Topic: Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0 (Windows/Linux) - page 83. (Read 6590565 times)

Dude run as ADMINISTRATOR

Did not work for me:(
I guess no more ETC mining:(

Sounds like compute mode is not on.

So, my miner's motherboard died and SSD crashed. Got new motherboard and SSD today.
I had no backup of my previous system, so I reinstalled everything from scratch using latest versions.

Gigabyte GA-H110-D3A, 6x RX580 4GB

Before:
Win 10, I think build 1703 of 1709 (not sure)
AMD beta blockchain drivers
Claymore 11.9

Now:
Win 10 build 1903 (latest)
Latest AMD drivers
Claymore 15

Before I got  +- 182MH/s ETH only (as expected for this rig), but now something like 72 (+-12 per card, all 6 cards working). Something's obviously off.

Quick suggestions on what I did wrong? Which OS version and driver version should I use? Thanks!

Thank you! I will try that driver for my 1050ti. I hope this gets fixed soon with an update or something, I do like keeping my drivers up to date

Hi everyone!

Yesterday i got the same issue with my 1050TI 4gb, not enough memory for dag generation. Checked everything, and it turned out that NV drivers 382.53 works fine on some of my rigs with this cards, so i downgraded drivers and now it works fine. Hope it helps ).

Claymore i think its time you for new version,
no one answered my question here.
V15.0 has some bugs like gtx1060 crashing after a while and many share rejects.
Waiting for new update......

Hi,

Need to help here to make sure Claymore isn't killing my GPUs.

I'm getting this error message:

NVIDIA NVML library not found, temperature monitoring for NVIDIA GPUs disabled.
Driver is not installed! Trying to install it...
Driver is in use, try to start miner with "-driver uninstall" parameter.

Basically I want to be able to control the temp and make sure I'm not running my rig at full blast, so can anyone help with setting up the NVML.dll so Claymore will recognize it. Apparently NVIDIA doesn't install with the driver installation from what I've read.

Its on

try rxboost option, get you extra 1 MH

Hi guys,

Im an monero guy, and today i tested this miner. Because with ramdomx monero mining will be not profitable, im testing ETH.  

I got one rig with vega 64, i can reach 51+ Mh/S on ethereum with strap 5 with no problems.  In monero it consume 195x watts for 8 gpus. With ETH the consume is 22xx watts. What an increase hehe

In other rig I got 8 mixed rx 580 with bios mod and one vega 56.  They handle 29-31,5 Mh/S. 29- 4gb elpida and 31,5 on 8gb samsung. When i use straps in this moded gpus i got no increase, sometimes an decrease. Maybe is normal, because they are biosmoded. ?

I got one vega 56 and got  43x Mh/S with strap 2-4.

I´m using overdriventool xmr settings, and doned some test that show i can lower a lot my gpu core with no impact in hashing performance. Will test this better when i merge to ETH. But i can lower core to something like 1000 mhz with no decrease in performance:

https://ibb.co/zrwTzXp

Strange or maybe the algo is not gpu core sensitive like xmr.

Anyone had passed for this too?  

Thnks  

I'm having the same issue today out of nowhere with all 5 of my 1050ti. Cannot write DAG. I've tried many things:
Updated driver, no go
Downgraded driver, no go
-eres 0, no go
increase page file size to 40gb, no go

please, we need help, our 1050ti's are dying lol

GPU-z says 77MB used

Check with gpu-z how match gpu memory is used from windows before starting the miner.

I already have these set:

setx GPU_FORCE_64BIT_PTR 0
setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_SINGLE_ALLOC_PERCENT 100

Has any one go this to work with a 4GB GTX 1050 ti recently

I have 8 x GTX 1050 Ti 4GB on Windows 10.
This morning they stopped with DAG creation error.
Any idea how to resolve this, the DAG is only 3.3GB I believe
My Virtual Memory is more than enough to cover all of my cards in the rig.

Need Help for strap
My rig has 3 GTX 1070 and 3 DTX 1060.

I am using 1@4600 for GTX 1070 to get an extra 7% of hash rate.

like

-strap 1@4600, 1@4600,1@4600,1,1,1

but sometime the 1070  would have an unknown error, causing the GPU out of line and miner would restart its own. Also one of 1060 is going to run at 4600 and causing 1060 has error and the miner keeping crash.

Someone has a solution for this. If the strap option can apply to PCI bus could solve this issue.

Thanks a lot

Larry

Hi,

Is this miner still maintained? Where can i see the changelog? 

Sorry, i didn' want to be rude.

Regardless of the opinion, THIS thread is about Claymore and HIS development. Not PM. If you were to open a thread sharing work that you've dedicated several hours to (and profiting from for those hours) just for someone to post reason(s) not to support him and to support someone else's work, would that be appropriate? On top of that express an inappropriate, rude and sexist attitude toward another member? Sure, I posted the original statement about PM software being updated. Obviously, that wasn't the best idea since now it put this douchwad on my ignore list (because he is clearly out of line) but I shared that information to spur development on Claymore and to INFORM other members of the advancement. At no point was my intention to ever persuade or coerce anyone to switch to PhoenixMiner. He on the other hand appears to be doing exactly that... in Claymore's thread. Which is definitely inappropriate. If he is such a PM supporter then what the hell is he doing posting in this thread? The sexist comments that he was posting was not only off topic but also libel. He doesn't know shit about me or anyone else on here, no less to even assume what a "normal female" is. His comments were definitely inappropriate, sexist, libel, rude, inappropriate and definitely off topic. Such are nothing but counterproductive.