Quote from: akuci on February 01, 2018, 10:23:25 AM

Any chance for some nVidia performance upgrade with the new Cuda 9.1?

We've done some preliminary testing but the results are not encouraging - the difference it is very small and not always in favor of CUDA 9.1. Combined with the nuisance of having to install Nvidia drivers that support CUDA 9.1 we don't see much reason to support it until the drivers are older and more stable.

Quote from: kurupilav on February 01, 2018, 10:50:09 AM

I have tried your miner in my 2 GPU rig, miner starts and after 20-25 seconds later simply exits without giving any warning or error

here is the log :
...
any help?

Can you please open a command line window and run PhoenixMiner from there? Or, if it is easier for you, run the start_miner.bat file - it includes the pause command after the execution of PhoenixMiner, so the console won't close automatically and you will be able to see if there is any error message that is not included in the log file.

Quote from: Tsum on February 01, 2018, 06:29:19 AM

RX 580 and the driver version is 18.1.1

Quote from: benosimo06 on February 01, 2018, 01:15:41 PM

I've got an RX 380 + 480 using the 17.30.1029 Blockchain driver. The firewall change didn't have any effect for me either.

We'll prepare a special version that will "spit out" more useful information about the error to give us some directions where to look for its causes. It will be PM'd to anyone who is suffering from the "Fatal error: debugger detected" problem.

Quote from: PercyPea on February 01, 2018, 02:28:14 PM

Just wanted to say, currently mining ether classic at 24ish mhs with version 2.4 on a 3GB 1060

verion 2.6b doesnt work as it has a cuda crashing saying not enough memory on the 1060 Sad

haven't tried 2.5d as windows defender reports some norty virus' (2.4 and 2.6b are virus free)

You can safely ignore the windows defender warning (but keep in mind that it can be quite aggressive and even delete PhoenixMiner.exe without asking). Many AV programs seem to freak-out when see code that thwarts any attempts to debug or monitor the executable and declare it a dangerous. As for the memory error, this is kind of strange as we have 1060 3GB in our test rigs and they work fine with 2.6b when mining ETC. You can try to specify the -eres 0 command line option which will turn off the DAG pre-allocation and force 2.6b to behave exactly like 2.4, which didn't have DAG pre-allocation.