CCminer(SP-MOD) Modded GPU kernels.

[chrysophylax]

I am hoping that our developers begin merging the improved code into the open source code, so that my Linux rigs can also run at optimal speed.  Not everyone can afford to donate 0.1 BTC, or $40-50 US.

I submitted a small speedup in the quark algo and lyra2v2 to github. Have a nice weekend linux users

we are having a nice weekend ... as will those that have the funds to outrun ANY optimizations that you make via windows sp ...

when you provoke - we react ...

i can ( and WILL ) upgrade to the level that thefarm outdoes any of the optimizations that your private miner does - simply by adding ONE extra card ...

why am i saying this now? ...

i have been moving house for the last couple of days - and finishing tomorrow ... in two months - thefarm gets relocated as well as my next house move ... this time to a more permanent location with options to extend thefarm to more than three locations ... which means i can focus a lot more on thefarm - the code - the coins - the miners - and the infrastructure that binds these miners together ( im working on the setup and distribution of miners to the system and the ease at which they can be created and destroyed - going back to my 'grid' roots ) ...

the point im making here is a simple one ... for miners like me - who are 'starting' to become larger farms - and the large farms themselves - its not an issue to grow to 'keep up' or surpass those around us ... even at the cost of a ( so called ) 'loss' of a small optimization on the windows private miners ... i donate to you and other devs ANYWAY ... no matter how small or big the donation may be ...

but to taunt us into submission by dangling the proverbial carrot of 'optimization by windows' - is basically excluding us as well as our donations in future and more importantly - our interest ... something which leaves a foul distaste in my mouth - let alone others in the game ...

this also means that the smaller miner is disadvantaged a great deal - as they not only have to come up with the money for the computer AND the cards AND the os ( windows is FAR from free - you pay for it when you buy the machine ... and to those who are using a pirated version - tut tut ) - but also the miner itself ... setting them up for a HUGE outlay before a single portion of btc is mined ...

no matter what respect i / we have for you and other devs ( and i / we do have a lot of respect ) - this course of planning will lead to more and more internalization of funds AND lack of donations in a massive way ... not the OTHER way that we have all been searching for - which is a GOOD way of donation to devs and a sustainable one ...

i have tried running windows miners in fedora 23 x64 - and they mostly fail ... and if they do work - whatever optimization you make in the miner is drowned out by the conversion of coding wine has to do - and it does degrade the hashrate ...

i would mine for you for 3months at a time at a percentage of the total mining hash using the optimized miners - as long as i got to compile it from source on MY machines ... this would be a hell of a lot more than 0.1btc or 0.5btc on thefarm ( all nvidia ) than small donations ... but you would rather internalize your optimizations and grow your own farm ... well - good for you ... and its your right ...

but to taunt linux users? ...

guess who will come out on top when the optimizations are crushed by internal coding done by the 'big' farms ... the devs here? ...

i applaud your entrepreneurial direction ... but no ...

there needs to be a better way ... and we need to find it ...

#crysx

[1714762203]

1714762203

[1714762203]

1714762203

[1714762203]

1714762203

[1714762203]

1714762203

[djm34]

here an independent analysis (not by me) of all these new private releases: http://postimg.org/image/kt84x7jav/

(I let people draw their own conclusion, here mine: there is no difference between the 1st private release and the last  )

[sp_]

From release 2 to 3 quark was improved. The other improvements was on compute 5.2 devices. Release private 4 has been released with more quark and lyra2v2 on the 750ti.


On the asus strix 750ti standard clocks. (vs release 74)
lyra2v2: +6.26%
quark: +2.1%

compared to release 3
lyra2v2: +0,46%
 quark:  +0.3%

In sp private #4 the gigabyte windforce g1 970 is peaking at 10.7 MHASH on factory clocks. This is around what the original djm34 kernal did on the gtx 980. My code turns your gtx 970 into a gtx 980..

[djm34]

From release 2 to 3 quark was improved. The other improvements was on compute 5.2 devices. Release private 4 has been released with more quark and lyra2v2 on the 750ti.


On the asus strix 750ti standard clocks. (vs release 74)
lyra2v2: +6.26%
quark: +2.1%

compared to release 3
lyra2v2: +0,46%
 quark:  +0.3%

In sp private #4 the gigabyte windforce g1 970 is peaking at 10.7 MHASH on factory clocks. This is around what the original djm34 kernal did on the gtx 980. My code turns your gtx 970 into a gtx 980..

0.46% means nothing (except that people buying into that are bad in maths...)

actually you always claim increase by either comparing to your old stuff (or mine) while there is no difference between your various private release

[djm34]

by the way your standard public release triggers an alert from malwarebyte. Can you tell exactly what you did (rel 74 too) ?
I don't get that kind of shit with mine, are editing register or things like that for some reason ?

ps: I am pretty sure it is benign, but none of the other release does that

[s7icky]

Uhhh...

https://www.virustotal.com/en/file/f4df1e75023abe281787dfb71d8767169ea77b922c003f972e71bfdcd222d4c0/analysis/

[Al_may]

by the way your standard public release triggers an alert from malwarebyte. Can you tell exactly what you did (rel 74 too) ?
I don't get that kind of shit with mine, are editing register or things like that for some reason ?

ps: I am pretty sure it is benign, but none of the other release does that

yes back with ver 64 i had a major problem but in 65 it went away. it showed back up in .74 and is still a problem. the program is running bitcoinminer.exe in the back ground, as per Malwarebyte

[zTheWolfz]

kind of sounds like what I suspected since I seen a loss in coins rather than a gain, perhaps he has impaled it to mine part of our hash to his own wallet?

[djm34]

by the way your standard public release triggers an alert from malwarebyte. Can you tell exactly what you did (rel 74 too) ?
I don't get that kind of shit with mine, are editing register or things like that for some reason ?

ps: I am pretty sure it is benign, but none of the other release does that

yes back with ver 64 i had a major problem but in 65 it went away. it showed back up in .74 and is still a problem. the program is running bitcoinminer.exe in the back ground, as per Malwarebyte

I don't think it is really running a bitcoinminer, but I don't like to release something (since I have to use it for the neoscrypt release, and the exe I compile from the source gets flagged by malwarebyte...) which is flagged by virus scanners... so it would be good that sp_ correct that asap

[sp_]

False positives guys.. Compile for youself and check. Sgminer and ccminer are being used by botnets. Antivirus software doesn't like mining software.

[djm34]

False positives guys.. Compile for youself and check. Sgminer and ccminer are being used by botnets. Antivirus software doesn't like mining software.

yes I compile for myself your release, and get these alert.
Strangely when I compile my version, I don't get them, so there must be something doing it

so it is something in your code which does that

[sp_]

False positives guys.. Compile for youself and check. Sgminer and ccminer are being used by botnets. Antivirus software doesn't like mining software.

yes I compile for myself your release, and get these alert.
Strangely when I compile my version, I don't get them, so there must be something doing it
so it is something in your code which does that

Because the botnets doesn't use your version because it is slow. Antivirus software is scanning patterns in exe files right?..

[djm34]

False positives guys.. Compile for youself and check. Sgminer and ccminer are being used by botnets. Antivirus software doesn't like mining software.

yes I compile for myself your release, and get these alert.
Strangely when I compile my version, I don't get them, so there must be something doing it
so it is something in your code which does that

Because the botnets doesn't use your version because it is slow. Antivirus software is scanning patterns in exe files right?..

have no idea what it is scanning...

[sp_]

Download ccminer1.2 from github in google chrome. It will be blocked.
https://github.com/cbuchner1/ccminer/releases

Most likely because a trojan/virus has included this exefile in their archieve.
Ccminer is safe.

[djm34]

Download ccminer1.2 from github in google chrome. It will be blocked.
https://github.com/cbuchner1/ccminer/releases

Most likely because a trojan/virus has included this exefile in their archieve.
Ccminer is safe.

I don't use chrome it is a piece of crap which raise alert with malwarebyte all the time

[sp_]

From release 2 to 3 quark was improved. The other improvements was on compute 5.2 devices. Release private 4 has been released with more quark and lyra2v2 on the 750ti.


On the asus strix 750ti standard clocks. (vs release 74)
lyra2v2: +6.26%
quark: +2.1%

compared to release 3
lyra2v2: +0,46%
 quark:  +0.3%
In sp private #4 the gigabyte windforce g1 970 is peaking at 10.7 MHASH on factory clocks. This is around what the original djm34 kernal did on the gtx 980. My code turns your gtx 970 into a gtx 980..

0.46% means nothing (except that people buying into that are bad in maths...)
actually you always claim increase by either comparing to your old stuff (or mine) while there is no difference between your various private release

But 6.26% does. Actually if you have a gtx 960 the gain is more than 10%. The +0.46% is just a free upgrade. The donators are on my list and will recieve new versions when the opensource miner get faster.

[djm34]

From release 2 to 3 quark was improved. The other improvements was on compute 5.2 devices. Release private 4 has been released with more quark and lyra2v2 on the 750ti.


On the asus strix 750ti standard clocks. (vs release 74)
lyra2v2: +6.26%
quark: +2.1%

compared to release 3
lyra2v2: +0,46%
 quark:  +0.3%
In sp private #4 the gigabyte windforce g1 970 is peaking at 10.7 MHASH on factory clocks. This is around what the original djm34 kernal did on the gtx 980. My code turns your gtx 970 into a gtx 980..

0.46% means nothing (except that people buying into that are bad in maths...)
actually you always claim increase by either comparing to your old stuff (or mine) while there is no difference between your various private release

But 6.26% does. Actually if you have a gtx 960 the gain is more than 10%. The +0.46% is just a free upgrade. The donators are on my list and will recieve new versions when the opensource miner get faster.

so basically you expect other dev to work on the opensource so to make your private miner faster
pallas says thank you to you 

[sp_]

My git is still active. Just submitted a 0.2% increase in lyra2v2.

[Eliovp]

These endless discussions. Got to give it to you guys. I would have given up a long time ago 

[djm34]

My git is still active. Just submitted a 0.2% increase in lyra2v2.

yeah I saw you commited something you removed... still the same activity as usual... half of the change which happens are just to correct some shit you put it in at the previous round...  

It isn't release 78, actually it is more likely 39 (in terms of real progress). 
Seriously can't you just check what you are doing before committing anything