CCminer(SP-MOD) Modded GPU kernels.

[sp_]

This is working on my 960 using my build(42) on windows 7

ccminer.exe -a whirlpoolx -u 1CTiNJyoUmbdMRACtteRWXhGqtSETYd6Vd -p 1 -o stratum+tcp://stratum.nicehash.com:3343

but you need to wait some seconds, and then Nicehash will set the diff correcly.

My 960 is also getting abit higher hashrate. I use -500Mzh on the memoryclockrate.

[1714918550]

1714918550

[1714918550]

1714918550

[1714918550]

1714918550

[sp_]

-a whirlpoolx

Well done on WhirlpoolX - my 290X is doing about 520MH/s.

That is really good. my 980 is around 300MHASH with overclocking. I tried to rewrite to uint4 but it went slower. What did you do? did you remove the lookuptables?

[flipclip]

To fix the default intensity(linuxI) you have to change the hardcoded value 26U in the whirlpoolx.cuda file. I forgot to add support for the intensity parameter. Will do it later.

Guys, this kernal is very fast, very profitable and very free. Please donate some beers to the authors

I updated line 47 of x15/whirlpoolx.cu to be:
uint32_t throughput = pow(2,19);

That got rid of the memory error, but I've encountered another issue.  When I hash with both GPUs (both 750Ti) one GPU gives me a crazy hashing number, for example:
[2015-03-08 14:17:02] GPU #1: GeForce GTX 750 Ti, 667438585 kH/s
[2015-03-08 14:17:03] GPU #1: GeForce GTX 750 Ti, 661272871 kH/s
[2015-03-08 14:17:03] GPU #1: GeForce GTX 750 Ti, 658636297 kH/s
[2015-03-08 14:17:04] GPU #1: GeForce GTX 750 Ti, 657527143 kH/s
[2015-03-08 14:17:05] GPU #1: GeForce GTX 750 Ti, 658030840 kH/s
and fills up the screen very quickly with these values (I believe the other GPU at some point does output a valid looking hash, but it scrolls by so quickly).  But if I use the -d switch and hash with only one of the GPUs, I get proper values of:
[2015-03-08 14:26:07] accepted: 6/6 (100.00%), 74935 khash/s yay!!!
[2015-03-08 14:26:18] GPU #1: GeForce GTX 750 Ti, 74852 kH/s
[2015-03-08 14:26:18] accepted: 7/7 (100.00%), 74926 khash/s yay!!!
[2015-03-08 14:26:25] GPU #1: GeForce GTX 750 Ti, 75030 kH/s
Or
[2015-03-08 14:26:52] accepted: 1/1 (100.00%), 66489 khash/s yay!!!
[2015-03-08 14:26:55] GPU #0: GeForce GTX 750 Ti, 66484 kH/s
[2015-03-08 14:26:55] accepted: 2/2 (100.00%), 66488 khash/s yay!!!
[2015-03-08 14:27:05] GPU #0: GeForce GTX 750 Ti, 66661 kH/s
[2015-03-08 14:27:05] accepted: 3/3 (100.00%), 66517 khash/s yay!!!

[PVmining]

... but I've encountered another issue.  When I hash with both GPUs (both 750Ti) one GPU gives me a crazy hashing number, for .

brother. i confirm. i have the same issue with every source out there (linux version!).

[sp_]

There are some warnings when compiling. I will take a look later. Klaus_t has fixed intensity and double nounce reporting. This should improve the speed on the pool for low difficulty. I am checking the changes and merging them now.

[smolen]

That is really good. my 980 is around 300MHASH with overclocking. I tried to rewrite to uint4 but it went slower. What did you do? did you remove the lookuptables?

I have, but it was slower - that implementation should be AWESOME on FPGA, though.

AVX-512 CPUs are going to have software LUT3 on 32(!) 512-bit registers

[scryptr]

This is working on my 960 using my build(42) on windows 7

ccminer.exe -a whirlpoolx -u 1CTiNJyoUmbdMRACtteRWXhGqtSETYd6Vd -p 1 -o stratum+tcp://stratum.nicehash.com:3343

but you need to wait some seconds, and then Nicehash will set the diff correcly.

My 960 is also getting abit higher hashrate. I use -500Mzh on the memoryclockrate.

OC SETTINGS and HASHRATE--

I went to using a command line batchfile and your OC settings (+260mhz core, -500mhz memory) and my hashrate improved to 153,200kh/s.  However, I still get no accepted shares:

:loop

ccminer -a whirlpoolx -o stratum+tcp://stratum.nicehash.com:3343 -u bitcoinaddress.worker -p d=0.02

pause
goto loop

I used the above both with and without the difficulty parameter.  Soon as your new release is posted, I will try it.       --scryptr

[sp_]

I have to reach a flight, don't have time to merge.

In the mean time you can check out klaust's fork. version 2:

https://github.com/KlausT/ccminer

He has submitted some fixes on the whirlpoolx.

[go6ooo1212]

I there some other algo improvements on this Klaust's fork. I really cant find any sense at whirlpoolx..

[scryptr]

I have to reach a flight, don't have time to merge.

In the mean time you can check out klaust's fork. version 2:

https://github.com/KlausT/ccminer

He has submitted some fixes on the whirlpoolx.

STRATUM CONNECTION ERROR--

I am now able to run both your and KlausT's ccminer with whirlpoolx algo.  I get about 150mh/s with +225mhz core and -500mhz memory settings.  My card crashes with +250mhz, or +260mhz settings when actively linked on stratum.

There was something wrong with the stratum link, not the hashing.  I kept getting the red "retry in 30 seconds" message.       --scryptr

[flipclip]

... but I've encountered another issue.  When I hash with both GPUs (both 750Ti) one GPU gives me a crazy hashing number, for .

brother. i confirm. i have the same issue with every source out there (linux version!).

tpruvot ccminer works with multiple GPUs under Linux.

[chrysophylax]

... but I've encountered another issue.  When I hash with both GPUs (both 750Ti) one GPU gives me a crazy hashing number, for .

brother. i confirm. i have the same issue with every source out there (linux version!).

tpruvot ccminer works with multiple GPUs under Linux.

guys ... sp ccminer works perfectly with all workers - and all workers have a minimum of 5 cards in them ...

cards are gigabyte 750ti oc lp ...

if you guys are overclocking - then the clocks are too high for one of the cards to handle ...

#crysx

[flipclip]

... but I've encountered another issue.  When I hash with both GPUs (both 750Ti) one GPU gives me a crazy hashing number, for .

brother. i confirm. i have the same issue with every source out there (linux version!).

tpruvot ccminer works with multiple GPUs under Linux.

guys ... sp ccminer works perfectly with all workers - and all workers have a minimum of 5 cards in them ...

cards are gigabyte 750ti oc lp ...

if you guys are overclocking - then the clocks are too high for one of the cards to handle ...

#crysx

no overclock on any of my cards.  cards work fine individually, but not together.  I only have two cards.  can you run whirlpoolx using -d 0,1 (so you'd only run two cards), and does that work? thanks

[scryptr]

... but I've encountered another issue.  When I hash with both GPUs (both 750Ti) one GPU gives me a crazy hashing number, for .

brother. i confirm. i have the same issue with every source out there (linux version!).

tpruvot ccminer works with multiple GPUs under Linux.

guys ... sp ccminer works perfectly with all workers - and all workers have a minimum of 5 cards in them ...

cards are gigabyte 750ti oc lp ...

if you guys are overclocking - then the clocks are too high for one of the cards to handle ...

#crysx

no overclock on any of my cards.  cards work fine individually, but not together.  I only have two cards.  can you run whirlpoolx using -d 0,1 (so you'd only run two cards), and does that work? thanks

LINUX and SP_ v42--

SP_'s build errored out under Linux on my 6x750ti FTW rig.  I got the "out of memory/scanhash" error reported a page back.  No cards would hash.  Using tpruvot's v1.6 rc, it compiled with no errors.  It runs smoothly at 458MH/s on the rig, about 76MH/s per card, for whirlpoolx.  This is the same rate as SP_ and KlausT are getting.       --scryptr

[chrysophylax]

... but I've encountered another issue.  When I hash with both GPUs (both 750Ti) one GPU gives me a crazy hashing number, for .

brother. i confirm. i have the same issue with every source out there (linux version!).

tpruvot ccminer works with multiple GPUs under Linux.

guys ... sp ccminer works perfectly with all workers - and all workers have a minimum of 5 cards in them ...

cards are gigabyte 750ti oc lp ...

if you guys are overclocking - then the clocks are too high for one of the cards to handle ...

#crysx

no overclock on any of my cards.  cards work fine individually, but not together.  I only have two cards.  can you run whirlpoolx using -d 0,1 (so you'd only run two cards), and does that work? thanks

what version of ccminer-sp are you running? ...

ill test when I get into the office in a couple of hours ...

what is the commandline parameters you use so that I can duplicate your setup? ...

I can test it easily as I have a worker with only two cards in it also ...

#crysx

[flipclip]

... but I've encountered another issue.  When I hash with both GPUs (both 750Ti) one GPU gives me a crazy hashing number, for .

brother. i confirm. i have the same issue with every source out there (linux version!).

tpruvot ccminer works with multiple GPUs under Linux.

guys ... sp ccminer works perfectly with all workers - and all workers have a minimum of 5 cards in them ...

cards are gigabyte 750ti oc lp ...

if you guys are overclocking - then the clocks are too high for one of the cards to handle ...

#crysx

no overclock on any of my cards.  cards work fine individually, but not together.  I only have two cards.  can you run whirlpoolx using -d 0,1 (so you'd only run two cards), and does that work? thanks

what version of ccminer-sp are you running? ...

ill test when I get into the office in a couple of hours ...

what is the commandline parameters you use so that I can duplicate your setup? ...

I can test it easily as I have a worker with only two cards in it also ...

#crysx

I'm not using any parameters, just --algo, --url, --user, --pass (without any difficulty passing), and --api-bind.  No need to test though, I just cp tpruvot's x15/cuda_whirlpoolx.cu and x15/whirlpoolx.cu over sp-hash's x15/cuda_whirlpoolx.cu and x15/whirlpoolx.cu, compiled and works with my two 750Ti's in linux  .

[smolen]

That is really good. my 980 is around 300MHASH with overclocking. I tried to rewrite to uint4 but it went slower. What did you do? did you remove the lookuptables?

I have, but it was slower - that implementation should be AWESOME on FPGA, though.

Rough idea - use lookup tables to pick up only linear independent bits and regenerate the rest by boolean matrix multiplication.

[PVmining]

guys ... sp ccminer works perfectly with all workers - and all workers have a minimum of 5 cards in them ...
cards are gigabyte 750ti oc lp ...
if you guys are overclocking - then the clocks are too high for one of the cards to handle ...
#crysx

thanks - but we talk exclusive about the new -a whirlpoolX algo.
this is broken on linux for multi GPU systems. anything else works.

tpruvot ccminer works with multiple GPUs under Linux.

yep. tpruvot saw the problem quite early and fixed it yesterday.
I was already away from the computer - otherwise I would gave you a hint.

[chrysophylax]

guys ... sp ccminer works perfectly with all workers - and all workers have a minimum of 5 cards in them ...
cards are gigabyte 750ti oc lp ...
if you guys are overclocking - then the clocks are too high for one of the cards to handle ...
#crysx

thanks - but we talk exclusive about the new -a whirlpoolX algo.
this is broken on linux for multi GPU systems. anything else works.

tpruvot ccminer works with multiple GPUs under Linux.

yep. tpruvot saw the problem quite early and fixed it yesterday.
I was already away from the computer - otherwise I would gave you a hint.

o ok ... thats all good then ...

#crysx

[sp_]

Some nice work there done in the other forks. I Will be away for a while so will take some time before I test and
merge. I see some more possible speedups in whirlpoolx as well.