Topic: The big VTC Vertcoin settings thread - page 43. (Read 264260 times)

welp nevermind.  restarted my miner and used vertminer with:

vertminer -o stratum+tcp://stratum.vertco.in:8080 -u x.x -p x --scrypt-vert --intensity 13 --worksize 256 --gpu-threads 2 --lookup-gap 2 --thread-concurrency 8192 --temp-target 75 --gpu-fan 35-95 --auto-fan --gpu-engine 1050 --gpu-memclock 1470 --gpu-vddc 1.081

and now im getting 315 and 285 on each gpu.  When in doubt, reboot.

Further update, changed clocks to gpu 1000 and mem 1640 and now pulling about 350 khs per card.  These are sapphire vapor-x 7970 ghz flashed w/ dual-x bios.

I have 2 sapphire 7970s i get 750khs each mining doge.  First i tried using the same settings i used for doge mining on the vert modified cgminer:

setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_USE_SYNC_OBJECTS 1

Set PARAM0=--scrypt -I 13 --worksize 256 --thread-concurrency 8192 --no-submit-stale
Set PARAM1=--lookup-gap 2 --gpu-threads 2
Set PARAM2=--gpu-engine 1080 --gpu-memclock 1496 --gpu-vddc 1.100 --gpu-powertune 20
Set PARAM3=-o  stratum+tcp://stratum.vertco.in:8080 -u x.x -p x
cgminer %PARAM0% %PARAM1% %PARAM2% %PARAM3%
 
with this i get 275 khs on gpu 1 and 14 khs on gpu 2

I tried this from earlier in this thread:

setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_USE_SYNC_OBJECTS 1

cgminer --scrypt -o stratum+tcp://stratum.vertco.in:8080 -u x.x -p x --gpu-platform 0 -w 256 -v 1 -I 13 -g 2 -l 1 -T --thread-concurrency 8192

I get an error that all devices are disabled and i cant mine

I found these settings as well and tried them w/ vert miner:

setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_USE_SYNC_OBJECTS 1

vertminer -o stratum+tcp://stratum.vertco.in:8080 -u x.x -p x --scrypt-vert --intensity 13 --worksize 256 --gpu-threads 2 --lookup-gap 2 --thread-concurrency 8192 --temp-target 75 --gpu-fan 35-95 --auto-fan --gpu-engine 1050 --gpu-memclock 1470 --gpu-vddc 1.081

Like this i'm back to 260khs and 14 khs.  What am i doing wrong?

my 280xs, which usually get 741kh/s @ 80ºc are getting 234kh/s @ 73ºc, although i saw the one cardjump to 290kh/s @ 76ºc. had to drop tc to 4096 and i to 12 to avoid tons of hw. i could probably tinker with them more, but i think i'll leave them be for now. been pushin them at 80ºc nearly nonstop for a while.

need 8G ram

Yes! Same problem here, what the hell is this.. Anything over 13 gives me loads of HW errors on Sapphire 7950 Boost.

Hurrah!
Have given this a go and no HW Errors.
Although i12 is the limit, but getting around 170kh from the 7850 (gpu = 1000 mem 1250)

Has anyone checked out Vertcoin price;

Solid rocket boosters AND main engines are definitely throttled up 

Sounds like you need different settings.  in particular it probably isn't handling two threads well.  Can you put it into a different rig?

I added +4 GB of RAM and I can mine. However, in a rig of 1 7950 and 1 280x, only the 280x is working properly with 290kh/s. The other one gets 7-8 kh/s. Any ideas?

No luck. The only way I was able to run it was with Intensity of 12:

]
,
"intensity" : "12",
"vectors" : "1",
"worksize" : "256",
"kernel" : "scrypt",
"lookup-gap" : "2",
"thread-concurrency" : "8192",
"shaders" : "0",
"expiry" : "120",
"gpu-threads" : "2"
}

Getting 170kh/s. Same as my 7850. Might be the limit of that chipset.

I'd uninstall and reinstall drivers. Sorry I don't have any better fix 

I have been using the -d flag with normal scrypt coins, but for some reason it keeps freezing my computer when I try to use it with vertcoin. First instance of the miner works just fine, but when I run the second one my computer always crashes.

With low intensity I sometimes try adding a second core (-g 2). Its worked on some of my cards. Not sure about your 4870.

I'm using a old 4870, and this is my config for modified vertminer

cgminer.exe --scrypt -o stratum+tcp://s.eu.vertcoin.org:3333 -u user.worker -p pass -I 10 -g 1 --thread-concurrency 1024 --temp-cutoff 90 -v 1 -w 256

I'm getting ~25Kh/s (pool says 10-15) while on normal cgminer (other coins) I'm getting ~100Kh/s. I lowered intensity (was 2048) and thread concurrency (was 11) so there are no HW errors.

Is there anything more I can do?

If you run different miners just put in the -d flag. -d 0 is for your first GPU and -d 1 is for your second, etc.

I think you can run both off of the same cgminer by putting a comma after the first value (e.g.: --thread-concurrency 4000,5000) but I have never done it this way and I think that you can't use two different inputs for the -g flag.

im using 2x 5870 with these settings

cgminer.exe --scrypt -o stratum+tcp://pool:port -u user -p x -v 1 -w 256 -I 12

(vert modified cgminer.exe)

but i dont know how to separate workers, can you please help me¡ do you suggest any changes?

Is there a way to run two instances of the miner at the same time? I have been mining with 280x and 5850 and since 280x likes 2 threads and 5850 likes just one I have been running two cgminers at the same time when I mine "normal" scrypt coins just to use the proper settings for each card. When I try to do that with the modified miner it just crashes my computer every time. I can mine with just one miner, but then I can't use the optimal setting since I have to use just one thread for each card. Am I missing something or should I just be happy that I can mine at all? 

Just posted this in the main thread thinking it was this thread . . whoops. Anyways . .

Just wondering if anyone here using 290 non x's is reading asic quality?

I'm running 5x 290 non x's and am seeing a big difference in my 3x hynix ones @ different asic qualities.

The %'s are: 81,77 and 73.

The 81 and 77 I can push fine @ 1075 core/1470 mem . . but the 73 gets sick unless I downclock it to 1000/1500 at any TC.

Not looking for the best config here. . .I'm going to spend some time tuning them better when I get time but it would be interesting to see if anyone else is running into similar issues?

I'm thinking that somewhere between 77 and 73% there's a point where the higher clocks get unstable.

If you don't know what I'm talking about, I'm reading asic quality from GPU-Z by right clicking the top bar and scrolling down to read asic quality in windows.

Also: for the elpida I'm always in the sub 70 level @ 69-69.4%