|
I'm getting 58.8Mhs@135Wt with Fury cards in lyra2rev3. Though they are 60CU, not 56.
|
|
|
|
|
I believe its just nicehash servers and not gminer. Both europe and usa "die" from time to time today, you can see that when their h\r drops to almost 0, it I doubt 100% of miners use gminer.
|
|
|
|
|
Could somebody give an example of config for nicehash please? Can't seem to make it work on my own)) Also what command should I use to specify GPUs: -device? Thanks.
|
|
|
|
|
iforgive
On my rigs of mixed 1070\1080\1080Ti memory allocation is ~3.6Gb on all cards, guess that's why.
|
|
|
|
this version is a complete bust for me for all my nvidia cards for nicehash for windows 10. Version 1.24 worked for my 1080ti/1070 for nicehash somewhat, However, version 1.25 i get:
connect to beam.usa.nicehash.com:3370 Failed: stream truncated
I get this error for all 1080ti/1070/1060 on four different systems with windows10 using 1.25. i get zero hashrates. this is for beam.. i.e. equihash150
zhash, i.e. equihash144 works fine on nicehash.
v1.25 works perfectly fine on nicehash |
|
|
|
|
I've clocked down memory to -500 + P2 state for x16r. It has no effect on hashrate(well, mb 0.1mhs compared to default mem. clocks), but it improves power consumption.
|
|
|
|
Thank you for the report within 1 day I'll publish version 1.14 with fix of this bug.
Thank you! Could you may be make Gminer split(cut) log file on restarts? It's writing single file continuously which makes it harder to track errors. |
|
|
|
Please show me your launch arguments and log from miner, I'll check.
These are default NHLM settings(from config files): gminer.exe --devices 0 1 2 3 4 5 --pec --pers auto --algo 144_5 --server zhash.hk.nicehash.com --user wallet.rigname --pass x --port 3369 --server zhash.hk.nicehash.com --user wallet.rigname --pass x --port 3369 --server zhash.in.nicehash.com --user wallet.rigname --pass x --port 3369 --server zhash.usa.nicehash.com --user wallet.rigname --pass x --port 3369 --server zhash.jp.nicehash.com --user wallet.rigname --pass x --port 3369 --server zhash.br.nicehash.com --user wallet.rigname --pass x --port 3369 --api 4003 I don't add any other parametres in settings. Here is log: https://yadi.sk/i/8kDAx37z_F-WqQ |
|
|
|
|
Im mining e144.5 on nicehash and getting "connection error: mailformed server message" like once in 5 mins at least, wtf is that?
|
|
|
|
|
Power consumption has become very spiky on x16r with last miner update.
|
|
|
|
how set intensity ?
GPU utilization is always 100%, there is no options to adjust intensity. Why do you need to decrease GPU usage? Cos if u mine on homePC even browsing becomes impossible... |
|
|
|
|
Is there a command line to set memory clocks for cards in this miner? When mining on nicehash its kinda necessary since most other profitable algos benefit on mem. downclock\stock clock, while equihash profits on mem. OC.
|
|
|
|
CryptoDredge now performs a bit better in skunk than T-rex  |
|
|
|
Never thought i could reach 40mh/s but thanks to this miner it's possible.
Depending on sequence I get 50mhs+ on 1080Ti's and 40mhs+ on 1080. On avg.(using -N 999999999) I get 26mhs and 20mhs respectively. On the other hand when i tried the latest cuda with latest driver with trex 0.7.2 it's jumping from 22mhs to 40mhs
0.7.x will improve your h\r over 0.6.x. Cuda10 drivers over cuda9.2 on the other hand won't... at least for now. |
|
|
|
It's definitely cuda 9.2 with 399.24 drivers (cuda 10 is experimental and slower).
yeah, I've also tried cuda 10 with 400 drivers and h\r was a blit slower with gtx1070\1080(ti), so I rolled back to 399.24 |
|
|
|
|
xmr-stak won't work on my R9 Furies, but xmrig did, I got 840kH\s@130Wt per card using 1920 intensity(tried 2024, 1984, 1952 but it didn't work).
ps all my furies have 60CU(50% unlock)
|
|
|
|
can one set the I higher than 6?
-i, --intensity Mining intensity (0 - 6)
Im running phi2 windows 10 on 1070's
Yes, at least for algos like lyra2z/rev2, neoscrypt |
|
|
|
@echo off
:loop
wildrig.exe --algo x16r --opencl-launch=18x128 --url=stratum+tcp://x16r.eu.nicehash.com:3366 --user=__ --pass=x
if ERRORLEVEL 1000 goto custom
timeout /t 5
goto loop
:custom
echo Custom command here
timeout /t 5
goto loop |
|
|
|
|
Initially they run stock clock and undervolted to be stable in games plus they all 60CU(50% unlock), so voltage is very high for mining anyway. I've tried stock voltage which is over 1200mv - it doest matter, besides even undervolted cards don't crash so being undervolted seems like impossible cause to me - they would have just crash after starting mining, but the thing is they don't even start in the first place.
Ps I tried mining x16r with sgminer - no problems there. So no issues with the cards themselves I guess.
|
|
|
|
|
I tried --opencl-launch 18x0 and 18x128
|
|
|
|
|
Show Posts
