[ANN][X11/X13] X11 (Darkcoin)/X13 (Marucoin) miner (based on sph-sgminer)

[djm34]

as french, I like revolutions so whose head shall we cut ?

x11 is like a banana republic, there is a revolution every now and then.
Really want to revolutionize things ? rewrite from scratch sgminer.

an advice: stop mistaking revolution and pissing contest...

God, SGMiner DOES need rewriting...

What will be the advantages to re-writing SGMiner?

You've obviously never seen the code. One thing that holds back development is that no one wants to touch it. CGMiner was bad enough, and that is horrid - now we have more misfit pieces of code grafted onto it. It works, but that's about the only good thing you can say about that code.

Everybody use it because, as far as I know, it's the best at handling OpenCL (so AMD cards) ;-)

No, everybody uses it because there's no reasonable alternative. It's not the best at OpenCL, it's the ONLY one - otherwise you must write it from scratch.

actually coin shield miner is better, but doesn't support standard protocol... (and use a bit too much cpu...)
But at least you have some freedom to add (or remove) variables for your kernel.

[pallas]

Maybe, but other opencl applications I've seen use half to a full core for every GPU, while cgminer can run six with the slowest single core cpu.
The problem is all the patches and forks it had.

[Ignition75]

So no speed improvements for the end user, just making life easier and cleaner for the devs.

[djm34]

So no speed improvements for the end user, just making life easier and cleaner for the devs.

making life easier for devs, means also improvement on speed because it is easier to implement.
My preferred example: xcn, nobody released anything (except claymore with a fee) because just thinking on how to do it in sgminer gives you nightmare. 

[Ignition75]

So no speed improvements for the end user, just making life easier and cleaner for the devs.

making life easier for devs, means also improvement on speed because it is easier to implement.
My preferred example: xcn, nobody released anything (except claymore with a fee) because just thinking on how to do it in sgminer gives you nightmare. 

You think there's a crowd funding opportunity here to completely re-write the mining software?

[JuanHungLo]

So no speed improvements for the end user, just making life easier and cleaner for the devs.

making life easier for devs, means also improvement on speed because it is easier to implement.
My preferred example: xcn, nobody released anything (except claymore with a fee) because just thinking on how to do it in sgminer gives you nightmare. 

You think there's a crowd funding opportunity here to completely re-write the mining software?

With "no speed improvements for the end user"  I doubt it.

[djm34]

So no speed improvements for the end user, just making life easier and cleaner for the devs.

making life easier for devs, means also improvement on speed because it is easier to implement.
My preferred example: xcn, nobody released anything (except claymore with a fee) because just thinking on how to do it in sgminer gives you nightmare. 

You think there's a crowd funding opportunity here to completely re-write the mining software?

With "no speed improvements for the end user"  I doubt it.

may be we should just release opencl file and ask the end user to make it to run... 

[Ignition75]

So no speed improvements for the end user, just making life easier and cleaner for the devs.

making life easier for devs, means also improvement on speed because it is easier to implement.
My preferred example: xcn, nobody released anything (except claymore with a fee) because just thinking on how to do it in sgminer gives you nightmare. 

You think there's a crowd funding opportunity here to completely re-write the mining software?

With "no speed improvements for the end user"  I doubt it.

may be we should just release opencl file and ask the end user to make it to run... 

How many man days are we talking here to re-write it from the ground up?

[djm34]

So no speed improvements for the end user, just making life easier and cleaner for the devs.

making life easier for devs, means also improvement on speed because it is easier to implement.
My preferred example: xcn, nobody released anything (except claymore with a fee) because just thinking on how to do it in sgminer gives you nightmare.  

You think there's a crowd funding opportunity here to completely re-write the mining software?

With "no speed improvements for the end user"  I doubt it.

may be we should just release opencl file and ask the end user to make it to run...  

How many man days are we talking here to re-write it from the ground up?

quite a lot, difficult to say would require a real planing and a real design thought from the beginning to incorporate various any type of algo (as well as API support and so on and the eventual dev fee everybody ask so they can remove it  )

and actually it is the problem usually you are in a hurry to program a new algo and you won't take the time to rewrite things...

[Ignition75]

So no speed improvements for the end user, just making life easier and cleaner for the devs.

making life easier for devs, means also improvement on speed because it is easier to implement.
My preferred example: xcn, nobody released anything (except claymore with a fee) because just thinking on how to do it in sgminer gives you nightmare.  

You think there's a crowd funding opportunity here to completely re-write the mining software?

With "no speed improvements for the end user"  I doubt it.

may be we should just release opencl file and ask the end user to make it to run...  

How many man days are we talking here to re-write it from the ground up?

quite a lot, difficult to say would require a real planing and a real design thought from the beginning to incorporate various any type of algo (as well as API support and so on and the eventual dev fee everybody ask so they can remove it  )

and actually it is the problem usually you are in a hurry to program a new algo and you won't take the time to rewrite things...

What would you write it in?  C++, Python?

[TorinT]

sgminer does not need rewrite. there still is much work with kernels

[djm34]

sgminer does not need rewrite. there still is much work with kernels

thanks for your opinion, great leader of the revolution 
(may-be once you will be done fiddling with x11 and code a new algo, you'll change your opinion)

[lokojones]

Hi there

2 x r9 280x on win 8.1 14.4 catalyst and sgminer 4.1 - everything working smooth and nice - 3.3Mh
but ver 5.0.1 do not start, is crushing instantly?!downloaded from http://cryptomining-blog.com/3921-new-windows-binary-of-sgminer-5-0-1-master-branch/
but when tested on other pc with win 7 same drivers, works fine?! any clue, pls?

Another noob question, original realese from https://github.com/sgminer-dev/sgminer/releases
dont contain .exe file - what I missing here? Could someone point me to working version ( i mean complete with .exe) 5.1 please - many thanks

[lokojones]

Hi there

2 x r9 280x on win 8.1 14.4 catalyst and sgminer 4.1 - everything working smooth and nice - 3.3Mh
but ver 5.0.1 do not start, is crushing instantly?!downloaded from http://cryptomining-blog.com/3921-new-windows-binary-of-sgminer-5-0-1-master-branch/
but when tested on other pc with win 7 same drivers, works fine?! any clue, pls?

Another noob question, original realese from https://github.com/sgminer-dev/sgminer/releases
dont contain .exe file - what I missing here? Could someone point me to working version ( i mean complete with .exe) 5.1 please - many thanks

No one want to contribute... come on guys... btw I still beleive in humanity:)

[kenshirothefist]

Hi there

2 x r9 280x on win 8.1 14.4 catalyst and sgminer 4.1 - everything working smooth and nice - 3.3Mh
but ver 5.0.1 do not start, is crushing instantly?!downloaded from http://cryptomining-blog.com/3921-new-windows-binary-of-sgminer-5-0-1-master-branch/
but when tested on other pc with win 7 same drivers, works fine?! any clue, pls?

Another noob question, original realese from https://github.com/sgminer-dev/sgminer/releases
dont contain .exe file - what I missing here? Could someone point me to working version ( i mean complete with .exe) 5.1 please - many thanks

Take a look here for sgminer downloads: https://bitcointalk.org/index.php?topic=632503.0

[kenshirothefist]

Guys, why are you keeping this thread alive? All the excellent work that has been done by the OP author - lasybear and other developers has already been included in the "official" sgminer release (it's a community work, therefore I guess it can be called in any way, if you don't like "official" then it can be the "sgminer-dev" version). Information gets lost if it is kept in several separate threads, therefore I welcome you to discuss topics regarding sgminer and related stuff in this thread: https://bitcointalk.org/index.php?topic=632503.0

If you agree, we could consider this thread closed and continue the debate here: https://bitcointalk.org/index.php?topic=632503.new#new

[Blawpaw]

Has anyone succeded in getting 2 GPUs -AMD 6950 \ 7970 working with X13 algo?
If so please show me your config file!

I've tried the kernel x13modold and it doesnt work...

any advice?

[restless]

latest x13 optimisations are not compatible with 6xxx and 5xxx radeons
Best way is to run with -d switch - pointing to your 7970 and another instance using x13modold/marucoin-modold , again with -d but pointing to 6950 card
Best speed achieved by 6970 is ~ 1.4MH/s for x13

[K1773R]

Has anyone succeded in getting 2 GPUs -AMD 6950 \ 7970 working with X13 algo?
If so please show me your config file!

I've tried the kernel x13modold and it doesnt work...

any advice?

it does, even with 5??? cards
tell us what isnt working. provide more details.

[thevictimofuktyranny]

More uptodate info on mods?

On these threads:
https://bitcointalk.org/index.php?topic=854257.0

and

https://bitcointalk.org/index.php?topic=632503.0