CCminer(SP-MOD) Modded GPU kernels.

[bensam1231]

sorry to post that here, but lbry cuda miner preview (binaries with sources) is on sale for 0.1 BTC... :p

PM me

will maybe get a tip from sp this time ? :p who know ?

tested working without issues on all cuda arch (3.0 to 6.1) on windows and linux

Code:

GPU #0: Gigabyte GTX 1080, 270.14 MH/s
GPU #1: GeForce GTX 1070, 192.61 MH/s
accepted: 339/339 (diff 0.115), 462.73 MH/s yes!
Stratum difficulty set to 16 (0.06250)
lbry block 11086, diff 9152.967
GPU #0: Gigabyte GTX 1080, 270.92 MH/s
accepted: 340/340 (diff 1.919), 462.74 MH/s yes!

How long is it going to be in 'preview' phase before you release it?

[fenomenhaa]

sorry to post that here, but lbry cuda miner preview (binaries with sources) is on sale for 0.1 BTC... :p

PM me

will maybe get a tip from sp this time ? :p who know ?

tested working without issues on all cuda arch (3.0 to 6.1) on windows and linux

Code:

GPU #0: Gigabyte GTX 1080, 270.14 MH/s
GPU #1: GeForce GTX 1070, 192.61 MH/s
accepted: 339/339 (diff 0.115), 462.73 MH/s yes!
Stratum difficulty set to 16 (0.06250)
lbry block 11086, diff 9152.967
GPU #0: Gigabyte GTX 1080, 270.92 MH/s
accepted: 340/340 (diff 1.919), 462.74 MH/s yes!

What about 970 and 750ti? )) 1080 crazy profitable bu iHaven't any o them. ( Still wating for rx 470's

[bensam1231]

Small request to developers, please show the intensity on GPUs when your miner is running somewhere. You shouldn't need to request the intensity for a GPU so you can tweak it. Saves both parties a extra step.

[pallas]

Small request to developers, please show the intensity on GPUs when your miner is running somewhere. You shouldn't need to request the intensity for a GPU so you can tweak it. Saves both parties a extra step.

That would be nice, unfortunately the default intensity of an algo is set inside that algo-specific file and is not store anywhere, so there is no nice way of doing it.
To do it the right way, there should be an algo structure holding default intensities for the various cards, so that it is initialised with either the user supplied one or the default, and printed at the beginning.
To be fair, sgminer doesn't even have algo-specific (let alone device-specific) default intensities, AFAIK.
Anyways it's pretty easy to spot them, just look at the file named as the algo plus ".cu", near the beginning you will see the code which sets intensities (throughput).

[bensam1231]

Small request to developers, please show the intensity on GPUs when your miner is running somewhere. You shouldn't need to request the intensity for a GPU so you can tweak it. Saves both parties a extra step.

That would be nice, unfortunately the default intensity of an algo is set inside that algo-specific file and is not store anywhere, so there is no nice way of doing it.
To do it the right way, there should be an algo structure holding default intensities for the various cards, so that it is initialised with either the user supplied one or the default, and printed at the beginning.
To be fair, sgminer doesn't even have algo-specific (let alone device-specific) default intensities, AFAIK.
Anyways it's pretty easy to spot them, just look at the file named as the algo plus ".cu", near the beginning you will see the code which sets intensities (throughput).

When it's mining it could simply do like (i-20) or (intensity 20) in the command line or when it first starts, show it next to each GPU it's winding up.

Yeah, I remember tweaking SGminer for intensity back when I mined with AMD.

[joblo]

Small request to developers, please show the intensity on GPUs when your miner is running somewhere. You shouldn't need to request the intensity for a GPU so you can tweak it. Saves both parties a extra step.

That would be nice, unfortunately the default intensity of an algo is set inside that algo-specific file and is not store anywhere, so there is no nice way of doing it.
To do it the right way, there should be an algo structure holding default intensities for the various cards, so that it is initialised with either the user supplied one or the default, and printed at the beginning.
To be fair, sgminer doesn't even have algo-specific (let alone device-specific) default intensities, AFAIK.
Anyways it's pretty easy to spot them, just look at the file named as the algo plus ".cu", near the beginning you will see the code which sets intensities (throughput).

It's already implemented in SP_MOD v80.

[Epsylon3]

Ok, i close the crowfunding for some days... stop sending me tips i reached the reasonable goal of 1 BTC

Each new donation reduce the overall possible profit... hard choice to do but the best way i found to reduce the hype effect over mining rewards

will maybe do another round for a smaller donation this week end

open source should be available during july (after the 20th july)

- mode accountant off -

[Amph]

Ok, i close the crowfunding for some days... stop sending me tips i reached the reasonable goal of 1 BTC

Each new donation reduce the overall possible profit... hard choice to do but the best way i found to reduce the hype effect over mining rewards

will maybe do another round for a smaller donation this week end

open source should be available during july (after the 20th july)

- mode accountant off -

yeah the hashrate is already increasing too much, better to limit it for the time being

and no i'm not even mining it

my dream was always to see a very high profitable coin that will remain profitable forever, no dumped to the ground after 1 week/month/etc...

[antantti]

0.1 btc was too much? And 6 x 1070 is cheap?

[Amph]

Wolfo released it for free, he was payed very well i presume?

[antantti]

Wolfo released it for free, he was payed very well i presume?

Looking at how amd does against nvidia wolf wasn't payed so well.

[rednoW]

Wolfo released it for free, he was payed very well i presume?

Looking at how amd does against nvidia wolf wasn't payed so well.

Yeah! Sure there is a much faster amd miner for private from Wolf ))

[giagge]

Epsylon3 you give a miner also to holders of nvidia, even slower than what you are using , please .

[bathrobehero]

0.1 btc was too much? And 6 x 1070 is cheap?

0.1 btc is not too much for most miners I think but it is too much if the source will be public in a week. That's a short period of time to make that 0.1 back (and then some).

[bensam1231]

0.1 btc was too much? And 6 x 1070 is cheap?

0.1 btc is not too much for most miners I think but it is too much if the source will be public in a week. That's a short period of time to make that 0.1 back (and then some).

Yup... basically depends on the size of the farm and time to RoI. Why I still think a miner fee is the best way to do things... still is... Scales whatever the size.


Network difficulty is at 420GH, it gained 120GH in the last six hours. Pretty certain it'll be on par with other algos in maybe a day or two.

[antantti]

0.1 btc was too much? And 6 x 1070 is cheap?

0.1 btc is not too much for most miners I think but it is too much if the source will be public in a week. That's a short period of time to make that 0.1 back (and then some).

It is not always about profit, I would have donated that 0.1 anyway. We need guys like epsylon, djm, c&c... and sp on top of that. If they all go awol what is left?

Talking about profits, hmm...

[bensam1231]

0.1 btc was too much? And 6 x 1070 is cheap?

0.1 btc is not too much for most miners I think but it is too much if the source will be public in a week. That's a short period of time to make that 0.1 back (and then some).

It is not always about profit, I would have donated that 0.1 anyway. We need guys like epsylon, djm, c&c... and sp on top of that. If they all go awol what is left?

Talking about profits, hmm...

True to your word, it's not too late to 'DONATE' your .1BTC today!

That aside, why I still believe a miner fee is a great way to go with things. If only there was some fair way to split it up among developers so they could all contribute to the same miner.

[sp_]

My undervolted R9 Nano is hashing @ 65MHASH / 100watt with the opensource sgminer  kernal..

Settings:

-X 1024 -g 2

At the current difficulty/pumped price I get around 3.5$ per card per day. (lbry)

Claymores etherum+sia only give around $2 a day per card.


1080@270 is pretty good. ($14 /day)

[antantti]

0.1 btc was too much? And 6 x 1070 is cheap?

0.1 btc is not too much for most miners I think but it is too much if the source will be public in a week. That's a short period of time to make that 0.1 back (and then some).

It is not always about profit, I would have donated that 0.1 anyway. We need guys like epsylon, djm, c&c... and sp on top of that. If they all go awol what is left?

Talking about profits, hmm...

True to your word, it's not too late to 'DONATE' your .1BTC today!

That aside, why I still believe a miner fee is a great way to go with things. If only there was some fair way to split it up among developers so they could all contribute to the same miner.

with only 4x970 i call it a donate.

amd guys got a nice jump start here but now there is epsys miner available. lbc not tanking.

[bathrobehero]

0.1 btc was too much? And 6 x 1070 is cheap?

0.1 btc is not too much for most miners I think but it is too much if the source will be public in a week. That's a short period of time to make that 0.1 back (and then some).

Yup... basically depends on the size of the farm and time to RoI. Why I still think a miner fee is the best way to do things... still is... Scales whatever the size.


Network difficulty is at 420GH, it gained 120GH in the last six hours. Pretty certain it'll be on par with other algos in maybe a day or two.

Most miners are small and there's not many big farms so a set price is far from optimal.

Let's see a theoritical and totally made up proportion of 3 sizes of miners (so don't get hung up on the numbers, just an example):
- number of small miners between  1 - 12 cards: 50
- medium miners between 12-60 cards: 20
- big miners with over 60+ cards: 5

Small miners can't afford 0.05-0.1 BTC and they wouldn't ROI anyway but they overall have 50 - 600 cards.
Medium miners probably can afford and might even profit from spending 0.05-0.1 BTC on a miner but it's risky and you're not going to make your money back about half of the time. They overall control somewhere between 240-1200 cards.
Big miners have no problem paying for 0.05-0.1 BTC for a miner and they are guaranteed to profit even from a tiny increase in profitability without any risk. They probably would pay much more for the same miner if they had to.

I don't mind donating to developers, I did so dozens of times and it's great way to thank them but a lot of people aren't donating and just rather wait for the open source release/leak and use that. And devs will likely open source their work shortly after selling them to a few miners in the hopes of more donations. I feel they should earn more by earning from more people.

I don't want to pour petrol on the bonfire that is the argument of a fee based miner and I get the risks and obstacles but that is the only method which taxes miners proportionately.
Team nvidia doesn't have a fee based miner at all so it would be free pickings and it would earn the developers a respectable amount which would further motivate them to work on their miner possibly pushing AMD back so eventually it probably also wouldn't matter if all nvidia miners were using the same miner. And it would also create a competition once more devs would see how much the other guy is earning.

I'm not by any means complaining, it just developing and optimizing miners for nvidia can't be sustainable with the odd donations and rare coins like LBRY.