T-Rex 0.26.8 ETHW, ETC (+dual mining) GPU miner

[DrunkenBrahman]

It starts very unstable, then one card with an error, then another. As a result, I somehow launched it by remohttp://ving the lock on the core, then I returned all the settings and it works fine, but the start of mining itself is a real torment

https://i.postimg.cc/T1J7Xy3t/attach-3.jpg

[1714787921]

1714787921

[1714787921]

1714787921

[1714787921]

1714787921

[FomoMiner]

I can confirm that indeed, it is very unstable. trex dev himself recognize it:

Code:

A few tips if you experience frequent LHR locks:

    Make sure you don't have monitors connected to your LHR cards, or at least turn them off during miner start up to let it calibrate correctly. Avoid using RDP if it puts load on your LHR cards.
    Do not delay applying overclock - this will mess up calibration results and the miner will display unrealistically high LHR tune values.
    Try reducing memory overclock of the GPU that frequently trips LHR.

So what do you do when it has "frequent locks" ? You lower your mem clock, again & again until this release become unsuable and you're better to go back to 0.25.8
nice try tho

[trexminer]

great results BUT it will lock and lock again & again non stop
you will have to lower your oc until it becomes the same hash as previous versions of trex.
nice try but failed (unfortunatly)

Please upload the full log file, I'll have a look. Thanks.

[DrunkenBrahman]

great results BUT it will lock and lock again & again non stop
you will have to lower your oc until it becomes the same hash as previous versions of trex.
nice try but failed (unfortunatly)

Please upload the full log file, I'll have a look. Thanks.

I set the settings of the cards on which the error appeared in the stock 0-0-0-0, but the error still occurs at the moment between "DAG generated" and "using kernel #4"
V 20220417 13:52:41 GPU #4: DAG generated [crc: 50f76ac8, time: 17775 ms], memory left: 2.83 GB
V 20220417 13:55:54 TREX: Can't find nonce with device [ID=3, GPU #3], cuda exception: CUDA_ERROR_LAUNCH_FAILED, try to reduce overclock to stabilize GPU state
X 20220417 13:52:41 GPU #4: using kernel #4

This is first variant, and the second is wnen card just skip "using kernel" step and show no hashrate, 0.00

[dedizonesv2]

great results BUT it will lock and lock again & again non stop
you will have to lower your oc until it becomes the same hash as previous versions of trex.
nice try but failed (unfortunatly)

Please upload the full log file, I'll have a look. Thanks.

I set the settings of the cards on which the error appeared in the stock 0-0-0-0, but the error still occurs at the moment between "DAG generated" and "using kernel #4"
V 20220417 13:52:41 GPU #4: DAG generated [crc: 50f76ac8, time: 17775 ms], memory left: 2.83 GB
V 20220417 13:55:54 TREX: Can't find nonce with device [ID=3, GPU #3], cuda exception: CUDA_ERROR_LAUNCH_FAILED, try to reduce overclock to stabilize GPU state
X 20220417 13:52:41 GPU #4: using kernel #4

This is first variant, and the second is wnen card just skip "using kernel" step and show no hashrate, 0.00

You OC ? and Driver Nvidia vers ?

[DrunkenBrahman]

great results BUT it will lock and lock again & again non stop
you will have to lower your oc until it becomes the same hash as previous versions of trex.
nice try but failed (unfortunatly)

Please upload the full log file, I'll have a look. Thanks.

I set the settings of the cards on which the error appeared in the stock 0-0-0-0, but the error still occurs at the moment between "DAG generated" and "using kernel #4"
V 20220417 13:52:41 GPU #4: DAG generated [crc: 50f76ac8, time: 17775 ms], memory left: 2.83 GB
V 20220417 13:55:54 TREX: Can't find nonce with device [ID=3, GPU #3], cuda exception: CUDA_ERROR_LAUNCH_FAILED, try to reduce overclock to stabilize GPU state
X 20220417 13:52:41 GPU #4: using kernel #4

This is first variant, and the second is wnen card just skip "using kernel" step and show no hashrate, 0.00

You OC ? and Driver Nvidia vers ?

HiveOS 0.6-215@220415 N470.103.01

I fixed problem by adding
"dag-build-mode": "2,2,2,2,2,2"
"kernel": "4,4,4,2,4,4"

[latevillager]

Can someone help me with this error? It happens when internet goes down and then miner restarts due to low hashrate. Upon booting up and still no internet connection the miner window will just close down after final error and do nothing.

Code:

20220415 15:47:16 ERROR: No connection with eth-eu1.nanopool.org:9999, retry after 10 seconds
20220415 15:47:26 WARN: Can't connect to the mining pool, resolve: The requested name is valid, but no data of the requested type was found
20220415 15:47:26 ERROR: No connection with eth-eu1.nanopool.org:9999, retry after 10 seconds
20220415 15:47:36 WARN: Can't connect to the mining pool, resolve: The requested name is valid, but no data of the requested type was found
20220415 15:47:36 ERROR: No connection with eth-eu1.nanopool.org:9999, retry after 10 seconds
20220415 15:47:37 TREX: T-Rex instance wasn't validated

After "TREX: T-Rex instance wasn't validated" window closes and then I have to manually restart after internet connection is back. So annoying.

[soda37]

Can someone help me with this error? It happens when internet goes down and then miner restarts due to low hashrate. Upon booting up and still no internet connection the miner window will just close down after final error and do nothing.

Code:

20220415 15:47:16 ERROR: No connection with eth-eu1.nanopool.org:9999, retry after 10 seconds
20220415 15:47:26 WARN: Can't connect to the mining pool, resolve: The requested name is valid, but no data of the requested type was found
20220415 15:47:26 ERROR: No connection with eth-eu1.nanopool.org:9999, retry after 10 seconds
20220415 15:47:36 WARN: Can't connect to the mining pool, resolve: The requested name is valid, but no data of the requested type was found
20220415 15:47:36 ERROR: No connection with eth-eu1.nanopool.org:9999, retry after 10 seconds
20220415 15:47:37 TREX: T-Rex instance wasn't validated

After "TREX: T-Rex instance wasn't validated" window closes and then I have to manually restart after internet connection is back. So annoying.

miner software mistakenly thinks you are trying to cheat it because it can't connect to dev server and verify SSL certificate (?) sounds like a bug.
ensure srv1.trex-miner.com:443 or srv2.trex-miner.com:443 or srv3.trex-miner.com:443 are reachable.

Not relate to the message above, just want to give a positive feedback:
As a former gminer user, I want to report that T-rex miner works just great and stable even with some of the worst Zotac 3080 Ti cards, hashrate is very stable and equals to a reported hashrate on a pool. Seems like the software is very sophisticated and well-coded.

Edit: small bug report, seems temp limit fan constraint is not working properly, e.g. I set t:57[25-100] and expect min fan speed to be always at 25%, but in fact on the latest 515 drives I see fan 0%. What's wrong? a bug?

[nik-qz]

great results BUT it will lock and lock again & again non stop
you will have to lower your oc until it becomes the same hash as previous versions of trex.
nice try but failed (unfortunatly)

I had similar symptoms several times, I removed all restrictions in the batch file for this card in terms of core blocking and voltage and set the tuner parameter to automatic mode, i.e. to -1, works well, but still sometimes blocks and reduces the selected value
the same symptom of blocking / unlocking is observed when I ask to close the program for half a minute or a minute

[nik-qz]

Please upload the full log file, I'll have a look. Thanks.

whatever parameters I set, for some reason the speed above 86 megahashes does not want to rise in any way, there is a similar video card there is easily 93-94 megahashes on the latest version

[capoeira88]

Please fix dual/triple mining with ZIL! It dies when switching to ZIL.

[rafamonteiroo]

Nicehash unlocked 100% LHR, eagerly awaiting Trex

[smax303]

Same here! Thanks for all your work guys.

[miner29]

Nicehash unlocked 100% LHR, eagerly awaiting Trex

NBminer already released 100% lhr. 

[0verseer]

Already tried nicehash and NBminer and it was unstable as f!ck. Got rebooted from these miners a few times with LHR cards, can't stay mining for long or it will reduce the hash after some time. Have to lower the mem or hash to make it run more stable. Looks like T-Rex isn't really behind competitors if T-Rex can get their miner running more stable than them.

[willi9974]

thanks for the new version, will test ist if it is final.
at the moment i use nbminer and these version works in my case 100% stable and delivers the full hashrate, but i change back to t-tex if it is available

Best regards,
Willi

[MegaVlad]

Latest T-Rex 0.26 is very unstable under RaveOs 8820-2060, even in stock OC.
GPU fault & Reboot every 20-60 minutes.

[arie137]

Latest T-Rex 0.26 is very unstable under RaveOs 8820-2060, even in stock OC.
GPU fault & Reboot every 20-60 minutes.

I'm seeing the same in Windows 10, also with stock OC. Unable to find nonce. Restarts every 10-20 minutes.

[sxemini]

Latest T-Rex 0.26 is very unstable under RaveOs 8820-2060, even in stock OC.
GPU fault & Reboot every 20-60 minutes.

I'm seeing the same in Windows 10, also with stock OC. Unable to find nonce. Restarts every 10-20 minutes.

Now the final version is out - very stable and more poolside hashrate than NBminer

[willi9974]

what exactly has been changed now so that you can get the full performance from the most of all LHR cards.
I am only interested in what has been implemented now, some key / certificate / code from NVIDIA itself.

Has something been leaked?