PhoenixMiner 6.2c: fastest Ethereum/Ethash miner with lowest devfee (Win/Linux)

[xsoft]

Any tip for older 4GB cards (like R7 370) which throws error when generating DAG:

Code:

GPU1: Free VRAM: 3.885 GB; used: 0.001 GB
GPU1: Disabling DAG pre-allocation (not enough VRAM)
GPU1: Allocating DAG for epoch #359 (3.80) GB
GPU1: Allocating buffers failed with: clCreateBuffer (-61).
Fatal error detected. Restarting.

Ive tried -gser  -gser 2, -clNew 0, altInit...
Tested on 5.1c and 5.0e. Same results with claymore, linux EthOS.

./PhoenixMiner  -proto 3 -coin eth -amd -altinit -gser 2 -worker xxx -log 0 -wdog 0 -pool stratum+tcp://pool:4444 -wal 0x000

[rasmukri]

What does the percent thing mean when it rolls off the temps and power? what does 1872% mean? I'm running 1 GPU a RX 5700XT OC 2100 on the core and 1888 memory

It says

Code:

2020.08.28:20:29:41.955: main GPU1: 41C 1872% 221W

Anyone

[motomart]

Code:

GPU1: Unable to set GPU clocks to 1775 MHz (Vddc 975 mv) - error -8

Can anybody give some insight on this error?

Unsupported drivers?

I don't think so, other 5 cards in the same rig are hashing ok....

Maybe you write more about your config? Cards, system, drivers, command line? More strings from log? Before asking...

I have 6x radeon vii and today I have this error:

GPU1: Unable to set GPU clocks to 1775 MHz (Vddc 975 mv) - error -8

Temps seem ok, cclock and cvddc are low, power usage low and card is hashing at only 36 MH/s. I rebooted the system several times, checked settings but can't figure out why it happened all of a sudden.


2020.08.24:15:46:58.314: main GPU1: 58C 23% 78W, GPU2: 60C 71% 195W, GPU3: 61C 63% 233W, GPU4: 67C 76% 215W, GPU5: 64C 75% 194W, GPU6: 63C 79% 200W
GPU1: cclock 700 MHz, cvddc 737 mV, mclock 1120 MHz, mvddc 850 mV, Tj 65C, Tmem 66C
GPU2: cclock 1765 MHz, cvddc 975 mV, mclock 1125 MHz, mvddc 850 mV, Tj 77C, Tmem 71C
GPU3: cclock 1751 MHz, cvddc 975 mV, mclock 1125 MHz, mvddc 850 mV, Tj 85C, Tmem 74C
GPU4: cclock 1744 MHz, cvddc 968 mV, mclock 1124 MHz, mvddc 850 mV, Tj 96C, Tmem 76C
GPU5: cclock 1748 MHz, cvddc 975 mV, mclock 1120 MHz, mvddc 850 mV, Tj 87C, Tmem 74C
GPU6: cclock 1750 MHz, cvddc 975 mV, mclock 1120 MHz, mvddc 850 mV, Tj 99C, Tmem 71C


Here is my startup parameter
-cclock 1775,1775,1775,1775,1775,1775 -mclock 1120,1125,1125,1125,1120,1120 -cvddc 975,970,970,965,970,975 -powlim 0,0,0,0,0,0

[motomart]

What does the percent thing mean when it rolls off the temps and power? what does 1872% mean? I'm running 1 GPU a RX 5700XT OC 2100 on the core and 1888 memory
https://ibb.co/jZp9q28
It says

Code:

2020.08.28:20:29:41.955: main GPU1: 41C 1872% 221W

Anyone

That is % of GPU fan speed. Sometimes card give the wrong info to miner. Disregard this, it might get back to normal as it runs, it happened to me before.

[melpheos]

Anyone is running windows 7 with 3gb of ram and nvidia 10xx (with more than 3gb) series and have good perf on eth ?
I know after checking that some drivers manage memory badly and you need more than 3Gb of ram on the motherboard for some reason to match at least the DAG file size.
If you have good perf, can you tell me the driver version you are using ?
(no this is not an issue with virtual memory)

[joseph32]

Anyone is running windows 7 with 3gb of ram and nvidia 10xx (with more than 3gb) series and have good perf on eth ?
I know after checking that some drivers manage memory badly and you need more than 3Gb of ram on the motherboard for some reason to match at least the DAG file size.
If you have good perf, can you tell me the driver version you are using ?
(no this is not an issue with virtual memory)

You need 4 GB Ram on the card(s), not the board!

[Iamtutut]

Would anyone share settings for Gigabyte RX64 gaming ? Thanks

[xeridea]

Got my 3090 delivered today. These hashrates are crazy!!! 120MH for Ethereum, 65MH for KAWPOW. Eats up the power though, need to tweak the power usage.

[melpheos]

Anyone is running windows 7 with 3gb of ram and nvidia 10xx (with more than 3gb) series and have good perf on eth ?
I know after checking that some drivers manage memory badly and you need more than 3Gb of ram on the motherboard for some reason to match at least the DAG file size.
If you have good perf, can you tell me the driver version you are using ?
(no this is not an issue with virtual memory)

You need 4 GB Ram on the card(s), not the board!

Why do you think I kept writing that this is a driver issue !
Check some previous post some months ago and you will see that for some unknown reasons and in specific configuration, you also need more physical memory on the motherboard as well.
I perfectly know how much memory you need on the graphic card right now to mine eth and I can guarantee you that if I remove a 2gb memory on my rig out of 6gb, my hashspeed will be reduced by more than half.
PhoenixMiner editor thinks this is a driver bug but I was unable to find a driver that doesn't have this issue and this issue occurs on my 3 rigs.
My GPU are 8Gb and they are working fine as long as i have more than 4gb of ram on the motherboard.

[melpheos]

Got my 3090 delivered today. These hashrates are crazy!!! 120MH for Ethereum, 65MH for KAWPOW. Eats up the power though, need to tweak the power usage.

ship on the 24th of september... I call BS

[xeridea]

Some "proofs"?

I'm lie. I would rather have 15 570s than a single 3090 any day!!! Mining today is crazy!!!

[rasmukri]

That is % of GPU fan speed. Sometimes card give the wrong info to miner. Disregard this, it might get back to normal as it runs, it happened to me before.

Thanks, I have it water-cooled it mines when I'm not gaming so the fan header is empty. Guess that's how it reports the lack of anything connected.

[GODLIKEE]

Would anyone share settings for Gigabyte RX64 gaming ? Thanks

the best you can get on that gpu is 43mh/s. if you raise the tRef over 4k it crash always

[jebidia]

Anyone have luck with the -straps function?

-straps <n> Memory strap level (Nvidia cards 10x0 series only). The possible values are 0 to 6. 0 is the default value and uses the default
     timings from the VBIOS. Each strap level corresponds to a predefined combination of memory timings (-vmt1, -vmt2,
     -vmt3, -vmr). Strap level 3 is the fastest predefined level and may not work on most cards, 1 is the slowest (but still faster than
     the default timings). Strap levels 4 to 6 are the same as 1 to 3 but with less aggressive refresh rates (i.e. lower "-vmr" values).

I'm getting an error 31 Must be run as administrator...Tried to do so but still getting the error...Wanted to see if this feature could be used instead of the ethlrgmentpl

[temporalminer]

Anyone have luck with the -straps function?

-straps <n> Memory strap level (Nvidia cards 10x0 series only). The possible values are 0 to 6. 0 is the default value and uses the default
     timings from the VBIOS. Each strap level corresponds to a predefined combination of memory timings (-vmt1, -vmt2,
     -vmt3, -vmr). Strap level 3 is the fastest predefined level and may not work on most cards, 1 is the slowest (but still faster than
     the default timings). Strap levels 4 to 6 are the same as 1 to 3 but with less aggressive refresh rates (i.e. lower "-vmr" values).

I'm getting an error 31 Must be run as administrator...Tried to do so but still getting the error...Wanted to see if this feature could be used instead of the ethlrgmentpl

I'm having the same problem. I would like some assistance please.

[rasmukri]

Is there any gain by over locking in the config file instead of using something like afterburner?
I use afterburner and it works pretty damn good. Just wondering if there is some sort of performance gain to do it in the config for the miner instead.

[xeridea]

Would anyone share settings for Gigabyte RX64 gaming ? Thanks

the best you can get on that gpu is 43mh/s. if you raise the tRef over 4k it crash always

Use claymore to set straps, then mine with TeamRedMiner, I get upwards of 50MH on reference Vega 56....

[xeridea]

Is there any gain by over locking in the config file instead of using something like afterburner?
I use afterburner and it works pretty damn good. Just wondering if there is some sort of performance gain to do it in the config for the miner instead.

Performance don't think so. I think clocks may be more set in stone on some cards with config though.  Also if you switch coins easier to set different for each one. Also easier to change en masse with config, or through remote manager. If tuning each card, using config make it easier to adjust for instability since numbers in log same as order you set them... Down to preference really.

[greasepit]

Is there any gain by over locking in the config file instead of using something like afterburner?
I use afterburner and it works pretty damn good. Just wondering if there is some sort of performance gain to do it in the config for the miner instead.

There are some miners with no clock settings built in, so you will need a third party software.
IMO Afterburner is for amauters or testing stability before going live because in case of power loss your cards can reset the clocks to default, setting on the miner cfg or either on the BIOS your cards will never miss clocks settings. You can set a script as well.

[john1010]

Is there any gain by over locking in the config file instead of using something like afterburner?
I use afterburner and it works pretty damn good. Just wondering if there is some sort of performance gain to do it in the config for the miner instead.

Using Afterburner will frustrate you, better to make your own experiment in doing some trial and error in the command line of Phoenix miner or you can modify your card and set a proper strap for every card yo have, but moding is too risky it may damage and brick your card.