CCminer(SP-MOD) Modded GPU kernels.

[bensam1231]

hello everyone ,

nvidia volta appeared , and maybe we`ll see the new cards by christmas

i`m courious to see first mining rigs build on new arhitecture (nvlink vs pcie ) any thoughts ?

performance of v100 seems to be 4x gtx 1060 ...and that is on one single card

http://www.pcworld.com/article/3196026/components-graphics/nvidias-monstrous-volta-gpu-appears-packed-with-21-billion-transistors-and-5120-cores.html

https://www.nvidia.com/en-us/data-center/volta-gpu-architecture/?ncid=van-volta-gpu-architecture

http://wccftech.com/nvidia-volta-gv100-gpu-tesla-v100-architecture-specifications-deep-dive/

That's on par with a year and a half to two year refresh cycle. Stuff like this usually comes out before desktop cards come out and it's scheduled for fall, so guessing we'll see cards early spring.

[grrrgrrr]

hello everyone ,

nvidia volta appeared , and maybe we`ll see the new cards by christmas

i`m courious to see first mining rigs build on new arhitecture (nvlink vs pcie ) any thoughts ?

performance of v100 seems to be 4x gtx 1060 ...and that is on one single card

http://www.pcworld.com/article/3196026/components-graphics/nvidias-monstrous-volta-gpu-appears-packed-with-21-billion-transistors-and-5120-cores.html

https://www.nvidia.com/en-us/data-center/volta-gpu-architecture/?ncid=van-volta-gpu-architecture

http://wccftech.com/nvidia-volta-gv100-gpu-tesla-v100-architecture-specifications-deep-dive/

These cards are designed specifically for scientific computing. They have a cuda core that can do 4x4 matrix multiplication in 1 GPU cycle. The kind of optimization like Bitcoin mining ASICs.

So here's my question: how will GPU mining evolve? Currently any gaming card can mine reasonably well. But as Nvidia specializing their cards to deep learning and AMD specializing their cards to memory performance, it's become harder to democratize PoW mining. 4x4 matrix multiplication (or some future ISA innovations) could prove huge in some algos but a $15000 card is not something that common people will buy.

[chrysophylax]

hello everyone ,

nvidia volta appeared , and maybe we`ll see the new cards by christmas

i`m courious to see first mining rigs build on new arhitecture (nvlink vs pcie ) any thoughts ?

performance of v100 seems to be 4x gtx 1060 ...and that is on one single card

http://www.pcworld.com/article/3196026/components-graphics/nvidias-monstrous-volta-gpu-appears-packed-with-21-billion-transistors-and-5120-cores.html

https://www.nvidia.com/en-us/data-center/volta-gpu-architecture/?ncid=van-volta-gpu-architecture

http://wccftech.com/nvidia-volta-gv100-gpu-tesla-v100-architecture-specifications-deep-dive/

These cards are designed specifically for scientific computing. They have a cuda core that can do 4x4 matrix multiplication in 1 GPU cycle. The kind of optimization like Bitcoin mining ASICs.

So here's my question: how will GPU mining evolve? Currently any gaming card can mine reasonably well. But as Nvidia specializing their cards to deep learning and AMD specializing their cards to memory performance, it's become harder to democratize PoW mining. 4x4 matrix multiplication (or some future ISA innovations) could prove huge in some algos but a $15000 card is not something that common people will buy.

we are interested in these also ...

we have contacted nvidia - as well as a few infrastructure companies - and looking at designing a new breed of miner for thefarm ...

none of this will be cheap - and none of it for the 'everyday' miner ... until the desktop editions come out of course - and even then they will be very high end systems ...

#crysx

[bensam1231]

hello everyone ,

nvidia volta appeared , and maybe we`ll see the new cards by christmas

i`m courious to see first mining rigs build on new arhitecture (nvlink vs pcie ) any thoughts ?

performance of v100 seems to be 4x gtx 1060 ...and that is on one single card

http://www.pcworld.com/article/3196026/components-graphics/nvidias-monstrous-volta-gpu-appears-packed-with-21-billion-transistors-and-5120-cores.html

https://www.nvidia.com/en-us/data-center/volta-gpu-architecture/?ncid=van-volta-gpu-architecture

http://wccftech.com/nvidia-volta-gv100-gpu-tesla-v100-architecture-specifications-deep-dive/

These cards are designed specifically for scientific computing. They have a cuda core that can do 4x4 matrix multiplication in 1 GPU cycle. The kind of optimization like Bitcoin mining ASICs.

So here's my question: how will GPU mining evolve? Currently any gaming card can mine reasonably well. But as Nvidia specializing their cards to deep learning and AMD specializing their cards to memory performance, it's become harder to democratize PoW mining. 4x4 matrix multiplication (or some future ISA innovations) could prove huge in some algos but a $15000 card is not something that common people will buy.

Same way ASICs work right now. Extremely specialized hardware that are only really good at a few things (almost never more then one algo), so similar. They'll probably be great in some algos with customized programming and trash in others, somewhat like the way FPGAs work right now. I don't anticipate these will change things that much.

The more interesting option is the desktop cards that will follow this as original post was talking about.

[sp_]

They have a cuda core that can do 4x4 matrix multiplication in 1 GPU cycle.

But few cryptocoin algos need fast matrix multiplications.. The only algos I can think of is M7 (cryptonite), primecoin, Nexus (prime solver) ..

[sp_]

I am working on tvpruvot / DJezo's opensource Equihash kernel and managed to gain 9%. I am pretty sure my changes are different from excevator and ewfb.. I might opensource my work for free.. My mod works best with 4 threads pr gpu

[du44]

I am working on tvpruvot's opensource Equihash kernel and managed to gain 9%. I am pretty sure my changes are different from excevator and ewfb.. I might opensource my work for free.. My mod works best with 4 threads pr gpu

That's good news. I like sp forks for the speed and what more important stability. Hope this kernel will be stable too

[scryptr]

I am working on tvpruvot / DJezo's opensource Equihash kernel and managed to gain 9%. I am pretty sure my changes are different from excevator and ewfb.. I might opensource my work for free.. My mod works best with 4 threads pr gpu

IF YOU KEEP IT OPEN SOURCE I WILL DONATE AGAIN--

Please keep the code open source.  The storm of nasty comments that fester up after each of your binary releases is too much for me to bear.  

A few new coders have already posted binaries, several had an "illegal memory error" that seemed similar.  Epsylon3's own version has a broken Jackpot Hashing Algorithm (JHA) that generates a similar error.  Whatever you release may be prone to this error on some users' systems.  Just a warning, because miners need stable software just as much as they want speed enhancement.       --scryptr

EDIT: Everyone who is interested in an enhanced, open source CCminer that mines Equihash should at least pledge a donation.  The CPU boost feature on sp_ 's CCminer can add up to 5% speed in adition to any assembly code improvements.       --scryptr

[mrada1204]

any link for download ?
couldn`t find it on github

[sp_]

any link for download ?
couldn`t find it on github

It's not done yet.

[Amph]

yeah make it open source, i don't think you need peanuts of 0.05 from random dude that will resell it to another random dude, you have your farm that will cover everything

[bensam1231]

Everyone wants shit for free. People bitch and moan so it's 'free'. Don't be afraid to charge for your work. Real support doesn't come from people who only give you kind words so they get free things.

SP I'm going to once again suggest using a dev fee as I did multiple times. Even in work that uses other peoples work you can have a dev fee, the source just needs to be provided. At 2% there aren't many people that are going to go out of their way to remove it, especially with readily available binaries. OR add it as a option that's enabled by default and can be disabled with -nofee or something along those lines. You can charge for work that uses others work as long as the source is available.

So there really is no difference between giving it away 'for free' or just having a dev fee already built in with source available. If people choose to remove it you're still earning more then you would be otherwise.

Good that you're improving equihash though.

yeah make it open source, i don't think you need peanuts of 0.05 from random dude that will resell it to another random dude, you have your farm that will cover everything

That literally has nothing to do with his work. His farm will regardless of him 'donating' work to other miners. Matter of a fact it earns more if he doesn't. Also earning .05BTC from one person is more then earning none from everyone.

[sp_]

Everyone wants shit for free. People bitch and moan so it's 'free'. Don't be afraid to charge for your work. Real support doesn't come from people who only give you kind words so they get free things.

SP I'm going to once again suggest using a dev fee as I did multiple times. Even in work that uses other peoples work you can have a dev fee, the source just needs to be provided. At 2% there aren't many people that are going to go out of their way to remove it, especially with readily available binaries. OR add it as a option that's enabled by default and can be disabled with -nofee or something along those lines. You can charge for work that uses others work as long as the source is available.

I will opensource my feecode.

Good that you're improving equihash though.

9% faster than the opensource is still a few percent slower than EFWB. When I release it, he will probobly integrate my changes into his fork and continue to be the fastest..

[pallas]

Good that you're improving equihash though.

9% faster than the opensource is still a few percent slower than EFWB. When I release it, he will probobly integrate my changes into his fork and continue to be the fastest..

not necessarily.
maybe he did the same optimisations as you are doing now.

[sp_]

not necessarily.
maybe he did the same optimisations as you are doing now.

Could be.. but I don't think so. Let's see later.

[Amph]

Everyone wants shit for free. People bitch and moan so it's 'free'. Don't be afraid to charge for your work. Real support doesn't come from people who only give you kind words so they get free things.

SP I'm going to once again suggest using a dev fee as I did multiple times. Even in work that uses other peoples work you can have a dev fee, the source just needs to be provided. At 2% there aren't many people that are going to go out of their way to remove it, especially with readily available binaries. OR add it as a option that's enabled by default and can be disabled with -nofee or something along those lines. You can charge for work that uses others work as long as the source is available.

So there really is no difference between giving it away 'for free' or just having a dev fee already built in with source available. If people choose to remove it you're still earning more then you would be otherwise.

Good that you're improving equihash though.

yeah make it open source, i don't think you need peanuts of 0.05 from random dude that will resell it to another random dude, you have your farm that will cover everything

That literally has nothing to do with his work. His farm will regardless of him 'donating' work to other miners. Matter of a fact it earns more if he doesn't. Also earning .05BTC from one person is more then earning none from everyone.

the point is that he don't need theat money, he probably sitting on millions already, he is making 150k a month for your info

if i had that income i would not care about money anymore certainly i would not care about 0.05, or are people really that greedy here? i would not be surprised

[sp_]

the point is that he don't need theat money, he probably sitting on millions already, he is making 150k a month for your info

No I am not.
I have alot of debt.

Mining is risky, you need expensive equipment to mine with a profit. And GPU's are improving faster than moore's law...

[scryptr]

the point is that he don't need theat money, he probably sitting on millions already, he is making 150k a month for your info

No I am not.
I have alot of debt.

Mining is risky, you need expensive equipment to mine with a profit. And GPU's are improving faster than moore's law...

MOORE'S LAW CAN ONLY FAIL--

Murphy's Law rules all.       --scryptr

[bensam1231]

Everyone wants shit for free. People bitch and moan so it's 'free'. Don't be afraid to charge for your work. Real support doesn't come from people who only give you kind words so they get free things.

SP I'm going to once again suggest using a dev fee as I did multiple times. Even in work that uses other peoples work you can have a dev fee, the source just needs to be provided. At 2% there aren't many people that are going to go out of their way to remove it, especially with readily available binaries. OR add it as a option that's enabled by default and can be disabled with -nofee or something along those lines. You can charge for work that uses others work as long as the source is available.

So there really is no difference between giving it away 'for free' or just having a dev fee already built in with source available. If people choose to remove it you're still earning more then you would be otherwise.

Good that you're improving equihash though.

yeah make it open source, i don't think you need peanuts of 0.05 from random dude that will resell it to another random dude, you have your farm that will cover everything

That literally has nothing to do with his work. His farm will regardless of him 'donating' work to other miners. Matter of a fact it earns more if he doesn't. Also earning .05BTC from one person is more then earning none from everyone.

the point is that he don't need theat money, he probably sitting on millions already, he is making 150k a month for your info

if i had that income i would not care about money anymore certainly i would not care about 0.05, or are people really that greedy here? i would not be surprised

It's not your job to determine who does and doesn't need money. You have no idea how much he's making or the costs he's incurring. Stating someone should do charity work 'just cause you said so' isn't really the way things work.

How about you donate oh 30% of your earnings to developers constantly. I think you've earned enough.

[Amph]

Everyone wants shit for free. People bitch and moan so it's 'free'. Don't be afraid to charge for your work. Real support doesn't come from people who only give you kind words so they get free things.

SP I'm going to once again suggest using a dev fee as I did multiple times. Even in work that uses other peoples work you can have a dev fee, the source just needs to be provided. At 2% there aren't many people that are going to go out of their way to remove it, especially with readily available binaries. OR add it as a option that's enabled by default and can be disabled with -nofee or something along those lines. You can charge for work that uses others work as long as the source is available.

So there really is no difference between giving it away 'for free' or just having a dev fee already built in with source available. If people choose to remove it you're still earning more then you would be otherwise.

Good that you're improving equihash though.

yeah make it open source, i don't think you need peanuts of 0.05 from random dude that will resell it to another random dude, you have your farm that will cover everything

That literally has nothing to do with his work. His farm will regardless of him 'donating' work to other miners. Matter of a fact it earns more if he doesn't. Also earning .05BTC from one person is more then earning none from everyone.

the point is that he don't need theat money, he probably sitting on millions already, he is making 150k a month for your info

if i had that income i would not care about money anymore certainly i would not care about 0.05, or are people really that greedy here? i would not be surprised

It's not your job to determine who does and doesn't need money. You have no idea how much he's making or the costs he's incurring. Stating someone should do charity work 'just cause you said so' isn't really the way things work.

How about you donate oh 30% of your earnings to developers constantly. I think you've earned enough.

no i don't earn enough at all, i don't have 1000 gpu mining with 150k return, as i said if i was rich i would donat easily, but it's not my case, otherwise i would not even bother mining probably...