixed Core or Power Limit? It hasn't been any update in ETHASH only for the 3060...
jr. member
Activity: 152
Merit: 3
Testing Silicion Lottery...

3080 : 96Mhs at 201W with Memory Temp 78C

Same Model : Without pushing it 205W and 82C Memory Temp... (I haven't push it up yet)



I tested the last version lolminer, and it's a shame because under ETC and ETH we consume more WATT than other software, I have uploaded this internally, I hope you will find how optimized
member
Activity: 639
Merit: 19
Testing Silicion Lottery...

3080 : 96Mhs at 201W with Memory Temp 78C

Same Model : Without pushing it 205W and 82C Memory Temp... (I haven't push it up yet)

member
Activity: 639
Merit: 19

It happens on all miners with the 5700 and old HiveOS version...
Install HiveOS with Kernel 5.4.0-hiveos #108, as I see you have a very old Driver AMD 20.20 the most stable is 20.40 that hasn't go a memory bug lead with the 5700xt.
If you take the last image realase of update inside HiveOS you will avoid that Free Ram that you have :-)

It was solved on this realease:

0.6-190linux
2021-01-08
LINUX SYSTEM IMAGE
New Stable branch
Hive Linux client: v0.6-190@201106
Universal boot mode: BIOS | UEFI
Linux kernel: v5.4.80-hiveos
Drivers version: AMD v20.40, Nvidia: v455.45.01
Universal image with "support from the box" AMD Polaris/Vega/Navi and Nvidia 10xx/16xx/20xx/30xx


Its not possible, i was using TRM for at least 3-4 months and after that 1 month PM without any memory problems. It started when i changed to lolMiner.
And if i change back to TRM, there is no problem with memory. It shows 6.6gb free ram all the time. (And if you havent noticed its RAM not VRAM)

(5.8gb now :-) )

But i can try to update that driver when i visit the mining site. Last time i tried 20.4, it didnt work, cant remember which card didn't show up.

and thanks for advice.

Driver 20.20 has a memory leak... is not miner depence.  Just update and you will have all solved... but update your complete image to 5.4.0-hiveos #108 the last avaliable in website. You will continue having all the same copying the rig.conf ☺

Updated driver to 20.40 and it broke everything, no OC or fan control on RX5600 and RX5700 -> Big problems with over heating... Now back at 20.20, everything working again, added auto restart every week for miner so no more problem Smiley
With 20.20 also hashrate higher and lower power consumption on RX570 and RX580.


I hope you update the image of HiveOS soon as I said you  Smiley then all will go perfectly ... meanwhile you will continue having the autoreboot when free memory ends...
DrX
member
Activity: 233
Merit: 20

It happens on all miners with the 5700 and old HiveOS version...
Install HiveOS with Kernel 5.4.0-hiveos #108, as I see you have a very old Driver AMD 20.20 the most stable is 20.40 that hasn't go a memory bug lead with the 5700xt.
If you take the last image realase of update inside HiveOS you will avoid that Free Ram that you have :-)

It was solved on this realease:

0.6-190linux
2021-01-08
LINUX SYSTEM IMAGE
New Stable branch
Hive Linux client: v0.6-190@201106
Universal boot mode: BIOS | UEFI
Linux kernel: v5.4.80-hiveos
Drivers version: AMD v20.40, Nvidia: v455.45.01
Universal image with "support from the box" AMD Polaris/Vega/Navi and Nvidia 10xx/16xx/20xx/30xx

Its not possible, i was using TRM for at least 3-4 months and after that 1 month PM without any memory problems. It started when i changed to lolMiner.
And if i change back to TRM, there is no problem with memory. It shows 6.6gb free ram all the time. (And if you havent noticed its RAM not VRAM)

(5.8gb now :-) )

But i can try to update that driver when i visit the mining site. Last time i tried 20.4, it didnt work, cant remember which card didn't show up.

and thanks for advice.

Driver 20.20 has a memory leak... is not miner depence.  Just update and you will have all solved... but update your complete image to 5.4.0-hiveos #108 the last avaliable in website. You will continue having all the same copying the rig.conf ☺

Updated driver to 20.40 and it broke everything, no OC or fan control on RX5600 and RX5700 -> Big problems with over heating... Now back at 20.20, everything working again, added auto restart every week for miner so no more problem Smiley
With 20.20 also hashrate higher and lower power consumption on RX570 and RX580.
member
Activity: 639
Merit: 19
Hi I mine using lolminer (SMOS) zombie mode on a rig with rx 580 4gb and a Vega 64 but the vega never start mining do you know if it is possible to do it on the same rig ?



Yes it is possible, which is your config line to check what's happening
member
Activity: 639
Merit: 19

It happens on all miners with the 5700 and old HiveOS version...
Install HiveOS with Kernel 5.4.0-hiveos #108, as I see you have a very old Driver AMD 20.20 the most stable is 20.40 that hasn't go a memory bug lead with the 5700xt.
If you take the last image realase of update inside HiveOS you will avoid that Free Ram that you have :-)

It was solved on this realease:

0.6-190linux
2021-01-08
LINUX SYSTEM IMAGE
New Stable branch
Hive Linux client: v0.6-190@201106
Universal boot mode: BIOS | UEFI
Linux kernel: v5.4.80-hiveos
Drivers version: AMD v20.40, Nvidia: v455.45.01
Universal image with "support from the box" AMD Polaris/Vega/Navi and Nvidia 10xx/16xx/20xx/30xx

Its not possible, i was using TRM for at least 3-4 months and after that 1 month PM without any memory problems. It started when i changed to lolMiner.
And if i change back to TRM, there is no problem with memory. It shows 6.6gb free ram all the time. (And if you havent noticed its RAM not VRAM)

(5.8gb now :-) )

But i can try to update that driver when i visit the mining site. Last time i tried 20.4, it didnt work, cant remember which card didn't show up.

and thanks for advice.

Driver 20.20 has a memory leak... is not miner depence.  Just update and you will have all solved... but update your complete image to 5.4.0-hiveos #108 the last avaliable in website. You will continue having all the same copying the rig.conf ☺
full member
Activity: 194
Merit: 100
Hi I mine using lolminer (SMOS) zombie mode on a rig with rx 580 4gb and a Vega 64 but the vega never start mining do you know if it is possible to do it on the same rig ?

DrX
member
Activity: 233
Merit: 20

It happens on all miners with the 5700 and old HiveOS version...
Install HiveOS with Kernel 5.4.0-hiveos #108, as I see you have a very old Driver AMD 20.20 the most stable is 20.40 that hasn't go a memory bug lead with the 5700xt.
If you take the last image realase of update inside HiveOS you will avoid that Free Ram that you have :-)

It was solved on this realease:

0.6-190linux
2021-01-08
LINUX SYSTEM IMAGE
New Stable branch
Hive Linux client: v0.6-190@201106
Universal boot mode: BIOS | UEFI
Linux kernel: v5.4.80-hiveos
Drivers version: AMD v20.40, Nvidia: v455.45.01
Universal image with "support from the box" AMD Polaris/Vega/Navi and Nvidia 10xx/16xx/20xx/30xx

Its not possible, i was using TRM for at least 3-4 months and after that 1 month PM without any memory problems. It started when i changed to lolMiner.
And if i change back to TRM, there is no problem with memory. It shows 6.6gb free ram all the time. (And if you havent noticed its RAM not VRAM)

(5.8gb now :-) )

But i can try to update that driver when i visit the mining site. Last time i tried 20.4, it didnt work, cant remember which card didn't show up.

and thanks for advice.
member
Activity: 639
Merit: 19
Hi,

I have noticed a problem with lolMiner 1.3, it's eating all the ram slowly. (all 8gb after about one week)
After there is no more free ram, computer crashes.
Anyone else with same problem?
OS: HiveOS 0.6-204@210608
Ram: 8Gb
MB: Asrock H81 Pro BTC
GPU: 2xRX580,RX570, RX5700 and RX5600XT
Amd Driver : 20.20

Install another 8 GB and reboot every 2 weeks.
Seriously, what are you mining?
My RX 470 4 GB video cards mine Ethereum Classic + Zill, on a farm with 4 GB of RAM and there are no problems. Try using an older Hive OS release.
I have never had any control over the amount of free RAM

Forgot to say that i'm mining Eth and lolMiner is only miner with this problem, working fine on TRM and PM. (But TRM RX570 starts making invalid shares after 24h, so i have to restart every 24h to avoid it Cheesy )
Maybe i keep restarting the miner every week, not a big problem.
(Current free memory 6.1)

It happens on all miners with the 5700 and old HiveOS version...
Install HiveOS with Kernel 5.4.0-hiveos #108, as I see you have a very old Driver AMD 20.20 the most stable is 20.40 that hasn't go a memory bug lead with the 5700xt.
If you take the last image realase of update inside HiveOS you will avoid that Free Ram that you have :-)

It was solved on this realease:

0.6-190linux
2021-01-08
LINUX SYSTEM IMAGE
New Stable branch
Hive Linux client: v0.6-190@201106
Universal boot mode: BIOS | UEFI
Linux kernel: v5.4.80-hiveos
Drivers version: AMD v20.40, Nvidia: v455.45.01
Universal image with "support from the box" AMD Polaris/Vega/Navi and Nvidia 10xx/16xx/20xx/30xx

Pages:
Jump to: