Topic: T-Rex 0.26.8 ETHW, ETC (+dual mining) GPU miner - page 3. (Read 174685 times)

So, I have a question.  I have a ASUS GeForce GTX 1050 Ti 4GB Phoenix (model PH-GTX1050TI-4G) mining Neoxa on neoxa.herominers.com.  My miner keeps crashing and restarting.  Here is my command line with Windows 11 pro

t-rex.exe -a kawpow -o stratum+tcp://us2.neoxa.herominers.com:1202 -u xxxx -p xxxxx -o stratum+tcp://us.neoxa.herominers.com:1202 -u xxxx -p xxxxx --autoupdate --fan 100 --pl 85 --cclock 200 --mclock 700

Unfortunately, running any flavor of linux is crashing my HP Pavilion 590-p0020 running with a m.2 nvme drive (apparently there seems to be a bug with linux on this model when booting off of an nvme drive).

My system is running a AMD Ryzen 3 2200G with 16GB of memory, using my ASUS GeForce GTX 1050 Ti in the only PCI 3.0 x16 slot on the motherboard.  Any ideas on how to fix this?  It mined ravencoin for more than a year with no issues until the DAG file got too large.  I get the same error on multiple pools.  This also occored with T-Rex NVIDIA GPU miner v0.26.4.

The text t-rex miner is giving me is below:

--------------------------20220820 10:47:34 ---------------------------
Mining at us2.neoxa.herominers.com:1202 [142.132.209.242], diff: 1.07 G
GPU #0: ASUS GTX 1050 Ti - 6.49 MH/s, [T:72C, F:99%], 342/342 R:0%
Shares/min: 0.773 (Avg. 0.401)
Uptime: 1 hour 29 mins 5 secs | Algo: kawpow | Driver: 461.92 | T-Rex 0.26.5
WD: 14 hours 50 mins 17 secs, shares: 342/342 , restarts 19
WD: ======== GPU CRASH LIST ========
WD: GPU#0: 19

20220820 10:47:42 kawpow epoch: 17, block: 134531, period: 44843, diff: 1.07 G
20220820 10:47:46 TREX: Can't find nonce with device [ID=0, GPU #0], cuda exception: CUDA_ERROR_OUT_OF_MEMORY
20220820 10:47:46 WARN: Miner is going to shutdown...
20220820 10:47:46 Main loop finished. Cleaning up resources...
20220820 10:47:46 ApiServer: stopped listening on 127.0.0.1:4067
20220820 10:47:48 T-Rex finished.
20220820 10:47:49 WARN: WATCHDOG: T-Rex does not exist anymore, restarting...
20220820 10:47:51 T-Rex NVIDIA GPU miner v0.26.5  -  [Windows]
20220820 10:47:51 r.b70faa49dd35
20220820 10:47:51
20220820 10:47:51
20220820 10:47:51 NVIDIA Driver v461.92
20220820 10:47:51
20220820 10:47:51 + GPU #0: [00:10.0|1c82] ASUS GeForce GTX 1050 Ti, 4096 MB
20220820 10:47:51
20220820 10:47:51 WARN: DevFee 1% (kawpow)
20220820 10:47:51
20220820 10:47:51 === MAIN POOL === | back to main in 10 mins |
20220820 10:47:51 URL : stratum+tcp://us2.neoxa.herominers.com:1202
20220820 10:47:51 USER: xxxx
20220820 10:47:51 PASS: xxxx
20220820 10:47:51
20220820 10:47:51 URL : stratum+tcp://us.neoxa.herominers.com:1202
20220820 10:47:51 USER: xxxx
20220820 10:47:51 PASS: xxxx
20220820 10:47:51
20220820 10:47:51 Starting on: us2.neoxa.herominers.com:1202
20220820 10:47:51 Using protocol: stratum3.
20220820 10:47:51 Extranonce is set to: 14ba
20220820 10:47:52 Authorizing...
20220820 10:47:52 ApiServer: HTTP server started on 127.0.0.1:4067
20220820 10:47:52 ---------------------------------------------------
20220820 10:47:52 For control navigate to: http://127.0.0.1:4067/trex
20220820 10:47:52 ---------------------------------------------------
20220820 10:47:53 Authorized successfully.
20220820 10:47:53 WARN: NVML: can't get power for GPU #0, not supported
20220820 10:47:53 GPU #0: intensity 20, fan 100%, mclock 700MHz, cclock 200MHz, pl 85%
20220820 10:47:53 kawpow epoch: 17, block: 134531, period: 44843, diff: 1.07 G
20220820 10:47:55 GPU #0: generating DAG 1.13 GB for epoch 17 ...
20220820 10:48:06 GPU #0: DAG generated [crc: 159c23a6, time: 10893 ms], memory left: 2.16 GB
20220820 10:48:16 kawpow epoch: 17, block: 134532, period: 44844, diff: 1.07 G
20220820 10:48:40 kawpow epoch: 17, block: 134533, period: 44844, diff: 1.07 G

--------------------------20220820 10:48:52 ---------------------------
Mining at us2.neoxa.herominers.com:1202 [142.132.209.242], diff: 1.07 G
GPU #0: ASUS GTX 1050 Ti - 6.71 MH/s, [T:67C, F:99%]
Shares/min: 0
Uptime: 1 min 1 sec | Algo: kawpow | Driver: 461.92 | T-Rex 0.26.5
WD: 14 hours 51 mins 35 secs, shares: 342/342 , restarts 20
WD: ======== GPU CRASH LIST ========
WD: GPU#0: 20

Hi, is it possible to dual mine with something else than ETH? I've been dual mining ETH and ALPH and wanted to switch to ETC and ALPH but it's not possible. I thought ETC was mostly the same algorithm than ETH but apparently it's not possible. In regard of ETH switching to PoS it would be interesting to have another dual mining option ready, like ETC+ALPH.

T-Rex 0.26.5
Make LHR unlocker compatible with older drivers (e.g. 470+ on Linux)
Bug fixes:

(ethash) Miner allocates excessive amount of GPU memory which causes 6GB cards to stop mining on Windows earlier than expected

https://github.com/trexminer/T-Rex/releases/tag/0.26.5

Go to first post of this thread …. click on download from trex github in that first post.   Read the words advising whats in the update.

where can I read what's new in version 0.26.5?

Actually he allready answered his own problem.

Trex see the GPU as 6144MB but trex says 39.38MB memory left

Trex sees the card as 6GB but there is not enough headroom as Windows also needs some of that.
Following are the limits with PhoenixMiner:
   Minimum free VRAM in MB (384MB for Windows and 128MB for Linux)

So you can see 39.38MB free VRAM is not going to cut it.

Not so strange, Dag is starting to hit the limit on 6 GB GPUs when using Windows. Some miners are still working for now but it  will
just get worse, a repeat of what happened at 4 GB.

Strange problem.

I was mining on 3 gaming notebook. (GPU: RTX3060 6GB) Win 10 Operating
I deleted the old version of trex. And I downloaded the latest one. I cannot mining on my 3 gaming notebook anymore.
I dont remember which one is the old version of trex. I couldnt find what is the problem.
I tried lolminer for test. lolminer can mine without any issue.
Could you check and tell me what is the problem ? I want to use trex.

GPU HWINFO IMAGE: https://i.imgur.com/MIUSo7U.png

TREX IMAGE: https://i.imgur.com/AaNI51D.png

lol image: https://i.imgur.com/ZActpJM.png

Maybe the problem: Trex see the GPU as 6144MB but trex says 39.38MB memory left after generating DAG.
So, Trex act as if the GPU is 5GB not 6GB.

T-Rex became run with an empty one window how to do that is run in one window?
empty window It`s C:\WINDOWS\system32\cmd.exe

also check if the DAG is higher than the memory of your card.

Windows memory reservation.
Use linux mining os and the problem goes away.

have the same issue, increasing the VM has not helped ... has anyone got any idea how to solve it?

what currency are you trying to mine?

please help me

error "CUDA_ERROR_OUT_OF_MEMORY" with card GTX1650

Bit late, but I fixed it by installing 511.65 driver and updating windows to latest. I noticed when dag rollover happens is usually first mines at reduced hashrate and when system restart occurs it then refuses to mine and throws out "out of memory" error. Ah that point i updated and now it works, for the time being.

what we needed ;

Bug fixes:

    (Windows) High CPU usage on NVIDIA drivers 512+

Traysi I recommended to you doing something with hardware acceleration with new Nvidia DisplayDrivers ver 511.77 and above, you have to decide problem with switch on acceleration mode in Windows and fix high cpu usage I needed that it is working like at Nvidia DisplayDrivers ver 466.77 with hardware acceleration with out heavy CPU usage or I have to go at excavator or another miner. I can not use anymore old Drivers 466 they do not support, not compatible any graphic programs   

yeah... issue is back, gminer has rebooting rig too a few times a day
today I installed hive, works fine at now, hashrate is the same

Yes, now gminer is ok, just a question for how long ... the old version trex solves the problem, but for a while... memory left 9,76mb

i switched rig to gminer and the problem is solved, 10 days uptime, all fine