Author

Topic: Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0 (Windows/Linux) - page 639. (Read 6590565 times)

newbie
Activity: 124
Merit: 0
Hi guys,

below is my set-up and I was wondering if I could use 1200w under 7GPU, or should I add another PSU to support the wattage.
Appreciate your guide on this  Grin

CORSAIR – 1200w
MSI M5 7 PCIE
4x - RX 570
3x – RX 580
MSI AB 4.4.0 beta 16


legendary
Activity: 2212
Merit: 1038
I can't imagine any farm of size messing with all of these little tools on every startup, so they must burn the bios with volts and clocks.

PBE can't do it, so ... what does?

Like they use Windows - that would be stupid. They use Linux, usually, and either VBIOS mods or tools like mine on *nix to access the VRM directly and control it.

I fried all the pcie slots on my windows box, may I ask what *nix tools you use? Specifically read and writing bios, a bios settings editor and monitoring GPU stats.

On windows I use ATIWinfLash, PolarisBiosEditor and GPU-Z.


atiflash - command line one. I'd have to go find the public one if you need a copy, but I could. OhGodATool (written by my owner & I) for VBIOS edits, as well as a couple private tools I wrote (amdtimingcopy, the wolfmctools suite for memory timing fine-tuning, and shit.) The only other tool I use for live monitoring (and the only one I know exists that isn't super under wraps) is my own; wolfamdvolt/ohgodavolt, which I've shown on BCT before (NSFW): http://lovehindpa.ws/screens/wolfamdvolt.jpg

Well if your owner allows it git would be a great place to make any tools you feel like sharing available.
newbie
Activity: 24
Merit: 0
The last good drive is still 17.7.1. No point installing new ones, they only break stuff and do nothing good, unless blackchain driver if you know how to overcome the issues.

Thanks for the info, the latest one i'm using is 17.7.2, if after switching to 17.7.1 it still fluctuate even though it works well on you, i can conclude that the problem lies somewhere else
full member
Activity: 158
Merit: 100
For RX 5xx users, May I ask which driver versions are you using?
Since AMD blockchain driver doesn't let me set mclk on one of my cards, i reverted back to using normal drivers, but now the current up to date drivers fluctuate both the core and mclk and drop the hashrate of both cards

The last good drive is still 17.7.1. No point installing new ones, they only break stuff and do nothing good, unless blackchain driver if you know how to overcome the issues.
newbie
Activity: 24
Merit: 0
For RX 5xx users, May I ask which driver versions are you using?
Since AMD blockchain driver doesn't let me set mclk on one of my cards, i reverted back to using normal drivers, but now the current up to date drivers fluctuate both the core and mclk and drop the hashrate of both cards
sr. member
Activity: 430
Merit: 254
Wolf0 I have a question. Does it help (in the context of hash rates) to have a certain Linux kernel (as in a recent one...perhaps that you compile yourself) or a certain driver (amdgpu vs amdgpu-pro)? And I doubt it makes a difference...but what about mesa version?

btw I'm asking within the confines of publicly available packages specifically.
sr. member
Activity: 430
Merit: 254
Howdy Devs and happy Miners!!

Question:  When the Claymore dual miner restarts (for whatever reason: coin switch etc.) it resets the core voltage value on my MSI R9 290x Gaming 4G to zero.  This is a huge problem because it hits 80C on this setting.  I have to manually go back into Afterburner and reset it to -100 everytime.  This gets the card back down to 70C without affecting my hashrate, so how do I access and change this particular setting within Claymore?  FWIW I also have an R9 390 in this rig but when Claymore resets the core voltage value this card is not negatively affected in the temperature department.

My Rig:
Running Nicehash Legacy
Crimson Blockchain drivers
MSI Afterburner 4.4.0 beta 12

Motherboard: 990fx-ud3
GTX 1070
Rx 580
R9 290x  <---problem card
R9 390

Thanks in advance!!
Chris

Edit the BIOS to have the desired voltage/clocks at load, then you don't need to use AB (or wattman or whatever) at all.
legendary
Activity: 2212
Merit: 1038
I can't imagine any farm of size messing with all of these little tools on every startup, so they must burn the bios with volts and clocks.

PBE can't do it, so ... what does?

Like they use Windows - that would be stupid. They use Linux, usually, and either VBIOS mods or tools like mine on *nix to access the VRM directly and control it.

I fried all the pcie slots on my windows box, may I ask what *nix tools you use? Specifically read and writing bios, a bios settings editor and monitoring GPU stats.

On windows I use ATIWinfLash, PolarisBiosEditor and GPU-Z.
full member
Activity: 238
Merit: 100
Hi guys i need a pro Tip )))

Im using two asus strix 1060 6 gb
mining ETC on nanopool with Claymore 9.8
sometimes its found  share very low

is it im doing something wrong ?
btw 46 mhs

TY


Yes - mining the wrong coin.
full member
Activity: 238
Merit: 100
Hello everyone. I have error on claymore, ETH: Stratum - Failed to connect. Network is normally, i tried to reset router, changed dns but nothing work. Plz help, first time i saw that issue and dont know how to fix it


What happens when you run default start.bat files which come with original Claymore's miner package?
full member
Activity: 238
Merit: 100
1 year back, one of my very new rigs was taking 450 Watts on the wall.

Now I've just checked and it's consuming 600 Watts for the very same hash rate.

Does the efficiency go down this much of the hardware? Roll Eyes

What about the temperature on the cards and ambient? Has that increased?

Power consumption is side effect of increased performance. Claymore has optimised the miner significantly since 1 year ago. Thus increased power consumption.

A big yes! Temperature has been extremely high too along with the increase of power. But hash-rate hasn't changed/increased.

Temperature now with 100% fan is around 90, never happened before. Roll Eyes

Risen temperature is where the extra power goes to - heating.
I would have checked the thermal paste and dust off and grease the fans.
Once you cool down the rig the wattage should come back to normal unless you decide to upgrade the miner.
Latest version more efficient but power hungrier.
newbie
Activity: 34
Merit: 0
I can't imagine any farm of size messing with all of these little tools on every startup, so they must burn the bios with volts and clocks.

PBE can't do it, so ... what does?
sr. member
Activity: 449
Merit: 251
All other software , including wattman seams to work best when the P1 to P6 are disable for GPU and P1 / P2 for Memory, using 1 profile at min/max .  Allowing GPU to run only 1 state like that it has no oscillation, would that be a solution ? Of course in idle you have a high power consummation, but no miner should stay in idle with the GPU's, or default settings of GPU could be reset before quit script.
Thanks,

Is any third-part software able to manage voltage with blockchain drivers?
MSI Afterburner 4.4 Beta 15 can manage voltage on 570/580s, but can only do a max of -96mv offset. CDM 9.8 seems to work fine for 400 series. I don't know about Vega, haven't found one for sale at reasonable price yet, but from what I have seen drivers are still pretty buggy with clocks and volts.
full member
Activity: 186
Merit: 100
So the blockchain drivers will not work on Win7 x64?

I will have to wait until they adapt a patch?

Really dont want to upgrade to Win 10. Too much at stake on my PC.
KGV
newbie
Activity: 55
Merit: 0
All other software , including wattman seams to work best when the P1 to P6 are disable for GPU and P1 / P2 for Memory, using 1 profile at min/max .  Allowing GPU to run only 1 state like that it has no oscillation, would that be a solution ? Of course in idle you have a high power consummation, but no miner should stay in idle with the GPU's, or default settings of GPU could be reset before quit script.
Thanks,

Is any third-part software able to manage voltage with blockchain drivers?

OverdriveNTool
hero member
Activity: 1036
Merit: 606
All other software , including wattman seams to work best when the P1 to P6 are disable for GPU and P1 / P2 for Memory, using 1 profile at min/max .  Allowing GPU to run only 1 state like that it has no oscillation, would that be a solution ? Of course in idle you have a high power consummation, but no miner should stay in idle with the GPU's, or default settings of GPU could be reset before quit script.
Thanks,

Is any third-part software able to manage voltage with blockchain drivers?

Undervolting and Power Limit adjustment works for me with Afterburner Beta 16 in Windows 10. RX 570/580 and 480 cards.
donator
Activity: 1610
Merit: 1325
Miners developer
All other software , including wattman seams to work best when the P1 to P6 are disable for GPU and P1 / P2 for Memory, using 1 profile at min/max .  Allowing GPU to run only 1 state like that it has no oscillation, would that be a solution ? Of course in idle you have a high power consummation, but no miner should stay in idle with the GPU's, or default settings of GPU could be reset before quit script.
Thanks,

Is any third-part software able to manage voltage with blockchain drivers?
donator
Activity: 1610
Merit: 1325
Miners developer
Hi Claymore : is there a new version of your dual miner in the pipe, ready soon ?
The actual one is very good, but we ever hope better ;-).

I plan to release new version in 3-4 days. I've almost finished ASM mode for ETH+LBC, also try to find some workaround for voltage control.
newbie
Activity: 42
Merit: 0
1 year back, one of my very new rigs was taking 450 Watts on the wall.

Now I've just checked and it's consuming 600 Watts for the very same hash rate.

Does the efficiency go down this much of the hardware? Roll Eyes
your rig is possibly becoming wiser with time and does something productive for itself on your expense

Dust or PSU, you need to blow it out w/ air

It should not happen: the worst PSU I had, went from 88% efficiency down to 78% in a year and bit.

You are talking about a huge drop: something like a PSU at 90% efficiency going down to 60% efficiency.


I have a few Antminers I picked up used.  When I first got them, they were caked in dust.  After a good cleaning, ASIC temps dropped, and I picked up almost 100 GH on one of them.

I'm also trying to dual mine on Decred and the Yimp pool is just generating too many errors and time outs.  Do you guys know how to make the suprnova pool work?  claymore won't take a -u and -p setting
legendary
Activity: 1862
Merit: 1505
1 year back, one of my very new rigs was taking 450 Watts on the wall.

Now I've just checked and it's consuming 600 Watts for the very same hash rate.

Does the efficiency go down this much of the hardware? Roll Eyes
your rig is possibly becoming wiser with time and does something productive for itself on your expense

Dust or PSU, you need to blow it out w/ air

It should not happen: the worst PSU I had, went from 88% efficiency down to 78% in a year and bit.

You are talking about a huge drop: something like a PSU at 90% efficiency going down to 60% efficiency.

Yes, that's why I'm curious. It'd have been fine if it was 50 Watts increase in power, but jumping from 450 Watts to 600 Watts is really ridiculous, specially for a new bronze Corsair power supply.

I'm going to open up my PSU, GPU's, whole system and suck some dust through the stuff through a light vacuum cleaner. Let's see what happens.
Jump to: