[ANN] sgminer v5 - optimized X11/X13/NeoScrypt/Lyra2RE/etc. kernel-switch miner

[rednoW]

Fuck that, I think that's mostly RE'd from Claymore's. Qubit and Quark are mine.

So will we see the source code soon?

[jch9678]

And I reached 11.39MH/s on 280X with my third pass working on it, temps slightly higher - 69C on 60% fan now (NSFW): https://ottrbutt.com/miner/quarkwolf-04202015.png

how you do it ? wich settings you use ? stock kernel or you write your own one ?

Wrote my own, of course.

if quoted speed is still true then i doubt it's your bin,
or you chose wrong settings which i cannot imagine as well
maybe 4th pass ?  

Hm, maybe it's not, then. What's their speed?

Odd that less than a week after someone gets my version, this happens...

280x
qubit - 12.5 Mh / S
Quark - 12.3 Mh / S

Doesn't work with standard sgminer however

[smolen]

Dammit, just realized - if he's distributing it, some jackass like Luke-Jr is gonna find out and demand the source. The host code will likely be enough to give away a major trick to getting Quark faster... shit. The OpenCL should be safe, it's input, not a derived work.

Hmm...

Code:

$ strings ./quarkcoinTahitigw64l4ku0.bin |less |grep -i opencl

Code:

__OpenCL_compile_options
__OpenCL_0_global
__OpenCL_search_metadata
__OpenCL_search_kernel
__OpenCL_search_header
__OpenCL_search1_metadata
__OpenCL_search1_kernel
__OpenCL_search1_header
__OpenCL_search2_metadata
__OpenCL_search2_kernel
__OpenCL_search2_header
__OpenCL_search3_metadata
__OpenCL_search3_kernel
__OpenCL_search3_header
__OpenCL_search4_metadata
__OpenCL_search4_kernel
__OpenCL_search4_header
__OpenCL_search5_metadata
__OpenCL_search5_kernel
__OpenCL_search5_header
__OpenCL_search6_metadata
__OpenCL_search6_kernel
__OpenCL_search6_header
__OpenCL_search7_metadata
__OpenCL_search7_kernel
__OpenCL_search7_header
__OpenCL_search8_metadata
__OpenCL_search8_kernel
__OpenCL_search8_header
-I "." -I "./kernel" -I "." -D WORKSIZE=64 -D BITALIGN -I "/usr/local/bin" -D SPH_COMPACT_BLAKE_64=0 -D SPH_LUFFA_PARALLEL=0 -D SPH_KECCAK_UNROLL=0  -save-temps@(#) OpenCL 1.2 AMD-APP (1642.5).  Driver version: 1642.5 (VM)
;ARGSTART:__OpenCL_search_kernel
;ARGEND:__OpenCL_search_kernel
;ARGSTART:__OpenCL_search1_kernel
;ARGEND:__OpenCL_search1_kernel
;ARGSTART:__OpenCL_search2_kernel
;ARGEND:__OpenCL_search2_kernel
;ARGSTART:__OpenCL_search3_kernel
;ARGEND:__OpenCL_search3_kernel
;ARGSTART:__OpenCL_search4_kernel
;ARGEND:__OpenCL_search4_kernel
;ARGSTART:__OpenCL_search5_kernel
;ARGEND:__OpenCL_search5_kernel
;ARGSTART:__OpenCL_search6_kernel
;ARGEND:__OpenCL_search6_kernel
;ARGSTART:__OpenCL_search7_kernel
;ARGEND:__OpenCL_search7_kernel
;ARGSTART:__OpenCL_search8_kernel
;ARGEND:__OpenCL_search8_kernel

[smolen]

Nope, officially not my shit, or modified. Mine has 12 kernels, as-is, but SHOULD be 9 or so after cleanups.

Interesting, is GPL applicable here? SPH_LUFFA_PARALLEL hints that this kernel is derived work.

[chup]

Nope, officially not my shit, or modified. Mine has 12 kernels, as-is, but SHOULD be 9 or so after cleanups.

Interesting, is GPL applicable here? SPH_LUFFA_PARALLEL hints that this kernel is derived work.

It is not, since his (maybe derivative) work is not published. He can sell, distribute, but not publish derivative works to comply to GPL...

[chup]

Interesting, is GPL applicable here? SPH_LUFFA_PARALLEL hints that this kernel is derived work.

It is not, since his (maybe derivative) work is not published. He can sell, distribute, but not publish derivative works to comply to GPL...

If he distributes it to someone who wants the source, then he has to give it - assuming he has used ANY GPL'd code in his kernel, the entire kernel has now been infected.

No, 100% positive that obligation for announcing the source code comes only if one publishes derivative work. Distribution comes at ask of end user, so no obligations afterwards even if the end user was paying for it. At this case, he is distributing not only kernels, but also variant of sgminer and that's the reason of not publishing the work, but distributing. He knows he is on a safe side.

[pallas]

Hmm...

Code:

$ strings ./quarkcoinTahitigw64l4ku0.bin |less |grep -i opencl

readelf -aW kinda nicer.

[pallas]

I've released my lyra2re miner sources, see my signature.

[MaxDZ8]

Interesting, is GPL applicable here? SPH_LUFFA_PARALLEL hints that this kernel is derived work.

It is not, since his (maybe derivative) work is not published. He can sell, distribute, but not publish derivative works to comply to GPL...

If he distributes it to someone who wants the source, then he has to give it - assuming he has used ANY GPL'd code in his kernel, the entire kernel has now been infected.

No, 100% positive that obligation for announcing the source code comes only if one publishes derivative work. Distribution comes at ask of end user, so no obligations afterwards even if the end user was paying for it. At this case, he is distributing not only kernels, but also variant of sgminer and that's the reason of not publishing the work, but distributing. He knows he is on a safe side.

I read this a few times and I cannot make any sense of it.

[chup]

No, 100% positive that obligation for announcing the source code comes only if one publishes derivative work. Distribution comes at ask of end user, so no obligations afterwards even if the end user was paying for it. At this case, he is distributing not only kernels, but also variant of sgminer and that's the reason of not publishing the work, but distributing. He knows he is on a safe side.

I read this a few times and I cannot make any sense of it.

Maybe I used wrong term "distribute", meaning direct (peer to peer; private channel) transfer. There is no GPL rule that can prohibit something that can not be prohibited. If the end user is willed to use improved binaries without source code, there is no power prohibiting that.
In fact, I believe that GPL is not meant to prohibit anything, but to make better environment and conditions for free software. You are free to use GPL licensed software for You in any way You like. You can change it before. And someone can help You changing it. Here is part of section 2 (Basic permissions) of GPL terms and conditions:

"You may convey covered works to others for the sole purpose of having them make modifications exclusively for you, or provide you with facilities for running those works, provided that you comply with the terms of this License in conveying all material for which you do not control copyright. Those thus making or running the covered works for you must do so exclusively on your behalf, under your direction and control, on terms that prohibit them from making any copies of your copyrighted material outside their relationship with you."

I said nothing about persons leaking after. If they want to publish GPL licensed derivative work, they have to find source code before. If "the leakage" is through private channels,... there is no (legal) power to stop that also.

[MaxDZ8]

So basically my understanding is that you have no clue what you're talking about and you're all in for propaganda. Nobody in business considers GPL a decent software license nowadays but the most extremist open source dudes. If you want free software, free is free. Free is MIT or ZLIB.

[splootch]

You wanna sell it just put the link where we can buy or ask, you wanna give it just give it but damn keep the ally gpu miners up to date if you wanna beat these asic whale...we are very little and very few. if we not going to make the money, asic will. No point to remain silence fearing you gonna loose 0.50$ by a rising gpu harsh rate from new miner released .

[chup]

So basically my understanding is that you have no clue what you're talking about and you're all in for propaganda. Nobody in business considers GPL a decent software license nowadays but the most extremist open source dudes. If you want free software, free is free. Free is MIT or ZLIB.

I have a clue that we are discussing here variant of sgminer and other GPL-licensed code. Suddenly someone that "has clue" jumps with "better" acronyms. You should go try convince Jeff Garzik that he has no clue choosing GPL-license for his work. If You can do that I will promote You in MASTER OF PROPAGANDA.

[splootch]

The more power gpu miners community has the less profitable asic is and the higher those asic whales has to hold for the btc price to go up. RELEASE THE THING.

[sammy007]

I noticed in this thread some links to optimized miner by author of "kachur" CryptoNight miner. This kachur shit was a malware/scamware miner exploiting plain stupid duplicate share vulnerability in node-cryptonote-pool. You warned, stay away from all blobs by this "dev".

https://bitcointalk.org/index.php?topic=583449.msg9262887#msg9262887

[Oscilson]

I noticed in this thread some links to optimized miner by author of "kachur" CryptoNight miner. This kachur shit was a malware/scamware miner exploiting plain stupid duplicate share vulnerability in node-cryptonote-pool. You warned, stay away from all blobs by this "dev".

https://bitcointalk.org/index.php?topic=583449.msg9262887#msg9262887

Do you also have a proof that this new miner submits duplicate shares?

[BitmoreCoin]

I noticed in this thread some links to optimized miner by author of "kachur" CryptoNight miner. This kachur shit was a malware/scamware miner exploiting plain stupid duplicate share vulnerability in node-cryptonote-pool. You warned, stay away from all blobs by this "dev".

https://bitcointalk.org/index.php?topic=583449.msg9262887#msg9262887

Do you also have a proof that this new miner submits duplicate shares?

of course it does, besides it switches off the rig power supply
to use preinstalled liquid antimatter source, no heat at all, just extreme radiation

Too much radiation is bad. But if it can use radiation to transmit coins, then energy is saved.

[Bananana]

What is with the quark mining so profitable all of sudden? Also anybody can share wolf0's program please .

[bensam1231]

Why do you think only wolf can create optimized binaries?
I think kachur can create it as kminer for monero. Isn't it possible?

Just want to point out Kachur made a very competitive Cryptonote miner with Claymore. He eventually stopped updating it and got banned for some reason. There was also some schenanagins about duplicated shares for a bit too.

[semajjames]

What is with the quark mining so profitable all of sudden? Also anybody can share wolf0's program please .

as far as I know Wolf0 has not released any bin's for Quark and has not released a miner