Quote from: Tay on March 24, 2014, 09:33:53 PM

Quote from: goatonastick on March 24, 2014, 08:04:01 PM

HELP!

I've tried using all my dogecoin settings, and even pasting some of the settings here, but have not managed to break more than 11 KH/s with vertminer Sad

My tried-and-true settings for dogecoin on my 7850 usually give me 346 KH/s:
-intensity 12 --gpu-threads 2 --gpu-engine 1000 --gpu-memclock 1200 --shaders 1024 --thread-concurrency 4096 --worksize 256

but even with the thread concurrency in half, I still can't get much higher than 12 KH/s.

what am I doing wrong? gpu has 2GB of memory, and system has 4GB of ram.

After much tweaking I've got my 7850's up to 175+kh/s on vertminer (BAMT)

   "worksize": "64",
   "intensity": "19",
   "thread-concurrency" : "8192",
   "shaders" : "1280",
   "gpu-engine" : "980",
   "gpu-memclock" : "1250",
   "gpu-vddc" : "0",
   "gpu-powertune" : "0",
   "gpu-threads" : "1",
   "lookup-gap" : "2",
   "queue": "0",

MSI HD7850 (single fan) 2GB 7850's, these are currently performing better than my Asus 270's.

Not used to using the vertminer config file, I simply rename vertminer.example.conf to vertminer.conf right? After I edit the URL and Login/Pass of course.

I get these errors:

Fatal JSON error in configuration file
Configuration file could not be used

here's my config file:

Code:

{
"pools" : [
{
"url" : "stratum+tcp://spa.pool.mn:3370",
"user" : "goatonastik.01",
"pass" : "01"
},
{
"url" : "stratum+tcp://spa.pool.mn:3370",
"user" : "goatonastik.01",
"pass" : "01"
}
]
,
"worksize": "64",
"intensity": "19",
"thread-concurrency" : "8192",
"shaders" : "1280",
"gpu-engine" : "980",
"gpu-memclock" : "1250",
"gpu-vddc" : "0",
"gpu-powertune" : "0",
"gpu-threads" : "1",
"lookup-gap" : "2",
"queue": "0",
}

Sorry for my noobiness but I'm used to using .bat files.

EDIT: good news / bad news. I put the commands into the bat file, and I get 177 KH/s, but I have around 20% hardware errors

EDIT 2: now they're averaging down to less than 10%. I know any HW errors are bad but this is so much better than before, thanks!