[ANN] TeamRedMiner v0.10.10 - Ironfish/Kaspa/ZIL/Kawpow/Etchash and More - page 10.

Cryptoseekert

newbie

Activity: 21

Merit: 1

Quote from: Bokiman on May 26, 2021, 02:25:32 AM

Quote from: Cryptoseekert on May 22, 2021, 07:39:56 PM

Hello guys

i hope you are all doing good

one question:

for ravencoin mining on windows 10, amd rx 570 4gb, what is the best way to control fans, voltages and clocks?

can i do it from the miner (like eth mining on phoenix or old claymore, where i can control temps, fans, etc via a command)

thanks

As far as I know, you cant control voltage and clocks,.. I do it throug OverdriveNTool program.

hi good sir, sorry to bother you.

ive been using overdriveNtool, but when i have a power cut problem or something, and the rig restarts, it will restart on the default clocks, not the overdrive ones.
how can i set them auto at win start?
thanks Cheesy

jiang80s

newbie

Activity: 5

Merit: 0

Hi,
I have some issues when using for xhv coin under linux ( RaveOS) comparing with windows.
Under linux my baffin cards like RX 550 2Gb, RX 560 2Gb works only 300 h/s. But it will get 500 h/s more in windows, with same core, mem, and voltage.
Let me know the problem and how to solve it.
Thank you.

bghaemi

newbie

Activity: 2

Merit: 0

Quote from: nik-qz on June 16, 2021, 12:30:45 AM

Quote from: Trietch on June 13, 2021, 06:24:36 AM

Hello, since the v0.8.2 update I keep getting "GPU 2 : detected DEAD" every hour on my stable rig.
When I look at the log, the card is still at normal hashrate when detected DEAD Undecided

Tried to reduce overclock and increase voltage, still having this issue.

Can this be caused by the miner detecting the card DEAD when it's not? Thanks Cheesy

you can check the usb cables, riser and finally reinstall the driver

I got the same issue on GPU0, despite of changed all hardware (Rizer, Cable, Motherboard, Power Supply) also the AMD driver from 18.6.1 to 20.11.2 but still was getting Dead GPU message.
Finally downgraded to v0.8.2 and continue mining ETH without hassle.

oblione

newbie

Activity: 71

Merit: 0

Quote from: GKumaran on June 08, 2021, 07:51:34 AM

Quote from: steve8lily on June 07, 2021, 07:50:48 PM

To the Teamredminer development team:

Is it possible to put in the Power throttle limit control to slow down the mining itself in order to protect GPU does not over temp under GPU temp/ Junction temp/ Vram temp , pls consider

Such a feature isnt there at the moment, but check if these commands help in your case to protect GPUs

Code:

--temp_limit=TEMP Sets the temperature at which the miner will stop GPUs that are too hot.
Default is 85C.
--temp_resume=TEMP Sets the temperature below which the miner will resume GPUs that were previously
stopped due to temperature exceeding limit. Default is 60C.

Are those temps the edge temps? or is there a way to turn a gpu off based on the TMEM temperate until it cools down?

DrX

member

Activity: 233

Merit: 20

Quote from: DrX on June 28, 2021, 05:32:57 AM

Hi,

any other fixes for HW errors than
--eth_dag_alloc_patch

Using Asrock H81 Pro BTC / Bios 2.4. and HiveOS (AMD driver 20.20)

Did try also different PCI-e slots but after about 24h, one of the XFX RX570 starts pushing errors and not one valid shares after that?
Also have tried different voltages and clocks.
only fix seems to be restart miner? is auto-restart timer on miner?

Reply to myself, fixed the problem by changing to lolminer...

demian198

newbie

Activity: 25

Merit: 0

i am having issues after updating amd gpu driver to 21.6.1 with navi14 card "failed to initialize device idx0 (-13)"

DrX

member

Activity: 233

Merit: 20

Hi,

any other fixes for HW errors than
--eth_dag_alloc_patch

Using Asrock H81 Pro BTC / Bios 2.4. and HiveOS (AMD driver 20.20)

Did try also different PCI-e slots but after about 24h, one of the XFX RX570 starts pushing errors and not one valid shares after that?
Also have tried different voltages and clocks.
only fix seems to be restart miner? is auto-restart timer on miner?

dastardlyman12

newbie

Activity: 21

Merit: 0

strange request

I have access to a vega frontier edition water cooled
it has 16gb of hbm2
the card looks great - but thats about it. LOL.

to get it to mine i have to use the amd driver 19.7.1 and do the "is gaming mode driver" registry hack.
then apply some trick timings using amdmemtweakxl ive got it to 45 megahash
with core at 1050 and mem at 1060

but B mode will not enable
the 45 megahash is using A mode or the competition's miner

you cannot use a modern driver with this card because it forces professional creator mode and you lose ALL
control of clocks etc. it defaults to high core clock and low hbm clock exactly what you DONT want for mining

ive not tried linux but i dont think its any better

can i force B mode on amd driver 19.7.1 or is it a lost cause??
many thanks

voodoop

newbie

Activity: 3

Merit: 0

I installed the latest AMD drivers (21.6.1) and I get :
failed to initialize device idx 0 (-13) .

On previous drivers it worked correctly .

LE : one Vega 56 .

UnclWish

sr. member

Activity: 1484

Merit: 253

Quote from: kerney666 on June 23, 2021, 05:42:03 PM

Quote from: UnclWish on June 23, 2021, 06:02:44 AM

Anybody knows why TR miner on Polaris cards (470/480/570/580/590) builds DAG so long?
TR miner build it in about 12-13 sec. on my 580 8Gb cards. 1st of all I'm thought that it's only 1st time after launching (cache, etc.). But every time when DAG changes it take so long as at launch.
But other miners (Claymore, lol, Phoenix) builds DAG about 7-8 seconds.

Well, we just got tired of reports of aggressive (and in some aspects suboptimal) Polaris straps that caused the DAG build read/write pattern to crash gpus, so we defaulted to a very slow and safe version. For continuous ethash mining it's 4-5 secs extra every 4.5 days, 0.0016% penalty, but naturally worse if you switch between epochs more often for some reason, or crash every 2-3h and restart the miner.

I honestly don't remember exactly what type of cap we ended up with, but try adding --eth_dag_slowdown=0 and check the DAG build time.

Oh, sorry. I didn't read usage.txt. Thanks for an option - I will try different variants of it.

EDIT: With -eth_dag_slowdown=0 DAG creation time decreased to 8.8 sec. It's faster but still slower than other miners (7.8-8.0 sec.). I'm use nicehash - his pool often changes DAG's, so for me DAG rebuilding time is important...

kerney666

member

Activity: 658

Merit: 86

Quote from: UnclWish on June 23, 2021, 06:02:44 AM

Anybody knows why TR miner on Polaris cards (470/480/570/580/590) builds DAG so long?
TR miner build it in about 12-13 sec. on my 580 8Gb cards. 1st of all I'm thought that it's only 1st time after launching (cache, etc.). But every time when DAG changes it take so long as at launch.
But other miners (Claymore, lol, Phoenix) builds DAG about 7-8 seconds.

Well, we just got tired of reports of aggressive (and in some aspects suboptimal) Polaris straps that caused the DAG build read/write pattern to crash gpus, so we defaulted to a very slow and safe version. For continuous ethash mining it's 4-5 secs extra every 4.5 days, 0.0016% penalty, but naturally worse if you switch between epochs more often for some reason, or crash every 2-3h and restart the miner.

I honestly don't remember exactly what type of cap we ended up with, but try adding --eth_dag_slowdown=0 and check the DAG build time.

kerney666

member

Activity: 658

Merit: 86

Quote from: dastardlyman12 on June 16, 2021, 05:45:58 PM

I manage about 30 amd rigs in the birmingham area of the uk.

really like your miner - it does all you claim and is slightly better than the competition.
its especially good on radeon VII and rx5700 .

at present the rigs all have the following set in the config.txt file
core clock
mem clock
core volts
mem volts
fixed fan speed
max temp before gpu throttles (this is vital)

theres other stuff but those 6 things are the main ones.

if your miner could set those 6 things within its config - i would dump the competition and use your miner on all rigs.
yes i know about overdriventool and msi afterburner but in my experience on rigs with 470, 570,480,580,5500, 5700, radeon vii -
those tools are hit and miss and too much trubble .
we use windows 10 ghost spectre (no updates, bloat or telemetry) and mostly amd driver 20.9.1
the one rig that isnt 20.9.1 - has a vega frontier edition 16gb water cooled - i should do a video on
that rig - what a palaver to get that card mining. but we did it with driver 19.7.1 and the registry hack

ghost spectre rocks - you will be amazed. no updates . rock solid . tiny ram usage . tiny disk usage.
best wishes from manchester uk

Clocks/voltage support is being worked on, most probably windows only in the first release.

Fan support already exists, see --fan_control in the USAGE.txt docs (incl tracking core and/or mem temp, using a static speed, defined min/max range and more), but we've also fixed some Navi fan control issues for the next release.

For temps, we don't throttle per se to keep the gpu running close to the limit, it's assumed the user should make sure to keep temps within the expected range and the protection is rather against broken fans or similar incidents. We do a hard cutoff at 85C (default) and stop mining until the gpu reaches 65C again. See --temp_limit and --temp_resume for more info.

Last, I'm sure you're well aware that VIIs do excel under linux with our C-mode becoming available after a few kernel parameter tweaks, just mentioning it just in case. I'm guessing running is homogeneous environment is more important for you guys.

rambetiko

member

Activity: 188

Merit: 10

what is all CN algo hash for 6900xt? haven heavy v8 lite etc.

GKumaran

member

Activity: 204

Merit: 10

Quote from: dastardlyman12 on June 16, 2021, 05:45:58 PM

I manage about 30 amd rigs in the birmingham area of the uk.

really like your miner - it does all you claim and is slightly better than the competition.
its especially good on radeon VII and rx5700 .

at present the rigs all have the following set in the config.txt file
core clock
mem clock
core volts
mem volts
fixed fan speed
max temp before gpu throttles (this is vital)

theres other stuff but those 6 things are the main ones.

if your miner could set those 6 things within its config - i would dump the competition and use your miner on all rigs.
yes i know about overdriventool and msi afterburner but in my experience on rigs with 470, 570,480,580,5500, 5700, radeon vii -
those tools are hit and miss and too much trubble .
we use windows 10 ghost spectre (no updates, bloat or telemetry) and mostly amd driver 20.9.1
the one rig that isnt 20.9.1 - has a vega frontier edition 16gb water cooled - i should do a video on
that rig - what a palaver to get that card mining. but we did it with driver 19.7.1 and the registry hack

ghost spectre rocks - you will be amazed. no updates . rock solid . tiny ram usage . tiny disk usage.
best wishes from manchester uk

Dev said they were working on, but no idea on eta !!

UnclWish

sr. member

Activity: 1484

Merit: 253

Anybody knows why TR miner on Polaris cards (470/480/570/580/590) builds DAG so long?
TR miner build it in about 12-13 sec. on my 580 8Gb cards. 1st of all I'm thought that it's only 1st time after launching (cache, etc.). But every time when DAG changes it take so long as at launch.
But other miners (Claymore, lol, Phoenix) builds DAG about 7-8 seconds.

dastardlyman12

newbie

Activity: 21

Merit: 0

I manage about 30 amd rigs in the birmingham area of the uk.

really like your miner - it does all you claim and is slightly better than the competition.
its especially good on radeon VII and rx5700 .

at present the rigs all have the following set in the config.txt file
core clock
mem clock
core volts
mem volts
fixed fan speed
max temp before gpu throttles (this is vital)

theres other stuff but those 6 things are the main ones.

if your miner could set those 6 things within its config - i would dump the competition and use your miner on all rigs.
yes i know about overdriventool and msi afterburner but in my experience on rigs with 470, 570,480,580,5500, 5700, radeon vii -
those tools are hit and miss and too much trubble .
we use windows 10 ghost spectre (no updates, bloat or telemetry) and mostly amd driver 20.9.1
the one rig that isnt 20.9.1 - has a vega frontier edition 16gb water cooled - i should do a video on
that rig - what a palaver to get that card mining. but we did it with driver 19.7.1 and the registry hack

ghost spectre rocks - you will be amazed. no updates . rock solid . tiny ram usage . tiny disk usage.
best wishes from manchester uk

nik-qz

member

Activity: 66

Merit: 45

Quote from: Trietch on June 13, 2021, 06:24:36 AM

Hello, since the v0.8.2 update I keep getting "GPU 2 : detected DEAD" every hour on my stable rig.
When I look at the log, the card is still at normal hashrate when detected DEAD Undecided

Tried to reduce overclock and increase voltage, still having this issue.

Can this be caused by the miner detecting the card DEAD when it's not? Thanks Cheesy

you can check the usb cables, riser and finally reinstall the driver

Trietch

newbie

Activity: 6

Merit: 0

Hello, since the v0.8.2 update I keep getting "GPU 2 : detected DEAD" every hour on my stable rig.
When I look at the log, the card is still at normal hashrate when detected DEAD Undecided

Tried to reduce overclock and increase voltage, still having this issue.

Can this be caused by the miner detecting the card DEAD when it's not? Thanks Cheesy

GKumaran

member

Activity: 204

Merit: 10

Quote from: steve8lily on June 07, 2021, 07:50:48 PM

To the Teamredminer development team:

Is it possible to put in the Power throttle limit control to slow down the mining itself in order to protect GPU does not over temp under GPU temp/ Junction temp/ Vram temp , pls consider

Such a feature isnt there at the moment, but check if these commands help in your case to protect GPUs

Code:

--temp_limit=TEMP Sets the temperature at which the miner will stop GPUs that are too hot.
Default is 85C.
--temp_resume=TEMP Sets the temperature below which the miner will resume GPUs that were previously
stopped due to temperature exceeding limit. Default is 60C.

GKumaran

member

Activity: 204

Merit: 10

Quote from: nik-qz on June 07, 2021, 11:52:30 AM

On the Autolykos protocol, when mining an ERGO coin (pool woolypooly), the miner constantly writes about duplicate work, is this not good?
a lot of expired shares and as a result of rejections

Quote from: Kerney666(In TRM Discord)

To be clear, it's not a bug in the miner, it's just an unnecessary warning, the workaround is 100% safe and works. Basically, it means we don't refresh the nonce range we test since the pool job we just received is 100% identical to the previous one, we continue hashing from the last nonce used before the new job arrived. It's more that it's occuring much more often in the ERGO mining world, pools don't seem to refresh the job from the daemon when they send out a "new" job for the same network block height. The warning was inherited from other mining protocols where this doesn't really happen naturally but should be removed for ERGO for users since it adds zero value, only confusion.

Topic: [ANN] TeamRedMiner v0.10.10 - Ironfish/Kaspa/ZIL/Kawpow/Etchash and More - page 10. (Read 211380 times)