CCminer(SP-MOD) Modded GPU kernels.

[polanskiman]

With ccminer v7 from DJm34 I have an average of:
x11 = 2605 Khash

With ccminer by SP_ release 8 I have an average of:
X11 = 2800 Khash

That's a 195 Khash average difference. It wont make me any richer but it is always welcome

You should try the 980. Up from 7MHASH to 9. 30% faster.

Naaa, wont invest in any GPU anymore. Not worth it.

[Travis9x]

With ccminer v7 from DJm34 I have an average of:
x11 = 2605 Khash

With ccminer by SP_ release 8 I have an average of:
X11 = 2800 Khash

That's a 195 Khash average difference. It wont make me any richer but it is always welcome

You should try the 980. Up from 7MHASH to 9. 30% faster.

Naaa, wont invest in any GPU anymore. Not worth it.

Unless you're a gamer.  Then the 900 series is very worth it. 

[polanskiman]

With ccminer v7 from DJm34 I have an average of:
x11 = 2605 Khash

With ccminer by SP_ release 8 I have an average of:
X11 = 2800 Khash

That's a 195 Khash average difference. It wont make me any richer but it is always welcome

You should try the 980. Up from 7MHASH to 9. 30% faster.

Naaa, wont invest in any GPU anymore. Not worth it.

Unless you're a gamer.  Then the 900 series is very worth it. 

This is a mining forum....

[Travis9x]

With ccminer v7 from DJm34 I have an average of:
x11 = 2605 Khash

With ccminer by SP_ release 8 I have an average of:
X11 = 2800 Khash

That's a 195 Khash average difference. It wont make me any richer but it is always welcome

You should try the 980. Up from 7MHASH to 9. 30% faster.

Naaa, wont invest in any GPU anymore. Not worth it.

Unless you're a gamer.  Then the 900 series is very worth it. 

This is a mining forum....

I know.  The point is, if you're a gamer, then the 900 series is worth it, because they'll play your games very well, and they also mine very well when you're not currently gaming.

[jjjordan]

Travis, what the heck man...
The dude said - "...won't invest... ...not worth it."
Definition of invest: expend money with the expectation of achieving a profit
And while we all agree the 980 is a great card overall  - it is not a good subject to invest in.
It's a very good thing to spend your money on though.
So if you are talking about investing with the intent to gain fun instead of profit,
that is not investing, but spending. Which I like to do too.
If your nitpicking is caused solely because you had time to "invest",
please do it somewhere else and let whoever has a comment for
"CCminer modded NVIDIA Maxwell kernals by SP" to post it here.

Oh, and if you want your post to be the last in the 980 investment discussion please do it
and let us chat and wait for the next modded ccminer in peace.

[scryptr]

NEOSCRYPT:

Make it worth the wait.  Then game your brains out after ROI.  First, kick some AMD biscuits.  Just for the thrill of it...

       --scryptr

p.s.  tpruvot appears to be adapting some cgminer code for monitoring the nVidia cards.       --scryptr

[djm34]

Travis, what the heck man...
The dude said - "...won't invest... ...not worth it."
Definition of invest: expend money with the expectation of achieving a profit
And while we all agree the 980 is a great card overall  - it is not a good subject to invest in.
It's a very good thing to spend your money on though.
So if you are talking about investing with the intent to gain fun instead of profit,
that is not investing, but spending. Which I like to do too.
If your nitpicking is caused solely because you had time to "invest",
please do it somewhere else and let whoever has a comment for
"CCminer modded NVIDIA Maxwell kernals by SP" to post it here.

Oh, and if you want your post to be the last in the 980 investment discussion please do it
and let us chat and wait for the next modded ccminer in peace.

thanks for the clarification dear mod 
but I think everybody is entitled to his own opinion (and actually you are the one nitpicking...)

[SS2006]

Release 8, best x11 performance ive seen on a gtx 970 (OC 200 MH Clock)
7500 KHASH, amazing.

[alexbg21]

Palit JETSTREAM GTX 970 X11 slightly OCed - 7600 - 7700 Mhs

[jpouza]

Release 8 is crashing with the new nvidia drivers 344.75 with the 980 cards.

[Travis9x]

Travis, what the heck man...
The dude said - "...won't invest... ...not worth it."
Definition of invest: expend money with the expectation of achieving a profit
And while we all agree the 980 is a great card overall  - it is not a good subject to invest in.
It's a very good thing to spend your money on though.
So if you are talking about investing with the intent to gain fun instead of profit,
that is not investing, but spending. Which I like to do too.
If your nitpicking is caused solely because you had time to "invest",
please do it somewhere else and let whoever has a comment for
"CCminer modded NVIDIA Maxwell kernals by SP" to post it here.

Oh, and if you want your post to be the last in the 980 investment discussion please do it
and let us chat and wait for the next modded ccminer in peace.

I wasn't nitpicking anything.  It was meant as somewhat of a joke, since most gamers consider all their gaming hardware to be investments.
Lighten up, man. 

[sp_]

Yesterday I improved Simd. Compress1 and compress2 can run in the same block on the maxwell,reduced the size of the constmem tables. (24 bit multiplication is faster on the maxwell). X11 seems to go 20-50KHASH faster on the 750ti, but slower on compute 3.5 and below devices.  The sourcecode is commited to github.

I am going to merge 1.4.9 with my kernal changes soon. Release 9 will work with the latest drivers and will be based on 1.4.9/1.5.0 with API support, Intensity parameter and bugfixes.

[Amph]

do you have a 970 test for every algo, or x algo at least?

[Epsylon3]

wow... you missed to add x11/cuda_x11_luffa512_Cubehash.cu in Makefile.am

Sorry, i made some other changes for pools and jansson but should be easier to cherry-pick

Code:

                          x11/cuda_x11_shavite512.cu x11/cuda_x11_simd512.cu x11/cuda_x11_echo.cu \
+                         x11/cuda_x11_luffa512_Cubehash.cu \
                          x13/x13.cu x13/cuda_x13_hamsi512.cu x13/cuda_x13_fugue512.cu \

but hmm this commit seems to break x11 algo (tested on your repo on linux 750Ti) and miss the 80 max reg tweak

[sp_]

Yeah, had to rollback the echo change. It broke the chain.

[sp_]

do you have a 970 test for every algo, or x algo at least?

I am testing on 650,750,750TI,970 and 980. but mostly on 750ti.

[Schleicher]

I've made a pull request on github.

[Epsylon3]

works fine with the echo Revert... and the change was not faster

What is the perf on x11 for the 650 ? i had one last year, just curious, it was a good card for games.

[sp_]

The 650 is doing pretty bad. 727KHASH@x11. The merged simd code (compress1/compress2) into one kernal, reduced the performance by 10KHASH. (1.3%) But Cudaarch is not possible to check in host code.

[sp_]

I've made a pull request on github.

Looks good, I have merged you changes.