EWBF's Cuda Equihash Miner 0.6 - page 5.

Overdrive5

member

Activity: 113

Merit: 10

Quote from: fistsofgod on October 13, 2018, 02:46:31 PM

EWBF_, can you add ability for miner to exit as soon as there is an error with the cuda cores? currently when there is a memory error or something like that it tries to restart the GPUs and stuff. I'd prefer the miner just quit so computer restarts faster. everytime the miner tries to restart GPUs one of them doesnt mine or the computer restarts anyways. It would be more efficient if the program exited right away

fistsofgod,

Have you tried the --eexit 3 setting?

Otherwise you will need a separate software watchdog that monitors GPU utilization.

Hope this helps

Overdrive5

fistsofgod

newbie

Activity: 164

Merit: 0

EWBF_, can you add ability for miner to exit as soon as there is an error with the cuda cores? currently when there is a memory error or something like that it tries to restart the GPUs and stuff. I'd prefer the miner just quit so computer restarts faster. everytime the miner tries to restart GPUs one of them doesnt mine or the computer restarts anyways. It would be more efficient if the program exited right away

BitSquare

newbie

Activity: 28

Merit: 0

Quote from: markiz73 on October 09, 2018, 07:52:51 AM

memory +1000 ? Am I right?
This is a very large overclocking for this graphics card. Try + 700 and add gradually if it is stable.
Perhaps it is better to keep the video card and not to put above 800.

My nVidia 1080Ti can only +700.

hell_slayer

full member

Activity: 728

Merit: 139

Quote from: uralcryptocoin on October 12, 2018, 12:52:57 PM

Looks really strange...
I try to suppose the following: 200W is enough low power limit for GTX1080ti with typical PL at 250W. So what is your OC settings and working core\memory clocks?
Maybe try to reduce OC or increase PL to 220W or more

Nothing special : +150 core clock , +500 mem , 80% power limit . Now I use version 0.5 again and everything works well, I just wonder what has changed in the new version and could cause such strange system behavior.

uralcryptocoin

sr. member

Activity: 1042

Merit: 328

SIGNATURE CREATION by uralcryptocoin

Quote from: hell_slayer on October 10, 2018, 12:27:30 PM

....
Only a single gpu is running on my gaming PC powered by 600w platinum psu . GPU power consumption is locked at 200W . Total energy consumption of the system is 410w , so it's not a problem

Looks really strange...
I try to suppose the following: 200W is enough low power limit for GTX1080ti with typical PL at 250W. So what is your OC settings and working core\memory clocks?
Maybe try to reduce OC or increase PL to 220W or more

hell_slayer

full member

Activity: 728

Merit: 139

Quote from: uralcryptocoin on October 10, 2018, 12:05:30 PM

Which error appears miner closing with? Use logging to get error message

No error messages . Step 1 , 2 are working fine without any errors , but step 3 causes system to reboot . As I said before 0.5 version is working fine , so I guess it's not a hardware issue

Quote from: uralcryptocoin on October 10, 2018, 12:05:30 PM

There is a lack of info. What is your rig configuration? What hardware espessially PSU and number of GPUs? Would you check power consumption of the GPU? NVSMI tool will help

Only a single gpu is running on my gaming PC powered by 600w platinum psu . GPU power consumption is locked at 200W . Total energy consumption of the system is 410w , so it's not a problem

uralcryptocoin

sr. member

Activity: 1042

Merit: 328

SIGNATURE CREATION by uralcryptocoin

Quote from: hell_slayer on October 10, 2018, 11:44:06 AM

Got a problem with new 0.6 version and1080ti. Win 10 system reboots in a specific scenario : 1) miner is running on 210.9 algo (I don't try on other algo's , so it maybe unimportant detail) on a full load 2)miner is closed and new miner with --intensity 2 option is executed 3) --intensity 2 miner is closed and full load miner is executed ===> system reboot Undecided

previous 0.5 version worked without errors , does it make sense to try other --intensity parameters?

There is a lack of info. What is your rig configuration? What hardware espessially PSU and number of GPUs? Would you check power consumption of the GPU? NVSMI tool will help

Quote from: hell_slayer on October 10, 2018, 11:44:06 AM

.... 2)miner is closed and new miner with --intensity 2 option is executed ....

Which error appears miner closing with? Use logging to get error message

hell_slayer

full member

Activity: 728

Merit: 139

Got a problem with new 0.6 version and1080ti. Win 10 system reboots in a specific scenario : 1) miner is running on 210.9 algo (I don't try on other algo's , so it maybe unimportant detail) on a full load 2)miner is closed and new miner with --intensity 2 option is executed 3) --intensity 2 miner is closed and full load miner is executed ===> system reboot Undecided

previous 0.5 version worked without errors , does it make sense to try other --intensity parameters?

fistsofgod

newbie

Activity: 164

Merit: 0

Quote from: swisstrader on October 09, 2018, 07:52:00 AM

Quote from: fistsofgod on October 08, 2018, 05:33:45 PM

Quote from: swisstrader on October 08, 2018, 04:23:31 PM

Second crash occurred with my 5 GPU rig, each 1070Ti with PL=120 W and OC=200/1000.

make sure your page file is big enough. i set mine to 7168 mb per card. then start lowering your OC until stability increases. I'd also decrease power for higher efficiency

Where do I have to set the page file settings? It is Ubuntu. My knowledge about page file how to set is in Windows, no clue about settings in Ubunto.
Many thanks in advance for your help.

Found this here:

https://i.imgur.com/Lzf6r1l.jpg

Yeah i dont know about ubuntu, but decrease clock for stability and decrease power for efficiency.

efficiency > hashrate

uralcryptocoin

sr. member

Activity: 1042

Merit: 328

SIGNATURE CREATION by uralcryptocoin

Quote from: swisstrader on October 08, 2018, 04:23:31 PM

Second crash occurred with my 5 GPU rig, each 1070Ti with PL=120 W and OC=200/1000.
....

It's typical "overclock" error. Reduce core clock at first. +200 is too much. Rare GPUs can work at +200 core clock especially at enough low PL

markiz73

hero member

Activity: 1190

Merit: 641

memory +1000 ? Am I right?
This is a very large overclocking for this graphics card. Try + 700 and add gradually if it is stable.
Perhaps it is better to keep the video card and not to put above 800.

swisstrader

member

Activity: 128

Merit: 10

Quote from: fistsofgod on October 08, 2018, 05:33:45 PM

Quote from: swisstrader on October 08, 2018, 04:23:31 PM

Second crash occurred with my 5 GPU rig, each 1070Ti with PL=120 W and OC=200/1000.

make sure your page file is big enough. i set mine to 7168 mb per card. then start lowering your OC until stability increases. I'd also decrease power for higher efficiency

Where do I have to set the page file settings? It is Ubuntu. My knowledge about page file how to set is in Windows, no clue about settings in Ubunto.
Many thanks in advance for your help.

Found this here:

fistsofgod

newbie

Activity: 164

Merit: 0

Quote from: swisstrader on October 08, 2018, 04:23:31 PM

Second crash occurred with my 5 GPU rig, each 1070Ti with PL=120 W and OC=200/1000.

https://i.imgur.com/kBSV019.jpg

make sure your page file is big enough. i set mine to 7168 mb per card. then start lowering your OC until stability increases. I'd also decrease power for higher efficiency

swisstrader

member

Activity: 128

Merit: 10

Second crash occurred with my 5 GPU rig, each 1070Ti with PL=120 W and OC=200/1000.

markiz73

hero member

Activity: 1190

Merit: 641

Quote from: Valer4ik on October 08, 2018, 11:37:03 AM

Who, with such a tariff, is invested in farms on the GPU now? More expensive 3p / kW already beyond reason. Ashiki is now much more interesting at a low price, at a rural tariff and how home heating is the most. Drying the basement, used as heating and drying in the production of lumber - super. And payback around the year, and considering that under normal conditions of operation and maintenance for 2-3 years, asik plows a minimum, all the rules with mining

Asics very quickly lose their relevance now.
Look: https://www.asicminervalue.com
If a video card can be sold after a payback, then the asyk usually turns into a piece of iron.
It may be beneficial to use asic for heating in cold countries, but this requires additional investments and significantly increases the payback period.
I still do not see prospects in any kind of mining. Old equipment works because it has paid for itself several times.
I do not buy new equipment since 2018.

Valer4ik

full member

Activity: 490

Merit: 101

Who, with such a tariff, is invested in farms on the GPU now? More expensive 3p / kW already beyond reason. Ashiki is now much more interesting at a low price, at a rural tariff and how home heating is the most. Drying the basement, used as heating and drying in the production of lumber - super. And payback around the year, and considering that under normal conditions of operation and maintenance for 2-3 years, asik plows a minimum, all the rules with mining

Dr_Victor

sr. member

Activity: 954

Merit: 250

Quote from: mySmile28 on October 07, 2018, 07:57:07 AM

which version of driver will be best for 1063 on 144_5?

399.24

markiz73

hero member

Activity: 1190

Merit: 641

Quote from: Chimer47 on October 06, 2018, 02:47:16 PM

~snip
Hmm still dissappointed. As markiz73 wrote, 2080ti delivers about 57.7 Mh/s with EthHash and costs 1200$ and above. For that money I can buy 3 RX Vegas each of them delivering about 42Mh/s when optimized. So in total I get 126Mh/s with 3 Vegas for 1200$. So you can say Vega beats the shit out of Turing for the same hardware invest. So no need to buy any Turing at all so far. Too bad I had high hopes for new Nvidia Generation Sad

For $ 1,200, you can buy 7 RX570 4.0 GB and have a hashrate of 205-210 Mh/s

And on the secondary market, for $ 1,160, you can buy a used farm with a hashrate of 175 Mh/s.
1 dollar = 67 rubles
https://www.avito.ru/ulyanovsk/tovary_dlya_kompyutera/mayning_ferma_175_mh_1720469603
With the current rates, new video cards have a very long payback period.
I would advise you to look at the secondary market.

uralcryptocoin

sr. member

Activity: 1042

Merit: 328

SIGNATURE CREATION by uralcryptocoin

Quote from: fistsofgod on October 05, 2018, 01:01:39 PM

is the mining software able to send a power limit to the gpu? or can the miner execute an nvidiainspector command line? I want to change my OC settings based on pool difficulty. i just want to change powerlimit as diff goes up and down

It's very specific feature. The most users don't need that. Because of profitability of these actions is enough disputable. But you can create your own VB script that will analyze console output of the miner and will correct GPU setting by nv-smi cli tool

CarlOrff

sr. member

Activity: 826

Merit: 440

Very great job with the 0.6 release : my old GPUs with 2 GB VRAM can now mining BTCz, like before the previous algo version !
Congratulation EWBF :-).

Topic: EWBF's Cuda Equihash Miner 0.6 - page 5. (Read 85805 times)