Topic: Signatum (skunk) ccminer-krnlx 30+mh 1070, 52.5+ mh 1080ti [FREE,OPENSOURCE] - page 3. (Read 39469 times)

I actually use krnlx, but i think u gone swap to epsylon ccminer build. He got improvement on 13 august on skunk, based on krnlx withtout the share problem

https://bitcointalksearch.org/topic/ann-ccminer-23-opensource-gpl-tpruvot-770064

Link to source code :

https://github.com/tpruvot/ccminer

But u maybe want to go to another crypto actually, sigt is less rentable actually, lot of people sell it.

wow I just bought some 1070s to replace my r9 cards there killing me on power compared to my rx cards but wanted to mine signatum now I read the miner issues what miner is recommend I  never owned a nivida  card in my life so the help is appreciated thanks

Interesting observation. Thank you for the information. I am going to do some investigation myself.

now it is not profit, I use 6 pieces 470 rx series vga

You can all try it for yourselves. Test with palgin and krnlx and spmod3/5 miners. I tested these with suprnova. Since the hash rate in the dashboard being shown by suprnova is continuously changing, I decided to test the miners with the SIGT coins I'm mining.

Yes, it was not a scientific test, I just wanted to have a rough estimate because I noticed the coins I'm mining suddenly dropped when I used krnlx. I wasn't sure at first if it was because of the difficulty in SIGT and the many miners in suprnova or it was due to krnlx. So I created 2 accounts in suprnova. I have  2 almost identical rigs, 6 x GTX1060 on rig1 and 5 x GTX 1060 on rig2. The first test was running spmod3 on rig2 and krnlx on rig1. I didn't configure the rigs to be of the exact number of gpus because I will lose mining time for a couple of hours if disable one gpu. After more than an hour, I had almost the same number of coins mined. It meant krnlx was mining slower than spmod3 because it was using 6 gpus. The next test was running palginmod on rig 1 and spmod3 on rig2. This time, palginmod mined more coins, which should be the case since it is running on 6 gpus. If I divide the number of coins per gpu, the average coin per gpu was the same for palginmod and spmod3, which meant in my crude test that both miners seem to be performing equally.

There seems to be a problem with krnlx even it is showing more hash rate per second. Those of you who have a lot more resources and time can test krnlx more extensively.  For now, I'm going back to palginmod or spmod. Check also a test run by jimboscott here https://bitcointalk.org/index.php?topic=2030529.4700   He has the same observations.

weird though. the latest ccminer from tpruvot also use this code. giving the same performance.

https://github.com/tpruvot/ccminer/releases

I did see a considerable jump in my reported hashrate with this miner but my average hashrate at the pool dropped 20%

Which pool would you recommend for low diff

super parasite   supernova high diff jumps is not new and is not a miner issue...

 I spend my free time to test this miner, and finds out the claimed speed in the miner window is not showing on the pool. This has been confirmed by many users on this thread already. The only way to fix this is to fix the code.

Stop talking shit. You are forcing people to buy your miner with worsening other miners.
You greedy son of a bitch.

Wath pool do u use for static diff ?
Can u explain where u grab the number of share u validate ? I dont found it on suprnova / pool.mn

Mining on lower diff increase the hashrate on the pool, but the miner is still broken.
Some solutions are skipped, so you loose a few percent. And suprnova and many other pools doesn't support static diff, because static diff can be abused and used to steal from the other miners..

so you're saying to fix this mine at a lower difficulty? got it....

send me spmod 6 and I will run the test again. It can be that there is an issue on variable diff. I didn't test that yet as I set it to static 1.0 diff.

This seems to be correct, from my experience.

Sp-mod #6 is 9% faster than sp-mod #3 on the gtx 1070 if you use the latest driver. (384.94) and 100% tdp. (Donate 0.05B BTC to subscribe to my private miner)
The krnlx kernel is broken if you mine on high difficulty (vardiff).  Try supernova with highend cards rigs.
You only get payed for 80% of the  hash showed in the miner.

btw mate ...

we will be putting of CWIgm-0.9.9 some time the end of the week or early next week by the looks of it ...

IF we can get a number of issues fixed in the miner ...

CWIgm-0.9.9 will not only have skunkhash and tribus in it ( like CWIgm-0.9.8 does ) - but we will also be adding timetravel10 to it ( bitcore - btx ) ... i know that a little off topic regarding other coins - but stability and hashrate will be a better in the next version ...

either way - we would be interested in these results also ...

#crysx

no mate ...

not gospel ... but we have been around for years - and have delved into the nether regions of code and core ... so we 'sorta' know what we are talking about ...

as for a newbie account - i am in total agreement ... the newbie level is so restrictive that you cant do much - or post much - or be much on here ... just keep at posting and doing what you do - it will gain momentum - dont worry ... before you know it - the account is a junior memebr - and even tho thats restrictive - its nothing like the newbie level ...

anyway - need to get back to work ...

got another holy book to write ...

#crysx

Thank you blissz.
Could you please run test till 24 hours so we have valid results.
Unfortunately I dont have access to sp mod 5 (the one I had reported lower hashrate compared to mod 3 so I didnt use it)
 
I would like the following combos tested if possible:
a) sp mod 3 vs. cwi 0.9.8 ( have this file and can forward to you)
b) sp mod 3 vs. krnlx (CUDA 7.5, X86)
c) cwi 0.9.8 vs. krnlx (CUDA 7.5, X86)

Also any way you could track coins generated instead of share rate please, with data labels so we can calculate % difference b/w miners.

Might need to edit the test length for the cwi miner since miner pool and diff can vary. But ultimately need to find out which one generates more coin.  
Thanks again for posting the results.