CCminer(SP-MOD) Modded NVIDIA Maxwell / Pascal kernels. - page 1112.

bathrobehero

legendary

Activity: 2002

Merit: 1051

ICO? Not even once.

Quote from: frazier34567 on April 30, 2015, 05:27:49 PM

Was it a Ribbon or USB cable riser?

On Topic, the 1.5.48 build on my single 750 TI card gets between 5866 to 7200 KH/s Quark using the -i 22 -g 5 usually closer to 6300. However, I see a drop on my 6 750 ti card unit running at 36600 vs 37083 on build 1.5.47 commit 749. I have gone back to using the 1.5.47 commit 749 for it, with no extra parameters it is getting 37150 kh/s on screen and seems like it is getting more consistently on westhash (36.6 to 44.3MH/s). Windows 8.1 Pro no OC or bios flash EVGA SC cards.

USB riser, open rig.

frazier34567

member

Activity: 95

Merit: 10

Quote from: Epsylon3 on April 30, 2015, 03:52:55 PM

Quote from: bathrobehero on April 30, 2015, 03:04:35 PM

Offtopic: Just as a reminder for people with TDP modded Gigabyte 750 Ti OC (GV-N75TOC-2GI) cards; expect your cards to die. Out of 6, now 2 of them died on me in 11 months (no fan, no screen, nothing). And I 'only' increased the TDP to 45W (instead of the usual 60W) and only after 3-4 months of stock use. I usually had +140 to +170 mhz on the core with stock mem clocks. Temps were always around 60°C with the max temp target set to 65°C.

I also had one of my black edition broken, which was just replaced (3 years warranty) and there is something interresting to report...

You can see here an exploded cap C704 on the center...

On the new one, all these caps were removed...

The problem happened after a while when i tried mining on a PCX riser (1x to 16x, i suspect the 3.3V to be unstable on that)..

Was it a Ribbon or USB cable riser?

On Topic, the 1.5.48 build on my single 750 TI card gets between 5866 to 7200 KH/s Quark using the -i 22 -g 5 usually closer to 6300. However, I see a drop on my 6 750 ti card unit running at 36600 vs 37083 on build 1.5.47 commit 749. I have gone back to using the 1.5.47 commit 749 for it, with no extra parameters it is getting 37150 kh/s on screen and seems like it is getting more consistently on westhash (36.6 to 44.3MH/s). Windows 8.1 Pro no OC or bios flash EVGA SC cards.

hashbrown9000

sr. member

Activity: 427

Merit: 250

good to know about the 750ti cards as i have 12 of those exact ones. I wonder how they will fare in open-air mining rigs like mine.

Epsylon3

legendary

Activity: 1484

Merit: 1082

ccminer/cpuminer developer

Quote from: bathrobehero on April 30, 2015, 03:04:35 PM

Offtopic: Just as a reminder for people with TDP modded Gigabyte 750 Ti OC (GV-N75TOC-2GI) cards; expect your cards to die. Out of 6, now 2 of them died on me in 11 months (no fan, no screen, nothing). And I 'only' increased the TDP to 45W (instead of the usual 60W) and only after 3-4 months of stock use. I usually had +140 to +170 mhz on the core with stock mem clocks. Temps were always around 60°C with the max temp target set to 65°C.

I also had one of my black edition broken, which was just replaced (3 years warranty) and there is something interresting to report...

You can see here an exploded cap C704 on the center...

On the new one, all these caps were removed...

The problem happened after a while when i tried mining on a PCX riser (1x to 16x, i suspect the 3.3V to be unstable on that)..

bathrobehero

legendary

Activity: 2002

Merit: 1051

ICO? Not even once.

Quote from: 5w00p on April 30, 2015, 03:16:39 PM

bathrobehero,

What is the advantage for mining when increasing the TDP of a 750Ti via a flashed BIOS?

I imagine that it allows a higher overclock and therefore higher hashrate. But, I have had a PNY 750Ti OC for over a year, and I never flashed it. I never saw anyone posting any hashrates for algorithms that I cared to mine that were significantly higher than what my card would do (with my particular card's stock BIOS - OC maximized). Sure, maybe a less than 5% faster hashrate than my card, but that just didn't seem worth the risk and effort.

Am I correct or mistaken?

It's both higher overclock and more stability with overclocks. Even though I'll never get how some people can overclock their 750 Ti cards to ~1500 mhz and be stable, with increased TDP I could overclock to around 1350-1380 mhz and still have the rig running for weeks without crashing (for most algorithms that is). I always ready to jump onto new coins and solomine as much as I can until the pools join and the difficulty shoots up so I appreciate higher hashrate over hashrate/power consumption figures.

Quote from: sp_ on April 30, 2015, 03:38:00 PM

Just replace the bios with the gforce black bios. Same cooler, same power setting, more hash

Never heard of it but I don't think that is any different from a TDP mod. I just think the voltage regulation modules on these particular cards are just not equipped to handle anything over stock figures for long periods of time.

sp_

legendary

Activity: 2954

Merit: 1087

Team Black developer

Quote from: sp_ on April 29, 2015, 03:38:53 PM

Quote from: flipclip on April 29, 2015, 03:35:02 PM

Can the g parameter be used per device, example two cards set differently with -g 4,3 ? (per if you are running a mix of cards)

you have to do
-d 3,4 and then -g x

Not sure it works.
[/quote]

Should work, but the stats doesn't display the results correct per card. On the pool it should be ok though

sp_

legendary

Activity: 2954

Merit: 1087

Team Black developer

Quote from: bathrobehero on April 30, 2015, 03:04:35 PM

Offtopic: Just as a reminder for people with TDP modded Gigabyte 750 Ti OC (GV-N75TOC-2GI) cards; expect your cards to die. Out of 6, now 2 of them died on me in 11 months (no fan, no screen, nothing). And I 'only' increased the TDP to 45W (instead of the usual 60W) and only after 3-4 months of stock use. I usually had +140 to +170 mhz on the core with stock mem clocks. Temps were always around 60°C with the max temp target set to 65°C.

Just replace the bios with the gforce black bios. Same cooler, same power setting, more hash

fenomenhaa

sr. member

Activity: 248

Merit: 250

@SP another donation on the way, yesterday some of my cards mined for you ,thanks for your hard work. Btc : 4d6d292c590173b279ae0b77ea0a2ab40c9369b8534aa7fe60c8c1f7af522dd5 .

Did anyone test g parameter on quark or other algos.?it helps or not. still i mine with default settings. Grin

5w00p

hero member

Activity: 644

Merit: 502

bathrobehero,

What is the advantage for mining when increasing the TDP of a 750Ti via a flashed BIOS?

I imagine that it allows a higher overclock and therefore higher hashrate. But, I have had a PNY 750Ti OC for over a year, and I never flashed it. I never saw anyone posting any hashrates for algorithms that I cared to mine that were significantly higher than what my card would do (with my particular card's stock BIOS - OC maximized). Sure, maybe a less than 5% faster hashrate than my card, but that just didn't seem worth the risk and effort.

Am I correct or mistaken?

bathrobehero

legendary

Activity: 2002

Merit: 1051

ICO? Not even once.

Offtopic: Just as a reminder for people with TDP modded Gigabyte 750 Ti OC (GV-N75TOC-2GI) cards; expect your cards to die. Out of 6, now 2 of them died on me in 11 months (no fan, no screen, nothing). And I 'only' increased the TDP to 45W (instead of the usual 60W) and only after 3-4 months of stock use. I usually had +140 to +170 mhz on the core with stock mem clocks. Temps were always around 60°C with the max temp target set to 65°C.

I read about others with similar issues a couple of times. An old post but this was close by:

Quote

blahij
May 11th, 2014 at 03:08
I had 6 flashed hashing @ 300kh for a month or two. Had two burn themselves out day before and then another the next day.
Temps were fine and I can see some discoloring on the back and one component looks blown-out?
Mine were Gigbyte and I used the posted BIOS on this site.

Quote from http://cryptomining-blog.com/1014-how-to-increase-the-geforce-gtx-750-ti-power-target-limit/

flipclip

member

Activity: 111

Merit: 10

Quote from: DragonSlayer on April 30, 2015, 11:06:35 AM

Quote from: flipclip on April 30, 2015, 09:39:22 AM

Quote from: go6ooo1212 on April 29, 2015, 04:18:40 PM

Quote from: go6ooo1212 on April 29, 2015, 03:57:02 PM

Which improvement we could expect on QUBIt with GTX970/750Ti

...
Especially when I have both of the cards on same rig - its total mess using the "g" parameter...

Couldn't you start up two instances of ccminer utilizing the -d? One instance would be for your 970 the other instance for your 750Ti, and you could set the -i -g parameters per instance.

To the best of my knowledge, -g does not work with -d parameters. I hope someone can prove me wrong.

I just got the idea from what sp_ posted earlier:

Quote from: sp_ on April 29, 2015, 03:38:53 PM

Quote from: flipclip on April 29, 2015, 03:35:02 PM

Can the g parameter be used per device, example two cards set differently with -g 4,3 ? (per if you are running a mix of cards)

you have to do

-d 3,4 and then -g x

Not sure it works.

DragonSlayer

full member

Activity: 181

Merit: 100

Quote from: flipclip on April 30, 2015, 09:39:22 AM

Quote from: go6ooo1212 on April 29, 2015, 04:18:40 PM

Quote from: go6ooo1212 on April 29, 2015, 03:57:02 PM

Which improvement we could expect on QUBIt with GTX970/750Ti

...
Especially when I have both of the cards on same rig - its total mess using the "g" parameter...

Couldn't you start up two instances of ccminer utilizing the -d? One instance would be for your 970 the other instance for your 750Ti, and you could set the -i -g parameters per instance.

To the best of my knowledge, -g does not work with -d parameters. I hope someone can prove me wrong.

tbearhere

legendary

Activity: 3164

Merit: 1003

Quote from: scryptr on April 29, 2015, 10:42:19 AM

Quote from: tbearhere on April 29, 2015, 10:06:17 AM

Quote from: scryptr on April 27, 2015, 11:10:55 PM

Quote from: mendoza1468 on April 27, 2015, 09:10:25 PM

I have a question, if sometimes on yaamp the connection is lost and isn't restarting, i would like to add another pool. How can i do this. (with last ccminer1.5.47)
Am i the only one who lost connection on yaamp?
For now i use :
ccminer.exe -R 5 -T 5 -q -a quark -o stratum+tcp://yaamp.com:4033 -u 5LA1mw7JZKKVf3LK9KLAJSzQZzNwU9Z3Rp -p d=0.002

Can someone explain me what i should add to insert another pool after this one? (round robin, failover or someting else)

P.S sometime if i closed ccminer and restart it im able to log again and mine ...

Thanks for the help

LOOPER--

This is how:

:loop

ccminer.exe -r 1 -R 5 -a quark -o stratum+tcp://us-west01.miningrigrentals.com:3333 -u me -p x -d 0,1,2

taskkill -t -f /im ccminer*
timeout -t 1

ccminer.exe -r 1 -R 5 -a quark -o stratum+tcp://us-central01.miningrigrentals.com:3333 -u me -p x -d 0,1,2

taskkill -t -f /im ccminer*
timeout -t 1

goto loop

This is a slight revision of Bathrobehero's script. Use the "-R" switch, and the "-r" switch. The script above puts in a 5 second delay (-R 5) before switching to the lower pool in the batch file, and to try to reconnect once (-r 1). Yaamp uses it to switch between algos. As used above, it switches between pools. Yamp.com has an example script linked on the homepage for algo switching.

If you execute the command "ccminer --help >> help.txt" in a windows command box, you will get a text file explaining all the ccminer commands.

--scryptr

This looks good to me but the problem still is at the goto loop I think. Again at the crash line ( goto loop ) it says can't find ccminer. Or is it the 1 second time delay that may keep it from crashing ?

CRASHES--

I think that may be the intention for "-t 1", but now I wonder if crashes are because of overclocking ccminer for performance. The batch file works, but ccminer triggers the "program stopped working" error messages. I don't know if ccminer is supposed to stay runing between connection attempts. --scryptr

My pool just went down and ccminer crashed on my other 2 computers. This one didn't and never does and all computers are overclocked. Ones that crashed are vista and windows 8.1 and this one is windows 8.1. Now I'll try the -t 1 right before the goto loop to see if that helps. That * right after ccminer* is the same as .exe like this ccminer* same as ccminer.exe ??
Thx
EDIT: I just disconnected my ithernet cable and it went through the commands .. but in the command window it says ccminer.exe can't be found. But it didn't crash that way.
ps What cbuchner1 put in cudaminer needs to be put into ccminer. With cudaminer these bat files will work..even a 1 bat file restart like...

:start
cudaminer.exe -l T5x24 -C 2 -i 0 -o stratum+tcp://us.coinium.org:3340 -u me -p x -d 0 --time-limit 900
goto start

flipclip

member

Activity: 111

Merit: 10

Quote from: go6ooo1212 on April 29, 2015, 04:18:40 PM

Quote from: go6ooo1212 on April 29, 2015, 03:57:02 PM

Which improvement we could expect on QUBIt with GTX970/750Ti

...
Especially when I have both of the cards on same rig - its total mess using the "g" parameter...

Couldn't you start up two instances of ccminer utilizing the -d? One instance would be for your 970 the other instance for your 750Ti, and you could set the -i -g parameters per instance.

chrysophylax

legendary

Activity: 2912

Merit: 1091

--- ChainWorks Industries ---

Quote from: sp_ on April 30, 2015, 07:30:01 AM

I think cryptonight should be merged into ccminer as well.

this fork is going to be one hell of a miner ... you could say a forking hell of a miner Wink

...

tanx so much sp ...

donations still mining for you ... will add the other algos tomorrow afternoon adelaide time ...

westhash is running fine with quark - which confirms it is a yaamp issue ...

#crysx

sp_

legendary

Activity: 2954

Merit: 1087

Team Black developer

I think cryptonight should be merged into ccminer as well.

antonio8

legendary

Activity: 1400

Merit: 1000

Quote from: sp_ on April 30, 2015, 06:02:30 AM

Quote from: chrysophylax on April 30, 2015, 05:47:10 AM

what i was wondering is whether the -a spread algo that in THIS fork of ccminer ( that doesnt work with pool mining ) is going to be replaced with the optimized version that you sent privately ( which works flawlessly with the pool ) and will be incorporated into this fork? ...
that would be fantastic if this is so ... and we would have no issue donating much more hashpower to see that happen ...

Ok. I can merge the kernals into the latest branch. Still private for the donators.

Any updates soon on the private cryptonight miner soon?

chrysophylax

legendary

Activity: 2912

Merit: 1091

--- ChainWorks Industries ---

Quote from: sp_ on April 30, 2015, 06:02:30 AM

Quote from: chrysophylax on April 30, 2015, 05:47:10 AM

what i was wondering is whether the -a spread algo that in THIS fork of ccminer ( that doesnt work with pool mining ) is going to be replaced with the optimized version that you sent privately ( which works flawlessly with the pool ) and will be incorporated into this fork? ...
that would be fantastic if this is so ... and we would have no issue donating much more hashpower to see that happen ...

Ok. I can merge the kernals into the latest branch. Still private for the donators.

if you want to keep it private - that is entirely up to you ...

but that would be great if you could ...

the last one we compiled here works great - and you have done some great work with the optimizations ...

that is so much appreciated sp ...

tanx ...

#crysx

sp_

legendary

Activity: 2954

Merit: 1087

Team Black developer

Quote from: chrysophylax on April 30, 2015, 05:47:10 AM

what i was wondering is whether the -a spread algo that in THIS fork of ccminer ( that doesnt work with pool mining ) is going to be replaced with the optimized version that you sent privately ( which works flawlessly with the pool ) and will be incorporated into this fork? ...
that would be fantastic if this is so ... and we would have no issue donating much more hashpower to see that happen ...

Ok. I can merge the kernals into the latest branch. Still private for the donators.

chrysophylax

legendary

Activity: 2912

Merit: 1091

--- ChainWorks Industries ---

Quote from: sp_ on April 30, 2015, 05:28:49 AM

Quote from: chrysophylax on April 30, 2015, 05:16:13 AM

ok - agreed ... but 75% of the farm we have are gigabyte 750ti oc lp cards ...
the spread algo in this ccminer fork wont work with the pool - but the private spreadminer does - and that is built for compute 5.0 also isnt it? ...
or am i missing the point here ... as i am really tired right this minute ...
#crysx

You are mixing threads here. There was a guy who asked if he could use this version to mine on the gtx660ti.
Your gigabytes will work just fine, and I will make you richer. he-he

not mixing threads at all sp ... there is -a spread algo in your fork - and i tried using it only to find it will not run with the pool that we mine ( which is nonce-pool ) ...

the private ( paid ) spreadminer that you optimized and forked from tsiv works just fine with the pool ...

what i was wondering is whether the -a spread algo that in THIS fork of ccminer ( that doesnt work with pool mining ) is going to be replaced with the optimized version that you sent privately ( which works flawlessly with the pool ) and will be incorporated into this fork? ...

that would be fantastic if this is so ... and we would have no issue donating much more hashpower to see that happen ...

#crysx

Topic: CCminer(SP-MOD) Modded NVIDIA Maxwell / Pascal kernels. - page 1112. (Read 2347601 times)