Topic: GMiner v3.24 Ergo+KASPA+ZIL/KAWPOW/Equihash/Cortex - page 120. (Read 307024 times)
i have "failed to initialize miner on gpu0:asus geforce gtx 10603gb:out of memory" on all 6 cards(asus o3g).devs are you plan to reduce memory usage or is this limitation of the algo?
Beam mining (150_5) is compatible for RTX cards (2070/80/ti..) which is great but the other algos are not.
Any chance you'll make the others (144_5, 192_7...) RTX compatible ?
Today we have received our RTX cards, we did few tests with our miner and RTX2070.Any chance you'll make the others (144_5, 192_7...) RTX compatible ?
Beam v1.15 works fine, Beam v1.16 doesn't work, AION v1.16 works.
In v1.16 for Beam we have optimized CUDA kernels, increase sol/s and also power consumption.
Seems when miner reach some power consumption limit driver return unknown error, we will research more this problem within next days.
I guess we will release v1.17 with RTX compatibility in few days.
+1
+1
+1
+1
Nice improvement on beam
Went from 75-78 -> 92-94 with 6*1070ti
Went from 75-78 -> 92-94 with 6*1070ti
Just an FYI : Tested on my 980Tis, maximum im able to achieve with this card is 7-8 Sols/s. I know its an old card and inefficient but for the moment its on free power
Im more used to this card being equal to a 1070 in most cases and algos hashrate wise, right now my 1060s are quite some distance faster
. Just letting you know as a hail mary, would like to get it running on this rig but not the end of the world if its not possible
Thanks for the great work sofar!
Im more used to this card being equal to a 1070 in most cases and algos hashrate wise, right now my 1060s are quite some distance faster
. Just letting you know as a hail mary, would like to get it running on this rig but not the end of the world if its not possibleThanks for the great work sofar!
what flavor are you talking about? 144,5 or 150,5?
Miner v1.16 does not work with my RTX2080 card. Removed OC still doesn't work.
edit: 0n 150_5
edit: 0n 150_5
Just an FYI : Tested on my 980Tis, maximum im able to achieve with this card is 7-8 Sols/s. I know its an old card and inefficient but for the moment its on free power
Im more used to this card being equal to a 1070 in most cases and algos hashrate wise, right now my 1060s are quite some distance faster. Just letting you know as a hail mary, would like to get it running on this rig but not the end of the world if its not possible
Thanks for the great work sofar!
Im more used to this card being equal to a 1070 in most cases and algos hashrate wise, right now my 1060s are quite some distance faster
. Just letting you know as a hail mary, would like to get it running on this rig but not the end of the world if its not possibleThanks for the great work sofar!
I removed completely overclocking from memory, too, I get memory errors, on spark and beepool, fix it as soon as possible.
Please wait, we will release miner with better performance and fixed bug ("illegal memory access") very soon, I guess today.
I removed completely overclocking from memory, too, I get memory errors, on spark and beepool, fix it as soon as possible.
I use RTX6000, meet this error.
+----------------------------------------------------------------+
| GMiner Equihash Miner v1.16 |
+----------------------------------------------------------------+
Algorithm: Equihash 150,5
Stratum server:
host: stratum+ssl://beam.sparkpool.com:2222
user: XXXXXXXXXXXXXXX
password: x
Power calculator: on
Color output: on
Watchdog: on
API: http://127.0.0.1:10050
Log to file: off
Selected devices: GPU0 GPU1
Temperature limits: 90C 90C
------------------------------------------------------------------
11:30:29 Failed to initialize NVML
Press any key to continue . . .
https://drive.google.com/file/d/1Td7Vk6veEWIutACwL2BmqS3EdBx9ULLZ/view?usp=sharing
https://drive.google.com/file/d/1wKEnGFBYdaABBt5RI_4P_KfAfub0Exx0/view?usp=sharing
+----------------------------------------------------------------+
| GMiner Equihash Miner v1.16 |
+----------------------------------------------------------------+
Algorithm: Equihash 150,5
Stratum server:
host: stratum+ssl://beam.sparkpool.com:2222
user: XXXXXXXXXXXXXXX
password: x
Power calculator: on
Color output: on
Watchdog: on
API: http://127.0.0.1:10050
Log to file: off
Selected devices: GPU0 GPU1
Temperature limits: 90C 90C
------------------------------------------------------------------
11:30:29 Failed to initialize NVML
Press any key to continue . . .
https://drive.google.com/file/d/1Td7Vk6veEWIutACwL2BmqS3EdBx9ULLZ/view?usp=sharing
https://drive.google.com/file/d/1wKEnGFBYdaABBt5RI_4P_KfAfub0Exx0/view?usp=sharing
We just updated GMiner v1.16 to https://www.minermonitoring.com/
Join our discord https://discord.gg/aGGaeqq
Join our discord https://discord.gg/aGGaeqq
Thanks all for feedbacks, we will fix "illegal memory access" bug ASAP.
I can confirm. 1.16 crashes all my 1070 rigs with illegal memory access error. Some are with CUDA 10, others are with CUDA 9, but both get the error. Hope it helps.
Maybe someone can help me, im getting error on V16 Anti-Hacking system detected. I use bitdefender. And i had it to exception but still not work. Tried to use as Admin not work also.
Any idea? Thanks
Also is there a discord server for this miner?
Any idea? Thanks
Also is there a discord server for this miner?
Spoke too soon, starting to get these illegal memory errors on beepol also going back to 1.15 until 1.17 released
illegal instructions even when downgrading to 1.15, so maybe leafpool issue.
Update: same on sparkpool
Update: same on sparkpool
Greetings! Are you have some test channle on telegram or discord? Miner 1.16 began crashed every time... Sparkpool algo 150_50. 1070ti or 1080ti... getting illegal memory access errors http://prntscr.com/m55ohf
first part 1.15 and the second 1.16
first part 1.15 and the second 1.16
1.16 miner was craching my rig and gaming machine and i kept getting illegal memory access errors Mining Beam 150,5... I learned that im only having this issue on Leafpool, after switching to beepool it appears stable and running as 1.15 was but with better hashrates.... Hope Leafpool fixes their API issue or whatever cuz i really like the pool
Hi Zminer777,
I'm experiencing errors mining 150,5 (BEAM).
I thought that memory overclock was causing it, but reverting to default errors still happens.
Also I needed to turn off watchdog, because with it miner just hangs, I can't kill the process only restart helps.
GPUs are 1080ti and 1080
driver version 417.58 (CUDA 10)
Windows Server 2016
Also I noticed that 1080ti how now stabilized GPU Power (voltage) it's like on other Equihash solvers, but 1080 still have quite varying.
I assume that 1080 is not much optmized because it has even same hashrate as 1070 and in every other solver 1080 has higher compred to 1070.
(see the screenshot, GPU0 is 1080ti rest are 1080).
https://i.imgur.com/yqETSbE.png
I'm experiencing errors mining 150,5 (BEAM).
I thought that memory overclock was causing it, but reverting to default errors still happens.
Also I needed to turn off watchdog, because with it miner just hangs, I can't kill the process only restart helps.
GPUs are 1080ti and 1080
driver version 417.58 (CUDA 10)
Windows Server 2016
Quote
22:48:02 Error on GPU3: an illegal memory access was encountered
22:48:02 Stopped Mining on GPU0
22:48:02 Stopped Mining on GPU1
22:48:02 Stopped Mining on GPU2
22:48:02 Stopped Mining on GPU3
22:48:02 Stopped Mining on GPU4
22:48:02 Stopped Mining on GPU5
22:48:02 Stopped Mining on GPU0
22:48:02 Stopped Mining on GPU1
22:48:02 Stopped Mining on GPU2
22:48:02 Stopped Mining on GPU3
22:48:02 Stopped Mining on GPU4
22:48:02 Stopped Mining on GPU5
Also I noticed that 1080ti how now stabilized GPU Power (voltage) it's like on other Equihash solvers, but 1080 still have quite varying.
I assume that 1080 is not much optmized because it has even same hashrate as 1070 and in every other solver 1080 has higher compred to 1070.
(see the screenshot, GPU0 is 1080ti rest are 1080).
https://i.imgur.com/yqETSbE.png
So even without memory OC i'm getting ilegal memory access errors on 1060s.
Do you plan to support the Cuckcoo Cycle algo?
Jump to: