CCminer(SP-MOD) Modded GPU kernels.

[CjMapope]

Denarius? Are you kidding me? I setup the wallet 3 weeks ago and it stopped working, cannot update it by some reason, so abandoned it. I have feeling that sp_ is fooling us

Not that miners care about stuff like this but DNR is Carsen Klock's newest, soon to be abandoned scamcoin - as it is his signature move after years of not releasing coins or maybe doing so anonymously. Doubt it though, considering his massive ego.

He released dozens of shitcoins which he all abandoned after the initial pump...

But he certainly evolved from shit like Carsencoin [COCK]: https://bitcointalk.org/index.php?topic=720684.0

Please fix power issue on skunk, because cards work very hot without overclocking and eat too much power. What oc settings do you use for 1060/1070 cards? I did 100% Power and -500 mem and 1060s runs ~72c, 1070s from 70 to 84

That means the algo is either ASIC friendly and that it's highly optimized and using the cards to close to their full potential.

ya, dont blame SP, he just making money off people with more $$ than math skills hehe
His mods are LEGIT, but math may change that daily

[D8V1D]

7 x 1080 TI FE

0.12 KWh
BTC 2775

DNR 0.000150
100% - 819 MH/s 87.6 DNR = $36.46 - 6.03 (2093 watts) = $30.43 per day
  70% - 748 MH/s 80.0 DNR = $33.30 - 4.00 (1390 watts) = $29.30 per day

96% of the profit come from the first 70% TDP.
People running at 100% are basically wasting 703 watts of electricity.
Not including the extra cooling cost to run at 100%.

SIGT 0.000012
 70% - 336 MH/s 1220 SIGT = $40.63 - 3.92 (1360 watts) = $36.71 per day

*DNR calculation are ruff estimates based on the earnings from the past 5 hours of mining.
 Watts are measured from the wall.

[ESJai]

Two of my rigs keep getting this error: one after about 7 hours, one after 5-10mins.
Both are using the latest driver.
I tried stock setting but it still occurs.
Also tried lower intensity, cuda schedule auto, cpu priority.
This is the skunkmod#3.

Cuda error in func 'skein512_cpu_setBlock_80' at line 1403 : unknown error

Can anyone help?
Thank you!

[mystere_miner]

Stick to one model or another SP, you're definitely going to piss off a lot of people (myself included) if you try to fuck people over for more money constantly.

I bought the skunk miner 2 days ago, and the price is decreasing, and the difficulty rising faster than I can make up for it.  I've literally mined hundreds of coins, and when I look at the current price, I've not made any additional money because the price has been dropping.  I'm hoping beyond hope for a bump so I can sell off, but at this rate i'm going to have spent .05BTC with nothing to show for it.  Except for losing the funds I would have made if I had just been doing my normal mining, instead of trying t pay back the price of this miner.

Now SP wants to charge twice that for the next release?  I have no hope of making my money back on the first one.. I'd have to have 20 or 30 miners to actually make a profit.

[strannik951159]

Two of my rigs keep getting this error: one after about 7 hours, one after 5-10mins.
Both are using the latest driver.
I tried stock setting but it still occurs.
Also tried lower intensity, cuda schedule auto, cpu priority.
This is the skunkmod#3.

Cuda error in func 'skein512_cpu_setBlock_80' at line 1403 : unknown error

Can anyone help?
Thank you!

Overclocking м

[enochi]

7 x 1080 TI FE

0.12 KWh
BTC 2775

DNR 0.000150
100% - 819 MH/s 87.6 DNR = $36.46 - 6.03 (2093 watts) = $30.43 per day
  70% - 748 MH/s 80.0 DNR = $33.30 - 4.00 (1390 watts) = $29.30 per day

96% of the profit come from the first 70% TDP.
People running at 100% are basically wasting 703 watts of electricity.
Not including the extra cooling cost to run at 100%.

SIGT 0.000012
 70% - 336 MH/s 1220 SIGT = $40.63 - 3.92 (1360 watts) = $36.71 per day

*DNR calculation are ruff estimates based on the earnings from the past 5 hours of mining.
 Watts are measured from the wall.

so SIGT is still more profitable than the sp's DNR ?

[mystere_miner]

SIGT 0.000012
 70% - 336 MH/s 1220 SIGT = $40.63 - 3.92 (1360 watts) = $36.71 per day

That's about what i'm making, but considering i'd be making ~$20/day mining normally, that means i'm only making ~16/day over what i'd be normally making.  That's significant, but I have to pay back the .05BTC I invested in this..  at this rate, which is shrinking hourly it seems.. it will take 15+ days to break even, assuming it doesn't keep dropping in price... and the difficulty doesn't keep going up rapidly.

I can't see how the .1 he wants for the tribus miner will ever pay back, much less pay off.. unless you're a HUGE farm.  Considering that just since SP released his miner, the price of DNR has dropped by more than half (.00024 a few hours ago, and now less than .0001)

[joblo]

Denarius? Are you kidding me? I setup the wallet 3 weeks ago and it stopped working, cannot update it by some reason, so abandoned it. I have feeling that sp_ is fooling us

Not that miners care about stuff like this but DNR is Carsen Klock's newest, soon to be abandoned scamcoin - as it is his signature move after years of not releasing coins or maybe doing so anonymously. Doubt it though, considering his massive ego.

He released dozens of shitcoins which he all abandoned after the initial pump...

But he certainly evolved from shit like Carsencoin [COCK]: https://bitcointalk.org/index.php?topic=720684.0

Please fix power issue on skunk, because cards work very hot without overclocking and eat too much power. What oc settings do you use for 1060/1070 cards? I did 100% Power and -500 mem and 1060s runs ~72c, 1070s from 70 to 84

That means the algo is either ASIC friendly and that it's highly optimized and using the cards to close to their full potential.

A lot of people get that wrong. A hotter card on ceartain algos means the cuda cores are working hard and aren't waiting
on memory. That's a good thing, sort of. Such compute intensive algos are easy to implement in ASIC. The only thing
preventing it is demand. All of these algos that reshuffle sha3 functions are likely to get ASICs eventually if the coins that
use them are successful.

Moving from CPU to GPU to ASIC improves compute performance by orders of magnitude, compare a CPU with 8 cores vs
a GPU with 2000. Memory capacity and bandwidth can't keep up with the number of cores on a GPU so on algos that are
memory hard the cores sit idle much of the time waiting for data from memory, and consequently run cooler.

[Slava_K]

Two of my rigs keep getting this error: one after about 7 hours, one after 5-10mins.
Both are using the latest driver.
I tried stock setting but it still occurs.
Also tried lower intensity, cuda schedule auto, cpu priority.
This is the skunkmod#3.

Cuda error in func 'skein512_cpu_setBlock_80' at line 1403 : unknown error

Can anyone help?
Thank you!

Don`t use stock settings. Use TDP not more then 90, better 70-80... Undercklocking memory is not necessary. And that's enough.

[bathrobehero]

Denarius? Are you kidding me? I setup the wallet 3 weeks ago and it stopped working, cannot update it by some reason, so abandoned it. I have feeling that sp_ is fooling us

Not that miners care about stuff like this but DNR is Carsen Klock's newest, soon to be abandoned scamcoin - as it is his signature move after years of not releasing coins or maybe doing so anonymously. Doubt it though, considering his massive ego.

He released dozens of shitcoins which he all abandoned after the initial pump...

But he certainly evolved from shit like Carsencoin [COCK]: https://bitcointalk.org/index.php?topic=720684.0

Please fix power issue on skunk, because cards work very hot without overclocking and eat too much power. What oc settings do you use for 1060/1070 cards? I did 100% Power and -500 mem and 1060s runs ~72c, 1070s from 70 to 84

That means the algo is either ASIC friendly and that it's highly optimized and using the cards to close to their full potential.

ya, dont blame SP, he just making money off people with more $$ than math skills hehe
His mods are LEGIT, but math may change that daily

Investments are risky and Sp_'s mods are no exceptions. I bought most of them but not all of them. I'll certainly not buy optimizations for tribus.

Unfortunately the risk depends on how big ones farm is. For some people 0.05-0.1 is nothing, for others it takes weeks to earn that much making it a bad investment. But then again, a fee only system is very hard to develop - as it was discussed here numerous times.

[ESJai]

Two of my rigs keep getting this error: one after about 7 hours, one after 5-10mins.
Both are using the latest driver.
I tried stock setting but it still occurs.
Also tried lower intensity, cuda schedule auto, cpu priority.
This is the skunkmod#3.

Cuda error in func 'skein512_cpu_setBlock_80' at line 1403 : unknown error

Can anyone help?
Thank you!

Don`t use stock settings. Use TDP not more then 90, better 70-80... Undercklocking memory is not necessary. And that's enough.

Of course I would use lower TDP but that gave me the error and that's why I tried stock setting......
I tried both stock memory and lower memory clock, nothing changes.

Overclocking м

I just said I also used stock setting and still get the error so overclocking is not the problem.

[Slava_K]

Two of my rigs keep getting this error: one after about 7 hours, one after 5-10mins.
Both are using the latest driver.
I tried stock setting but it still occurs.
Also tried lower intensity, cuda schedule auto, cpu priority.
This is the skunkmod#3.

Cuda error in func 'skein512_cpu_setBlock_80' at line 1403 : unknown error

Can anyone help?
Thank you!

Don`t use stock settings. Use TDP not more then 90, better 70-80... Undercklocking memory is not necessary. And that's enough.

Of course I would use lower TDP but that gave me the error and that's why I tried stock setting......
I tried both stock memory and lower memory clock, nothing changes.

Overclocking м

I just said I also used stock setting and still get the error so overclocking is not the problem.

Test memory, PSU wattage, bios option about 4G limitation - today I have same problem after bios settings failing...

[ESJai]

Two of my rigs keep getting this error: one after about 7 hours, one after 5-10mins.
Both are using the latest driver.
I tried stock setting but it still occurs.
Also tried lower intensity, cuda schedule auto, cpu priority.
This is the skunkmod#3.

Cuda error in func 'skein512_cpu_setBlock_80' at line 1403 : unknown error

Can anyone help?
Thank you!

Don`t use stock settings. Use TDP not more then 90, better 70-80... Undercklocking memory is not necessary. And that's enough.

Of course I would use lower TDP but that gave me the error and that's why I tried stock setting......
I tried both stock memory and lower memory clock, nothing changes.

Overclocking м

I just said I also used stock setting and still get the error so overclocking is not the problem.

Test memory, PSU wattage, bios option about 4G limitation - today I have same problem after bios settings failing...

Memory, you mean ram?
Wattage isn't the problem Im sure.
What is 4G limitation? I've never heard of that. Any links for a reference to it?

[Garuda55]

Why price for skunk cheap then tribus?
Denarius will to the moon? but i didnt see that, market order to low

just hope sp_ from futere hahaha
but im already hopeless, price down 50% 

[ESJai]

Why price for skunk cheap then tribus?
Denarius will to the moon? but i didnt see that, market order to low

just hope sp_ from futere hahaha
but im already hopeless, price down 50% 

Maybe because people are stealing his work? From the conversations I've been reading in this post(?)

[bensam1231]

7 x 1080 TI FE

0.12 KWh
BTC 2775

DNR 0.000150
100% - 819 MH/s 87.6 DNR = $36.46 - 6.03 (2093 watts) = $30.43 per day
  70% - 748 MH/s 80.0 DNR = $33.30 - 4.00 (1390 watts) = $29.30 per day

96% of the profit come from the first 70% TDP.
People running at 100% are basically wasting 703 watts of electricity.
Not including the extra cooling cost to run at 100%.

SIGT 0.000012
 70% - 336 MH/s 1220 SIGT = $40.63 - 3.92 (1360 watts) = $36.71 per day

*DNR calculation are ruff estimates based on the earnings from the past 5 hours of mining.
 Watts are measured from the wall.

Measure wattage at wall NOT TDP. TDP is not a number that's set in stone. It varies by models, brand, and manufacturer. In addition to that you can have an algo that will use 78% TDP. That means when you're running at 70% you're only reducing power by 8% (once again this varies a lot) and when you increase it to 100% it DOES NOT use 100% of your available TDP as the algo does not use it.

Even if say the algo is using up to 100% TDP and you increase it from 70% that does not also mean it's using 30% more power. Certain algos use more power then others and in different ways. That means a 30% increase could only be using 15% more power measured at the wall.

[Garuda55]

Why price for skunk cheap then tribus?
Denarius will to the moon? but i didnt see that, market order to low

just hope sp_ from futere hahaha
but im already hopeless, price down 50% 

Maybe because people are stealing his work? From the conversations I've been reading in this post(?)

yeah myabe cause some ppl stealing his work and sell again, but better he bring high price for good profit, customer happy = seller rich
But he already doing good work 
Just hope price mod tribus down, and i can get free mod skunk hahahah 

[D8V1D]

7 x 1080 TI FE

0.12 KWh
BTC 2775

DNR 0.000150
100% - 819 MH/s 87.6 DNR = $36.46 - 6.03 (2093 watts) = $30.43 per day
  70% - 748 MH/s 80.0 DNR = $33.30 - 4.00 (1390 watts) = $29.30 per day

96% of the profit come from the first 70% TDP.
People running at 100% are basically wasting 703 watts of electricity.
Not including the extra cooling cost to run at 100%.

SIGT 0.000012
 70% - 336 MH/s 1220 SIGT = $40.63 - 3.92 (1360 watts) = $36.71 per day

*DNR calculation are ruff estimates based on the earnings from the past 5 hours of mining.
 Watts are measured from the wall.

Measure wattage at wall NOT TDP. TDP is not a number that's set in stone. It varies by models, brand, and manufacturer. In addition to that you can have an algo that will use 78% TDP. That means when you're running at 70% you're only reducing power by 8% (once again this varies a lot) and when you increase it to 100% it DOES NOT use 100% of your available TDP as the algo does not use it.

Even if say the algo is using up to 100% TDP and you increase it from 70% that does not also mean it's using 30% more power. Certain algos use more power then others and in different ways. That means a 30% increase could only be using 15% more power measured at the wall.

All the wattage above were measure from the wall with a Kill-A-Watt meter. I even specifically said "Watts are measured from the wall." in the post.
Most fully optimized algorithms will use up to 1400 watts at 70% TDP, that's true for DNR (1390), SIGT (1360), and ZEC (1400).
The highest I've test for at 100% TDP is 2200 watts on EWBF ZEC. DNR is using 2093 watts at 100%.

When you spend money to buy 21 x 1080 Ti, you do test to find optimized settings. If people want to run them at 100% that's completely up to them.
I'm just putting out results that I have found best.

[bensam1231]

7 x 1080 TI FE

0.12 KWh
BTC 2775

DNR 0.000150
100% - 819 MH/s 87.6 DNR = $36.46 - 6.03 (2093 watts) = $30.43 per day
  70% - 748 MH/s 80.0 DNR = $33.30 - 4.00 (1390 watts) = $29.30 per day

96% of the profit come from the first 70% TDP.
People running at 100% are basically wasting 703 watts of electricity.
Not including the extra cooling cost to run at 100%.

SIGT 0.000012
 70% - 336 MH/s 1220 SIGT = $40.63 - 3.92 (1360 watts) = $36.71 per day

*DNR calculation are ruff estimates based on the earnings from the past 5 hours of mining.
 Watts are measured from the wall.

Measure wattage at wall NOT TDP. TDP is not a number that's set in stone. It varies by models, brand, and manufacturer. In addition to that you can have an algo that will use 78% TDP. That means when you're running at 70% you're only reducing power by 8% (once again this varies a lot) and when you increase it to 100% it DOES NOT use 100% of your available TDP as the algo does not use it.

Even if say the algo is using up to 100% TDP and you increase it from 70% that does not also mean it's using 30% more power. Certain algos use more power then others and in different ways. That means a 30% increase could only be using 15% more power measured at the wall.

All the wattage above were measure from the wall with a Kill-A-Watt meter. I even specifically said "Watts are measured from the wall." in the post.
Most fully optimized algorithms will use up to 1400 watts at 70% TDP, that's true for DNR (1390), SIGT (1360), and ZEC (1400).
The highest I've test for at 100% TDP is 2200 watts on EWBF ZEC. DNR is using 2093 watts at 100%.

When you spend money to buy 21 x 1080 Ti, you do test to find optimized settings. If people want to run them at 100% that's completely up to them.
I'm just putting out results that I have found best.

Vice versa, not sure how you're exceeding an additional 30% power on TDP. While you run into scenarios where you'll end up using less power increasing TDP (versus the percentage increase), when it goes the opposite direction it shouldn't increase power usage more then 30%. For you a 30% increase in TDP increased your power by 50% and had almost no impact what so ever on your hashrate.

There are definitely algos that use more power then Equihash, SIB for instance. Equihash also has a weird quirk where there isn't a direct correlation between reducing TDP and reducing your hashrate, almost every algo has a direct correlation between power usage and hashrate, at least for Nvidia. With AMD you can reduce your core voltage at the same TDP level (which effectively increases your TDP headroom without increasing power).

That's neat you have GPUs... You aren't the only one nor is that a lot.

And yup, DNR has been around for awhile and it's lackluster compared to alternatives even with the modded miner. I didn't say anything because SP likes getting everyone excited before they figure out what they're doing and buy his miners. There are easy ways to check out earnings though without doing all of that, but I'll let you figure it out since you have 21x1080tis.

[sp_]

I working on a fixed spreadminer #12. Compiling with cuda 8.0 and compute 6.1 worked with the latest driver, but I get some performance lost.