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

[skov86]

On PHI2 it's constantly crushing, testing low power usage on tribus and lyra2z algo with 1080ti getting amazing results for 45% power draw!!

[Saphire]

I tested for Lux mining with  PHI2 it gives good hashes for +100mhz core but crashes between 5-10 minutes.
For +50 mhz works fine.
Increasing memory brings nothing.

For Lyra2z it is good and have no problem.

Thank you dev. 

[nitrobg]

Phi2 needs much lower core offset compared to other coins.

[jperser]

Testing version 0.5.1 hash rate is 17% higher than ccminer 2.3, but I need improvements in the API before I can deploy t-rex to my machines:

1. Need to add GPU Temperature to the API.  I have had fans already fail and this is my only indication that is GPU is over-heating.  The increased profit is not worth the risk of burring up a $600 GPU.  ccminer does this already, you can check the code to see where to get the information.

2. Need to add URL to the API.  I switch coins.  I mine multiple coins.  I use the API to double check what each miner is doing.  I need to be able to extract the server name and port from the API (for example: mine.loki.fairpool.xyz:5577).

I also noticed that "gpus" "intensity" is always zero in the API.

We will add this in the future releases with json response. Do you use HTTP api or telnet?

Currently using HTTP.  It was the quickest to get working with my custom scripts.  But I could use either and you should add this information to both (for wider range to customers).  The best solution would to to imitate ccminer's API.  Every monitoring software works with ccminer.

[litosfera]

L2Z, 3x 1080 ti, - i 21, 75% power, mem 0, clock + 160

20180725 08:44:00 [ OK ] 85/85 - 13.87 MH/s
20180725 08:44:01 GPU #0: 4635.46 kH/s
20180725 08:44:01 GPU #1: 4639.08 kH/s
20180725 08:44:01 GPU #2: 4574.66 kH/s
20180725 08:44:01 [ OK ] 86/86 - 13.87 MH/s
20180725 08:44:01 [ OK ] 87/87 - 13.87 MH/s
20180725 08:44:02 [ OK ] 88/88 - 13.87 MH/s


FEW MINUTES LATER

20180725 08:48:19 GPU #0: 4592.62 kH/s
20180725 08:48:19 GPU #1: 4592.13 kH/s
20180725 08:48:20 GPU #2: 4471.35 kH/s
20180725 08:48:26 [ OK ] 406/406 - 13.65 MH/s

[Saphire]

I finally got this running on my ASUS motherboards apperantly only one Nvidia drivers works stable with 19GPU mining board.

T-Rex algorithm performances for 1080Ti:
lyra2z 65TDP +151 core -400 mem = 4288 Kh/s 59'C
lyra2z 65TDP +151 core -502 mem = 4413 Kh/s 58'C
lyra2z 70TDP +151 core -400 mem = crashing
lyra2z 70TDP +151 core -502 mem = 4490 Kh/s 63'C
So far for lyra2z best available miner!

Can you tell me why you lower memory clocks? What does it bring to you?

[Dr_Victor]

I can tell you - when you have your mem decreased gpu takes a higher frequency (proven by experience).

[Saphire]

I can tell you - when you have your mem decreased gpu takes a higher frequency (proven by experience).

I tried with my MSI GTX 1080Ti Gaming
Decreased memory clocks by -400 mhz
and the result is the same.
I got 31.5 Mh/s for 7x1080ti rig befor and after decreasing mem clock i got now 31.6 mh

So if decreasing memory clock provides extra life for my graphic card there is no different with 0mh and -400 mh.

[trexminer]

L2Z, 3x 1080 ti, - i 21, 75% power, mem 0, clock + 160

20180725 08:44:00 [ OK ] 85/85 - 13.87 MH/s
20180725 08:44:01 GPU #0: 4635.46 kH/s
20180725 08:44:01 GPU #1: 4639.08 kH/s
20180725 08:44:01 GPU #2: 4574.66 kH/s
20180725 08:44:01 [ OK ] 86/86 - 13.87 MH/s
20180725 08:44:01 [ OK ] 87/87 - 13.87 MH/s
20180725 08:44:02 [ OK ] 88/88 - 13.87 MH/s


FEW MINUTES LATER

20180725 08:48:19 GPU #0: 4592.62 kH/s
20180725 08:48:19 GPU #1: 4592.13 kH/s
20180725 08:48:20 GPU #2: 4471.35 kH/s
20180725 08:48:26 [ OK ] 406/406 - 13.65 MH/s

Some users faced this issue solved it by running browser in background with opened youtube page with video paused. Try it. Play and pause video then leave browser in background.

[mak013]

Are there news for HiveOS users?

There is an example there. Just choose a custom miner. I`ve tried it, works fine with lyra2z and with phi2 1 rig is crushing all the time. May be i need to decrease OC but i`m using another algo and it does not matter for me.

[ZloiOpex]

i've got about 1550 from 1063 and 14 MH/s from 9x1063 reported by pool, it is more than 140% of ccminer performance $_$

[Hairbo]

Love the software! Keep it up, do you know more or less when full api integration for Awesome miner will occur?

[Ragnarokt]

Are there news for HiveOS users?

There is an example there. Just choose a custom miner. I`ve tried it, works fine with lyra2z and with phi2 1 rig is crushing all the time. May be i need to decrease OC but i`m using another algo and it does not matter for me.

Could you get API data like a hashrate from each card in web-interface?

[GoshaGT]

My miner 0.5.1 does not want to work. Has anyone encountered such a problem?

20180728 11:45:34 WARN: New difficulty: 12 (0.04688)
20180728 11:45:51 WARN: Cuda error unspecified launch failure, device_id=0
20180728 11:45:51 WARN: Cuda error unspecified launch failure, device_id=4
20180728 11:45:51 WARN: Cuda error unspecified launch failure, device_id=1
20180728 11:45:54 WARN: Graceful shutdown...
20180728 11:45:51 WARN: Cuda error unspecified launch failure, device_id=5
20180728 11:45:51 WARN: Cuda error unspecified launch failure, device_id=2
20180728 11:45:52 WARN: Graceful shutdown...
20180728 11:45:54 WARN: Graceful shutdown...
20180728 11:45:51 WARN: Cuda error unspecified launch failure, device_id=3
20180728 11:45:54 ApiServer: stopped listening on 0.0.0.0:4067
20180728 11:45:54 ApiServer: stopped listening on 0.0.0.0:4068

[infPICO]

1080ti runnig really 4,5-4,7 mh/s but very hot, compared to previous miners. guess it is tax for perfomance.

[xlockdownx]

Was anyone able to make this miner work with awesome miner ?

[cost82el]

Was anyone able to make this miner work with awesome miner ?

yes... is supported by default in last awesome miner update

https://bitcointalk.org/index.php?topic=676942.msg42624417#msg42624417

[Vicks-Berg]

My miner 0.5.1 does not want to work. Has anyone encountered such a problem?

20180728 11:45:34 WARN: New difficulty: 12 (0.04688)
20180728 11:45:51 WARN: Cuda error unspecified launch failure, device_id=0
20180728 11:45:51 WARN: Cuda error unspecified launch failure, device_id=4
20180728 11:45:51 WARN: Cuda error unspecified launch failure, device_id=1
20180728 11:45:54 WARN: Graceful shutdown...
20180728 11:45:51 WARN: Cuda error unspecified launch failure, device_id=5
20180728 11:45:51 WARN: Cuda error unspecified launch failure, device_id=2
20180728 11:45:52 WARN: Graceful shutdown...
20180728 11:45:54 WARN: Graceful shutdown...
20180728 11:45:51 WARN: Cuda error unspecified launch failure, device_id=3
20180728 11:45:54 ApiServer: stopped listening on 0.0.0.0:4067
20180728 11:45:54 ApiServer: stopped listening on 0.0.0.0:4068

It may be related to Nvidia drivers version and you may need to install the latest drivers that works with 9.1 or 9.2 cuda support.

You also can to set the gpu back to default clock. Because when the clocks are too high it could cause the issue.

[xlockdownx]

Was anyone able to make this miner work with awesome miner ?

yes... is supported by default in last awesome miner update

https://bitcointalk.org/index.php?topic=676942.msg42624417#msg42624417

Nice !! Thx for info

[tazmako]

Thx for good program for miner.