CCminer(SP-MOD) Modded GPU kernels.

[s7icky]

So... SP_ please explain how this works? Or can you just come out and admit you are screwing people? this is your decred #4

Ofcouse it works

Here is a snapshot of #6 beta:

Here is why it works.  (I have removed the intensity parameter.)

Code:

	int dev_id = device_map[thr_id];
	//int intensity = 29;
	//	if (device_sm[dev_id] < 350) intensity = 22;
//		if (device_sm[dev_id] > 500)
//		{
//			intensity = 30;
//		}
	//
	//uint32_t throughput = cuda_default_throughput(thr_id, 1U << intensity);
	
	uint32_t throughput = (0xffffffff-5) / 5;
	if (device_sm[dev_id] > 500)
	{
		throughput = (0xffffffff - 6) / 2;
	}

k dude im not a coder... thus why im ASKING you why its showing up different on the pools... and sure you can copy and paste/edit code.... cool for you WHY IS THERE SUCH A HUGE DIFFERENCE ON ANY POOL??And it looks like i am not the only one seeing this... so like wtf?

[sp_]

k dude im not a coder... thus why im ASKING you why its showing up different on the pools... and sure you can copy and paste/edit code.... cool for you WHY IS THERE SUCH A HUGE DIFFERENCE ON ANY POOL??

Because #4 contain a bug. Still 20% faster than then opensource if you mine @yiimp or solomine

[s7icky]

do you want to blame it on my 5960x @ 4.4ghz? or 3930k @ 4.9ghz?



sorry but that is not any faster that is actually slower....

[sp_]

I can teach you why the kernal is displaying low numbers for low difficulty shares. (and for single cards with a low hashrate)


Take a look at at
https://github.com/tpruvot/ccminer/blob/windows/Algo256/decred.cu

and line 259

      if (h[7] == 0 && cuda_swab32(h[6]) <= highTarget) {

We don't care about the 4 first bytes of the hash and assume they are 0. This is to spead up the search.

      
Here is a found decred block:(hash)      
      
0000000000000d1ec551b6e0b734dbfa31e95059aa84e50f7db2ab244499c87c

0000000 000000d1ec551b6e0b734dbfa31e95059aa84e50f7db2ab244499c87c


but when a pool asks for  a low difficulty share
f.eks find a number that is less than this:
00000ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff ffffffffffffffffffffffffffff
We are not submitting the shares that have anything else than 00 in the low byte of the first 32 bits. (why should we when we know that the solution doesn't give a block)

If I mod the kernal to search the top 32bits only (h[7]) <= Target the mining software will work exellent at low difficulty, but not so good on higher diff.. We are searching for blocks with 14 leading zero's So I believe the current kernal is good, and the pools are fucked..

Why should we when we know it doesn't give a block? BECAUSE POOLS WORK THAT WAY, DUDE. Not submitting those is literally burning cash. The problem isn't the pools, it's you.

He have fixed it in the opencl version..

With a nice comment

   /* Debug code to help you assess the correctness
    * of your hashing function in case someone decides
    * to try to optimize.
   if (!((pre7 ^ V7 ^ VF) & 0xFFFF0000)) {


https://github.com/tpruvot/sgminer/commit/d78cfbbc77df7a0d2b8c87168a4750e4eb8830a4

[s7icky]

do you want to blame it on my 5960x @ 4.4ghz? or 3930k @ 4.9ghz?



sorry but that is not any faster that is actually slower....

good way to ignore it dude.

[sp_]

good way to ignore it dude.

Change pool and use the correct intensity.

yiimp.ccminer.org

If you have a 980ti card with release #4 use -i 31.9 as a parameter for the intensity. Mine for 24Hours. A Single card. The gtx 970 works good with -i 30. the 750ti with -i 29.6.

[s7icky]

good way to ignore it dude.

Change pool and use the correct intensity.

yiimp.ccminer.org

If you have a 980ti card with release #4 use -i 31.9 as a parameter for the intensity. Mine for 24Hours. A Single card. The gtx 970 works good with -i 30. the 750ti with -i 29.6.

sorry i already did that no change... any other reasons it would be showing 3ghs difference with 2 x 980ti? or 500ish mhs on 960's?




i have done tests on 3 pools.


ccminer.exe -d 0,1 -i 24,29.6 -a decred -o http://dcr.suprnova.cc:9112 -u s7icky.1 -p x
pause

^^^ 980ti x2

ive been running that soooooooo............................... i kinda figured you would say that.and the other setup i have is 960 x 2 and 1 x 750ti...

ccminer.exe -i 29.6 -d 0 -a decred -o http://dcr.suprnova.cc:9111 -u s7icky.2 -p x
pause
ccminer.exe -i 29.6 -d 1 -a decred -o http://dcr.suprnova.cc:9111 -u s7icky.2 -p x
pause
ccminer.exe -i 29.6 -d 2 -a decred -o http://dcr.suprnova.cc:9111 -u s7icky.2 -p x
pause


so can you explain the 3ghs? :/ cause i am at a dead end bro.

[antantti]

i have done tests on 3 pools.


ccminer.exe -d 0,1 -i 24,29.6 -a decred -o http://dcr.suprnova.cc:9112 -u s7icky.1 -p x
pause

^^^ 980ti x2

ive been running that soooooooo............................... i kinda figured you would say that.and the other setup i have is 960 x 2 and 1 x 750ti...

ccminer.exe -i 29.6 -d 0 -a decred -o http://dcr.suprnova.cc:9111 -u s7icky.2 -p x
pause
ccminer.exe -i 29.6 -d 1 -a decred -o http://dcr.suprnova.cc:9111 -u s7icky.2 -p x
pause
ccminer.exe -i 29.6 -d 2 -a decred -o http://dcr.suprnova.cc:9111 -u s7icky.2 -p x
pause


so can you explain the 3ghs? :/ cause i am at a dead end bro.

Have you tried port 2252?

[bensam1231]

k dude im not a coder... thus why im ASKING you why its showing up different on the pools... and sure you can copy and paste/edit code.... cool for you WHY IS THERE SUCH A HUGE DIFFERENCE ON ANY POOL??

Because #4 contain a bug. Still 20% faster than then opensource if you mine @yiimp or solomine

So this is fixed in #5?

Care to post a screenshot of the miner next to pool dashboard on mnool, coinmine, or suprnova?

[sp_]

So this is fixed in #5?

No.

[sp_]

k dude im not a coder... thus why im ASKING you why its showing up different on the pools... and sure you can copy and paste/edit code.... cool for you WHY IS THERE SUCH A HUGE DIFFERENCE ON ANY POOL??

Because #4 contain a bug. Still 20% faster than then opensource if you mine @yiimp or solomine

So this is fixed in #5?
Care to post a screenshot of the miner next to pool dashboard on mnool, coinmine, or suprnova?

I am working on it #6, but my gtx 960 is crashing. I need mod the code to case on the cards..

[Zekku]

Just bought it and there's a bug.

Will test for 24h on each and compare.

2x GTX 970 EVGA SSC+

Edit: stock clocks

[s7icky]

k dude im not a coder... thus why im ASKING you why its showing up different on the pools... and sure you can copy and paste/edit code.... cool for you WHY IS THERE SUCH A HUGE DIFFERENCE ON ANY POOL??

Because #4 contain a bug. Still 20% faster than then opensource if you mine @yiimp or solomine

So this is fixed in #5?

Care to post a screenshot of the miner next to pool dashboard on mnool, coinmine, or suprnova?

just admit ur bullshitting people doood

[s7icky]

i have done tests on 3 pools.


ccminer.exe -d 0,1 -i 24,29.6 -a decred -o http://dcr.suprnova.cc:9112 -u s7icky.1 -p x
pause

^^^ 980ti x2

ive been running that soooooooo............................... i kinda figured you would say that.and the other setup i have is 960 x 2 and 1 x 750ti...

ccminer.exe -i 29.6 -d 0 -a decred -o http://dcr.suprnova.cc:9111 -u s7icky.2 -p x
pause
ccminer.exe -i 29.6 -d 1 -a decred -o http://dcr.suprnova.cc:9111 -u s7icky.2 -p x
pause
ccminer.exe -i 29.6 -d 2 -a decred -o http://dcr.suprnova.cc:9111 -u s7icky.2 -p x
pause


so can you explain the 3ghs? :/ cause i am at a dead end bro.

Have you tried port 2252?

yessum

[sp_]

Decred beta #6 is getting 100% more hash @ yiimp.com than zpool.

[antantti]

I don't know what am I doing wrong but it just works for me. Testing mnpool atm, this is the best so far. Windows 7 x64, 4x970, -i 28.9, latest dcrmod.

Pool keeps difficulty at 1, miner showing 7.566GH, pool 7.0-8.0GH.

CPU (4690K) usage 0%.

[sp_]

I don't know what am I doing wrong but it just works for me. Testing mnpool atm, this is the best so far. Windows 7 x64, 4x970, latest dcrmod.
Pool keeps difficulty at 1, miner showing 7.566GH, pool 7.4-7.6GH.

My windows 7 rigs are also doing good. Some of my windows 8.1 celeron rigs are using 25-50% cpu but less on higher difficulties.

[tbearhere]

Good sample GPU, overclock CPU +230Mhz , Full Tower Desktop With 3+ fans 200mm, 1x140 mm 1x120mm.
My sample 960 overclocked GPU to 1600Mhz - cool and stable

How much hash rate increase do you think you got by oc'ing the cpu?  Thx

[sp_]

I have added 50khash in the quark algo on the 750ti.

Also more boost in x11,x13,x14,x15,x17,nist5

Blake-512-80 20% faster. (x11 0.5% faster)

https://github.com/sp-hash/ccminer/commit/e195e988187999b7e035e18cd61ce4de9356f7cd

Supported by decred/vcash donations

[sp_]

Decred beta #6 is getting 100% more hash @ yiimp.com than zpool.

Broken... More work needed. still 20% faster, but not 100% faster.