CCminer(SP-MOD) Modded GPU kernels.

[sp_]

Hmmm..... sp mod 52 crashed when I tried algo x15 intensity 22.  
...
this is on EVGA 750ti running @ stock
...
Intensity 19 is the best @ 3175 khs.  20 gave a net loss of 150k per hash vs no setting.

x15? THe hashrate you are getting is closer to x11 performance. X15 should be around 2100 khash on this card.

[1714917808]

1714917808

[1714917808]

1714917808

[1714917808]

1714917808

[1714917808]

1714917808

[sp_]

-Faster x11
-Faster quark

1.5.53(sp-MOD) is available here: (17-06-2015)

https://github.com/sp-hash/ccminer/releases/tag/1.5.53

The sourcecode is available here:

https://github.com/sp-hash/ccminer

[sp_]

Make sure to delete this file if it's old:

(in the ccminer folder)

cudart32_65.dll

I had some problems on one of my rigs with release 53 and old drivers...

[djm34]

could you remove the cudareset from the main ?
It takes forever to start (especially if there is another miner running on an other card) and it interacts with other mining session.
Not mentionning you are doing it at the start of every algo...

[sp_]

done

[CapnBDL]

@sp

Quark algo:

Testing release_53. This version is running faster by ~25kh & so far very stable. Your 7z unzipped to be 18369kb in size while my Win Visual compile sized at 18381kb. Not sure why the difference. What I have found is that the older releases took up to 50 solved shares before reaching 'full speed', while this release ramped up after ~10 shares. All of this, of course, is variable due to network speeds, algo, etc. Using -i 22.9 -r 5 -R 10 --cpu-priority 2 switches. vRam usage is low and my intensity could be increased, but I like it as is. I am basing this info while mining 'quark' on WestHash (usa.nicehash.com). Getting ready to try on donation link.

ASUS i7 z77 intelchip, Win8.1, 16GB ram, ASUS 750Ti-Series_54 2G, GPUTweak OCed.

LOOKIN' GOOD sp_, djm, & gang.

@sp_, djm34
edit: I keep getting 'failure to connect' on this; donate-sp.granitecoin.com:7919 for quark algo. Is there another?

Thx

[flipclip]

any linux guys having trouble with commit # 829?

If I try to run the newest miner, the command prompt drops to a "#" (pound sign) and no mining happens.  Reverting to an earlier version of the miner shows the same results as well. 

I compiled and ran fine at commit #829 (that is "Optimized Jh512 from the Tanguy Pruvot branch" correct?) last night (Ubuntu 14.04).  Are you running it with a bunch of switches?  Have you tried just running ccminer or ccminer --help? to see if it runs.

[CapnBDL]

Testing Release_53

x13 algo: error; " Cuda error in func 'quark_blake512_cpu_setBlock_80' at line 714 : an illegal instruction was encountered. "

I made no changes to previous setup being used with switches; -i 18.9 -r 5 -R 10 --cpu-priority 2 -a x13
       required a graphics driver reinstall or reboot.

x11 algo: error " Cuda error in func 'quark_blake512_cpu_setBlock_80_multi' at line 706 : an illegal instruction was encountered. "

Again, no changes from previous switches; -g 2 -i 18.3 -r 5 -R 10 --cpu-priority 2 -a x11 
       required graphics driver reinstall or reboot.

OK...that's all the testing I can do today. Am going back to 51.

later-

[chrysophylax]

@sp

Quark algo:

Testing release_53. This version is running faster by ~25kh & so far very stable. Your 7z unzipped to be 18369kb in size while my Win Visual compile sized at 18381kb. Not sure why the difference. What I have found is that the older releases took up to 50 solved shares before reaching 'full speed', while this release ramped up after ~10 shares. All of this, of course, is variable due to network speeds, algo, etc. Using -i 22.9 -r 5 -R 10 --cpu-priority 2 switches. vRam usage is low and my intensity could be increased, but I like it as is. I am basing this info while mining 'quark' on WestHash (usa.nicehash.com). Getting ready to try on donation link.

ASUS i7 z77 intelchip, Win8.1, 16GB ram, ASUS 750Ti-Series_54 2G, GPUTweak OCed.

LOOKIN' GOOD sp_, djm, & gang.

@sp_, djm34
edit: I keep getting 'failure to connect' on this; donate-sp.granitecoin.com:7919 for quark algo. Is there another?

Thx

its back up and running ...

the vps seemed to have a spat and locked the java threads that run the proxies ...

all good now ...

will look more indepth at them tomorrow ...

we are also running through these donation links ...

https://bitcointalk.org/index.php?topic=1089744 ...

#crysx

[sp_]

Testing Release_53

x13 algo: error; " Cuda error in func 'quark_blake512_cpu_setBlock_80' at line 714 : an illegal instruction was encountered. "

I made no changes to previous setup being used with switches; -i 18.9 -r 5 -R 10 --cpu-priority 2 -a x13
       required a graphics driver reinstall or reboot.

x11 algo: error " Cuda error in func 'quark_blake512_cpu_setBlock_80_multi' at line 706 : an illegal instruction was encountered. "

Again, no changes from previous switches; -g 2 -i 18.3 -r 5 -R 10 --cpu-priority 2 -a x11 
       required graphics driver reinstall or reboot.

OK...that's all the testing I can do today. Am going back to 51.

later-

Each release have different optimal intensity. You need to retune or run with default intensities.

[sp_]

@sp

Quark algo:

Testing release_53. This version is running faster by ~25kh & so far very stable. Your 7z unzipped to be 18369kb in size while my Win Visual compile sized at 18381kb. Not sure why the difference. What I have found is that the older releases took up to 50 solved shares before reaching 'full speed', while this release ramped up after ~10 shares. All of this, of course, is variable due to network speeds, algo, etc. Using -i 22.9 -r 5 -R 10 --cpu-priority 2 switches. vRam usage is low and my intensity could be increased, but I like it as is. I am basing this info while mining 'quark' on WestHash (usa.nicehash.com). Getting ready to try on donation link.

ASUS i7 z77 intelchip, Win8.1, 16GB ram, ASUS 750Ti-Series_54 2G, GPUTweak OCed.

LOOKIN' GOOD sp_, djm, & gang.

@sp_, djm34
edit: I keep getting 'failure to connect' on this; donate-sp.granitecoin.com:7919 for quark algo. Is there another?

Thx

I have compiled with the 31/5/2015 driver. Different drivers give different code sizes. And old drivers can crash the miner

[bensam1231]

Bugreport for Neoscrypt. Running 5 970s on the same system causes it to spit out "Cuda error in func 'scanhash-neoscrypt' at line 67 : out of memory."

Limiting it to any 4 devices with -d eliminates the problem. Tried running separate instances with different devices, doesn't work. System memory is fine. DJM's miner has the same problem.

you need more memory... or pagefile... (which is required to allocate that memory to the gpu...)
alternatively you can lower the intensity, that should free up some memory
other alternative, try to start 1 or 2 ccminer session

As mentioned running more then one instance does not fix this, system memory is fine. It doesn't get maxed out or anywhere close to it while monitoring it. I'll throw some more memory into the system and see what happens though. Currently it's at 4GB.

Haven't tried a lower intensity, this seems more like a bug then anything if that's the case.

Memory is not necessary, pagefile will most likely work. I have 14GB (!) of pagefile on my rigs because I think there was a cudaminer algo that only ran if I had tons of pagefile but it never used more than a few megabytes.
Same goes for sgminer to a certain degree, although I was never able to run it with more than 4 cards enabled.

I will try a bigger pagefile. The pagefile sits at the smallest possible size when the miner is in use and starting up, just like memory. I don't know why this will fix it. My current pagefile already maxes out at 8192, but sits at 1024.

In order to run 5 AMD cards you need to be on W8.1. W7 only allows 4. 6 cards requires a lot of hacking about to make it work no matter what.

Tripled the amount of memory and it fixed it. I assume just adding a extra gig would've done it too. The PC had 4GB in it and four cards ran on it, 5GB probably would've worked. Pagefile was the same size. Makes me wonder if the pagefile idea would've fixed it. I remember something similar when I used to mine Monero coin, even though it never really used the system memory, it still had to preload it through it.

[CapnBDL]

@sp

Quark algo:

Testing release_53. This version is running faster by ~25kh & so far very stable. Your 7z unzipped to be 18369kb in size while my Win Visual compile sized at 18381kb. Not sure why the difference. What I have found is that the older releases took up to 50 solved shares before reaching 'full speed', while this release ramped up after ~10 shares. All of this, of course, is variable due to network speeds, algo, etc. Using -i 22.9 -r 5 -R 10 --cpu-priority 2 switches. vRam usage is low and my intensity could be increased, but I like it as is. I am basing this info while mining 'quark' on WestHash (usa.nicehash.com). Getting ready to try on donation link.

ASUS i7 z77 intelchip, Win8.1, 16GB ram, ASUS 750Ti-Series_54 2G, GPUTweak OCed.

LOOKIN' GOOD sp_, djm, & gang.

@sp_, djm34
edit: I keep getting 'failure to connect' on this; donate-sp.granitecoin.com:7919 for quark algo. Is there another?

Thx

I have compiled with the 31/5/2015 driver. Different drivers give different code sizes. And old drivers can crash the miner

Yup...I know about tuning...I was just letting you know. I have the latest nVidia drivers, but would like to know if this code was written with Cuda6.5 or 7? I am still using 6.5. I can see I'll have to make some switch changes. Going to default, for now.

edit: never mind...I see it's 6.5.

[Angora]

@SP

Release 53 x11 I am getting around a 5k increase (3180K on the upper end)..   Haven't changed intensity from version 52. 
I'm mining StartCoin on suchpool and it is showing my hash @4.06M w/v53. v52 it showed my hash has 2.6M.

Pool readings were taken after running 1 hr each time. 
4.06M has been holding steady for 3 hrs now, which is ~800K faster than the card is showing on confirms.

Cuda 7.0 and latest drivers (353.06).

[Slava_K]

_SP!!!!! plz make next commit!!! current commit number is 666... 

[joblo]

Testing Release_53

x13 algo: error; " Cuda error in func 'quark_blake512_cpu_setBlock_80' at line 714 : an illegal instruction was encountered. "

I made no changes to previous setup being used with switches; -i 18.9 -r 5 -R 10 --cpu-priority 2 -a x13
       required a graphics driver reinstall or reboot.

x11 algo: error " Cuda error in func 'quark_blake512_cpu_setBlock_80_multi' at line 706 : an illegal instruction was encountered. "

Again, no changes from previous switches; -g 2 -i 18.3 -r 5 -R 10 --cpu-priority 2 -a x11 
       required graphics driver reinstall or reboot.

OK...that's all the testing I can do today. Am going back to 51.

later-

That looks like a problem with  the  ccminer executable. Try recompiling or redownloading as appropriate.

[djm34]

Testing Release_53

x13 algo: error; " Cuda error in func 'quark_blake512_cpu_setBlock_80' at line 714 : an illegal instruction was encountered. "

I made no changes to previous setup being used with switches; -i 18.9 -r 5 -R 10 --cpu-priority 2 -a x13
       required a graphics driver reinstall or reboot.

x11 algo: error " Cuda error in func 'quark_blake512_cpu_setBlock_80_multi' at line 706 : an illegal instruction was encountered. "

Again, no changes from previous switches; -g 2 -i 18.3 -r 5 -R 10 --cpu-priority 2 -a x11 
       required graphics driver reinstall or reboot.

OK...that's all the testing I can do today. Am going back to 51.

later-

That looks like a problem with  the  ccminer executable. Try recompiling or redownloading as appropriate.

most likely intensity too high

[CapnBDL]

Testing Release_53

x13 algo: error; " Cuda error in func 'quark_blake512_cpu_setBlock_80' at line 714 : an illegal instruction was encountered. "

I made no changes to previous setup being used with switches; -i 18.9 -r 5 -R 10 --cpu-priority 2 -a x13
       required a graphics driver reinstall or reboot.

x11 algo: error " Cuda error in func 'quark_blake512_cpu_setBlock_80_multi' at line 706 : an illegal instruction was encountered. "

Again, no changes from previous switches; -g 2 -i 18.3 -r 5 -R 10 --cpu-priority 2 -a x11  
       required graphics driver reinstall or reboot.

OK...that's all the testing I can do today. Am going back to 51.

later-

That looks like a problem with  the  ccminer executable. Try recompiling or redownloading as appropriate.

Yep...I have to agree. I made some small changes to the intensity, etc., and still ran into the errors. I'll do a new compile and re-test as I don't believe that it would take that much of a change in the -i switch between the different releases. djm34 could be right but...Thx for the idea.

[go6ooo1212]

Guys, which values do you reach on quark with r.53?
I've got 6.3-6.4 with 750ti, and 15.8-16.2 with 970...

[Zels]

6600 kh on quark with 750ti here