Author

Topic: CCminer(SP-MOD) Modded NVIDIA Maxwell / Pascal kernels. - page 1031. (Read 2347601 times)

legendary
Activity: 1764
Merit: 1024
New release.


-Faster quark (optimized bmw-512)
-Added the credit algorithm (from the djm34 branch)
-Added the c11 algorithm(from the tprovot branch, but all kernals are replaced)

1.5.56(sp-MOD) is available here: (22-07-2015)

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

The sourcecode is available here:

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


Credits looks relatively unused and c11 doesn't seem to have a coin besides protoshares that died out. Is that true?
sp_
legendary
Activity: 2926
Merit: 1087
Team Black developer
haven't checked, but it is probably still there  Grin

Please merge and share  Grin

Donations are waiting.
legendary
Activity: 1400
Merit: 1050
did some optimization on quark: +75% (roughly)  Grin Grin
yeah... well  Grin
10 Mh/s on 750Ti?? That's really great!!
will you merge the mods, make a new repo or sell it?
well... that was a bug it is rather +1MH/s for 980/780 and 0.5MH/s for the 750ti


Since you posted your increase I have almost reached your increases in the opensourceminer.

Quark increase in the sp-mod written by me:

1.5.50(sp-mod) 29-may-2015

gtx 970: 15500   
gtx 750ti: 5650     


1.5.56(sp-mod) 22-july-2015

16254 +754KHASH (4.8%)
5920  +270khash (4.8%)

haven't checked, but it is probably still there  Grin
sp_
legendary
Activity: 2926
Merit: 1087
Team Black developer
did some optimization on quark: +75% (roughly)  Grin Grin
yeah... well  Grin
10 Mh/s on 750Ti?? That's really great!!
will you merge the mods, make a new repo or sell it?
well... that was a bug it is rather +1MH/s for 980/780 and 0.5MH/s for the 750ti


Since you posted your increase I have almost reached your increases in the opensourceminer.

Quark increase in the sp-mod written by me:

1.5.50(sp-mod) 29-may-2015

gtx 970: 15500  
gtx 750ti: 5650    


1.5.56(sp-mod) 22-july-2015

16254 +754KHASH (4.8%)
5920  +270khash (4.8%)
sp_
legendary
Activity: 2926
Merit: 1087
Team Black developer
New release.


-Faster quark (optimized bmw-512)
-Added the credit algorithm (from the djm34 branch)
-Added the c11 algorithm(from the tprovot branch, but all kernals are replaced)

1.5.56(sp-MOD) is available here: (22-07-2015)

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

The sourcecode is available here:

https://github.com/sp-hash/ccminer
sp_
legendary
Activity: 2926
Merit: 1087
Team Black developer
I might add bitcredits later.

is this your code dga?

https://github.com/dave-andersen/cudapts
dga
hero member
Activity: 737
Merit: 511
The AMD 280x  does about 4,400 cpm . Not sure how to convert this number.


When you say 420 MHash, that's individual sha512 executions that you're measuring?

Assuming it is, then:

In expectation, there are 2 collisions per "round" of momentum (creating 2^26 hashes).  Those create 4 "reported" collisions because they're counted symmetrically -- A colliding with B also lets you test B colliding with A.

So you should be getting *roughly* 1500 CPM on a 750ti.  The older miners were at about 800-900cpm.

NaN's miner on a 280x AMD gets about 6700 @ 200+ watts.

But I'm confused about the code - it seems to just be doing the sha hashing, not the collision-finding part of momentum?

  -Dave

Ahhhh, got it.  Credits.  Not bitcredits. Smiley
dga
hero member
Activity: 737
Merit: 511
The AMD 280x  does about 4,400 cpm . Not sure how to convert this number.


When you say 420 MHash, that's individual sha512 executions that you're measuring?

Assuming it is, then:

In expectation, there are 2 collisions per "round" of momentum (creating 2^26 hashes).  Those create 4 "reported" collisions because they're counted symmetrically -- A colliding with B also lets you test B colliding with A.

So you should be getting *roughly* 1500 CPM on a 750ti.  The older miners were at about 800-900cpm.

NaN's miner on a 280x AMD gets about 6700 @ 200+ watts.

But I'm confused about the code - it seems to just be doing the sha hashing, not the collision-finding part of momentum?

  -Dave
legendary
Activity: 1400
Merit: 1050
isn't it pRotoshares, not pHotoshares?
yeah, you are right. The Momentum algorithm.
understand nothing about this algo and coin.Is it bitshares?ans how to mine ?ıs there any pool or just solomine? Smiley)) Confsed... Smiley))

No it's bitcredits:

https://bitcointalksearch.org/topic/bitcredit-migrating-to-uniqredit-896133

but this algo is used in other coins as well.

not momentum algo related but the inertia on this thread is huge....
this is this coin: https://bitcointalksearch.org/topic/credits-cre-sha256-sidechain-new-block-deposit-feature-official-thread-1037244
not bitcredits...

sp_
legendary
Activity: 2926
Merit: 1087
Team Black developer
isn't it pRotoshares, not pHotoshares?
yeah, you are right. The Momentum algorithm.
understand nothing about this algo and coin.Is it bitshares?ans how to mine ?ıs there any pool or just solomine? Smiley)) Confsed... Smiley))

No it's bitcredits:
https://bitcointalksearch.org/topic/bitcredit-migrating-to-uniqredit-896133
but this algo is used in other coins as well. The momentum algo.
sp_
legendary
Activity: 2926
Merit: 1087
Team Black developer
I have added the C11 algo from the tpruvot branch. I have replaced all the kernals with my modded kernals, and the speed is faster.

8MHASH on the gtx 970
3MHASH on the 750ti
sr. member
Activity: 248
Merit: 250
isn't it pRotoshares, not pHotoshares?

yeah, you are right. The Momentum algorithm.
understand nothing about this algo and coin.Is it bitshares?ans how to mine ?ıs there any pool or just solomine? Smiley)) Confsed... Smiley))
legendary
Activity: 1470
Merit: 1114
i do like the office pc miner!  what cards are in there? they seem smaller than even the single fan evga type.

They are the EVGA 750ti's SC.  Single fan and no 6pin.  Powered USB risers FTW! Grin

what's the word on card orientation? I remember reading somewhere that the specific type of fan mount on GPUs was made for the card to be in the horizontal position (as if it were installed in a standard tower desktop), and that having them in the vertical orientation (such as in your rig) is what causes the fans to fail early.

rumor / misinformation?

Hmmm - I never heard the card/fan orientation rumor.  I have been running the cards in an open case vertical design now for about a year with no issues.  I dust the system about once every other month with canned air.  Other than that no problems.

I think it's rumour. Orientation makes sense with a large spinning mass. If horizontal the bearings don't have to support
the mass and wear longer. It applied to hard drives many years ago when they were much larger and less
reliable but a GPU fan is just too light to matter.

legendary
Activity: 1400
Merit: 1050
The AMD 280x  does about 4,400 cpm . Not sure how to convert this number.


When you say 420 MHash, that's individual sha512 executions that you're measuring?
if it is from my fork, it is just sha256... never coded momentum algo as far as I know  Grin
just wrong coin name  Roll Eyes
dga
hero member
Activity: 737
Merit: 511
The AMD 280x  does about 4,400 cpm . Not sure how to convert this number.


When you say 420 MHash, that's individual sha512 executions that you're measuring?
sp_
legendary
Activity: 2926
Merit: 1087
Team Black developer
The AMD 280x  does about 4,400 cpm . Not sure how to convert this number.
dga
hero member
Activity: 737
Merit: 511
isn't it pRotoshares, not pHotoshares?

yeah, you are right. The Momentum algorithm.

Oooh, sweet.  How does the speed compare to mine and to to NaN's AMD one?  (Mine's slow and hasn't been updated for about two years.  NaN's is... impressive.)
sp_
legendary
Activity: 2926
Merit: 1087
Team Black developer
isn't it pRotoshares, not pHotoshares?

yeah, you are right. The Momentum algorithm.
legendary
Activity: 1400
Merit: 1050

isn't it pRotoshares, not pHotoshares?
depending on the energy scale, it probably doesn't matter
legendary
Activity: 2716
Merit: 1094
Black Belt Developer
Jump to: