CCminer(SP-MOD) Modded GPU kernels.

[reb0rn21]

Gigabyte G1 Gaming GTX1070 *6(8pin)
drivers: 372.54
ccminer1.8.0 ~1544M(~257M) temp:59C power:1140W via NVIDIA-SMI (nvidia-smi -ac 4004,1999 & nvidia-smi -pl 190 / Core +150)
SP-MOD#3 ~1800M(~300M) temp:59C power:1090W via NVIDIA-SMI (nvidia-smi -ac 4004,1999 & nvidia-smi -pl 190 / Core +150)

~15% increase

ccminer1.8.1 crashes my system using the same command used below for ccminer 1.8.0 and SP-MOD#3

Command used to achieve the above rates:

ccminer1.8.0 - ccminer-x64.exe -t 6 -a lbry -o stratum+tcp://lbry.pool.mn:8878 -u XXXX -p XXXX --show-diff --cpu-priority 4
SP-MOD#3 - ccminer.exe -t 6 -a lbry -o stratum+tcp://lbry.pool.mn:8878 -u XXXX -p XXXX --show-diff --cpu-priority 4

I use almost the same speed with my rigs but with 22% less power usage and same speed, is it you bad config I don`t know

With SP mod if you got 300mhs with 1090w usage that is gr8, but you ccminer 1.8 results are just terrible

[1714787087]

1714787087

[1714787087]

1714787087

[scryptr]

Guessing Nik bought it from Scrypt... He isn't responding because you didn't buy it from him.

CCminer has a bug with MRR, you have to use the dynamic port it tries to reconnect to as it doesn't tolerate stratum redirection for some reason. Epsylons version has this bug as well.

Still saying SP you should lower fee to 2% and it should be a free release.

HAVE NOT TRIED SP_ #3--

I don't have a copy to sell.  If what Nikolaj says is true, at least the way I read it, sp_ 's #3 is not behaving as expected with MiningRigRentals.  Actually, someone renting hash through a fee  based miner is probably getting skinned 2 times.  I like NiceHash because it is so easy, but as I did more math, I appreciate a pool like Epsylon3's more and more.  Never hooked-up to MiningRigRentals.  Don't have a real opinion there.

As for the basic "WTF IS GOING ON HERE?", I take it that the miner restarts the fee cycle with each reconnect, and never mines for Nikolaj.  Just guessing.

As much as I've supported sp_ with 0.1 BTC in the past, it used to mean $20-$25.  The upfront fee is too steep for me just now.       --scryptr

[go6ooo1212]

I haven't tryed it to MRR either , and I can't say much in that case.

[sp_]

Did version #2 work bether with MRR? 2 minute donation instead of 4.

[sp_]

Gigabyte G1 Gaming GTX1070 *6(8pin)
drivers: 372.54
ccminer1.8.0 ~1544M(~257M) temp:59C power:1140W via NVIDIA-SMI (nvidia-smi -ac 4004,1999 & nvidia-smi -pl 190 / Core +150)
SP-MOD#3 ~1800M(~300M) temp:59C power:1090W via NVIDIA-SMI (nvidia-smi -ac 4004,1999 & nvidia-smi -pl 190 / Core +150)

~15% increase

ccminer1.8.1 crashes my system using the same command used below for ccminer 1.8.0 and SP-MOD#3

Command used to achieve the above rates:

ccminer1.8.0 - ccminer-x64.exe -t 6 -a lbry -o stratum+tcp://lbry.pool.mn:8878 -u XXXX -p XXXX --show-diff --cpu-priority 4
SP-MOD#3 - ccminer.exe -t 6 -a lbry -o stratum+tcp://lbry.pool.mn:8878 -u XXXX -p XXXX --show-diff --cpu-priority 4

I use almost the same speed with my rigs but with 22% less power usage and same speed, is it you bad config I don`t know

With SP mod if you got 300mhs with 1090w usage that is gr8, but you ccminer 1.8 results are just terrible

Reduce the memclock as much as possible. 850mhz instead of 4000. And save some power.
My kernel isn't using any memory.

[rednoW]

Reduce the memclock as much as possible. 850mhz instead of 4000. And save some power.
My kernel isn't using any memory.

what tool did you use to low memory freq? nv_smi? nvidia_inspector?

[thesmokingman]

Reduce the memclock as much as possible. 850mhz instead of 4000. And save some power.
My kernel isn't using any memory.

what tool did you use to low memory freq? nv_smi? nvidia_inspector?

I use Afterburner 4.3.0 Beta as it works with Pascal cards. -502 is the most AFB will allow me to clock the memory down without a BIOS mod and things are still running just fine...

[sp_]

Nvidi-smi tool. You need to list the supported clocks first.

[rednoW]

Nvidi-smi tool. You need to list the supported clocks first.

Ok, I'll give it a try. Now I use Palit ThunderMaster, it can downclock-1000 only

[maxim000]

frequency and power limit possible to set in .bat file

4x1070 example:
ccminer_LBRY_priv3  -i 27  -c LBRY.conf  --mem-clock=3299,3299,3299,3299  --gpu-clock=1949,1949,1949,1949  --plimit=85,85,85,85  --tlimit=77,77,77,77

[sp_]

frequency and power limit possible to set in .bat file

4x1070 example:
ccminer_LBRY_priv3  -i 27  -c LBRY.conf  --mem-clock=3299,3299,3299,3299  --gpu-clock=1949,1949,1949,1949  --plimit=85,85,85,85  --tlimit=77,77,77,77

But then you have to run ccminer with admin rights.

[rednoW]

frequency and power limit possible to set in .bat file

4x1070 example:
ccminer_LBRY_priv3  -i 27  -c LBRY.conf  --mem-clock=3299,3299,3299,3299  --gpu-clock=1949,1949,1949,1949  --plimit=85,85,85,85  --tlimit=77,77,77,77

boost mem clock NVAPI not supported (((
please, can you give me a command line for nvidia-smi to low mem clock.

I tried nvidia-smi -ac 810,2037
which is in list of possible clocks but it didn't change anything.
And yes, I tried to restart miner.

[sp_]

boost mem clock NVAPI not supported (((

You can change the memclock when the gpu is in p2 state mode. Or mod the bios.

more info here:

http://cryptomining-blog.com/7341-how-to-squeeze-some-extra-performance-mining-ethereum-on-nvidia/

Here they set the memclock higher, but we want to set it lower.

[rednoW]

boost mem clock NVAPI not supported (((

You can change the memclock when the gpu is in p2 state mode. Or mod the bios.

more info here:

http://cryptomining-blog.com/7341-how-to-squeeze-some-extra-performance-mining-ethereum-on-nvidia/

Here they set the memclock higher, but we want to set it lower.

I check that it is in p2 state mode with nvidia inspector and try to change memory clock both with nvidia smi and nvidia inspector. It is still limited -1000 resulting in 1400mhz (5600 effective)

[sp_]

I check that it is in p2 state mode with nvidia inspector and try to change memory clock both with nvidia smi and nvidia inspector. It is still limited -1000 resulting in 1400mhz (5600 effective)

Then you are left with a bios mod for your card model. You need a modified version of nvflash in order to bypass the certificate checks..

[Epsylon3]

ccminer (mine) handle --plimit on all platforms, but windows x86 value is in percent (like afterburner)

In afterburner i have 50% for minimum value , i try set --plimit 45 on .bat ( windows 10 x64 ) but PL in ccminer set automatically on 50% ? .

yep, there is a hardcoded range in the card bios i think

advanced nvapi support was added in the 1.8... but as i said in "my" readme beware with clocks, the cards use offsets and when already changed, its the mess to find the original one prefer only --plimit and --tlimit for pascal

and before you ask.. no, tlimit doesnt exist on linux. nvapi is windows only (same api used by oc tools)

[Nikolaj]

Guessing Nik bought it from Scrypt... He isn't responding because you didn't buy it from him.

CCminer has a bug with MRR, you have to use the dynamic port it tries to reconnect to as it doesn't tolerate stratum redirection for some reason. Epsylons version has this bug as well.

Still saying SP you should lower fee to 2% and it should be a free release.

sp, not scryptr.

the issue of sp kernels about miningrigrentals it's very serious, because of the "donation" it's 100%, not 3.3.

Apart this, if you loose connection, it restarts with the donation.

If every time that the kernel starts you "donate", these are the kind of issues that you're forced to cope with, and not easy ones.

I hadn't such issues with claymore.

[Nikolaj]

Guessing Nik bought it from Scrypt... He isn't responding because you didn't buy it from him.

CCminer has a bug with MRR, you have to use the dynamic port it tries to reconnect to as it doesn't tolerate stratum redirection for some reason. Epsylons version has this bug as well.

Still saying SP you should lower fee to 2% and it should be a free release.

I agree, but with the kernel from epsylon you haven't got a fee, and a "forced donation" the first 4 minutes it starts (or reconnects)

[sp_]

Guessing Nik bought it from Scrypt... He isn't responding because you didn't buy it from him.

CCminer has a bug with MRR, you have to use the dynamic port it tries to reconnect to as it doesn't tolerate stratum redirection for some reason. Epsylons version has this bug as well.

Still saying SP you should lower fee to 2% and it should be a free release.

I agree, but with the kernel from epsylon you haven't got a fee, and a "forced donation" the first 4 minutes it starts, or reconnects.

There might be an issue in #3. If you loose connection for a long time, it will switch to donation mining and do another 4mins.

4 minutes for me 116 for you. if you loose connection within the 116mins, the process restarts. (if you use the -R reconnect parameter in the bat, you should put it to a high value or remove it.)

With -R 5 it will try to reconnect to your pool 5 times.

[tbearhere]

1 GHASH of LBRY is currently paying 0.02085BTC
1 980ti can do 300MHASH with my mod. 0.3*0.02085 = 0,006255BTC = $3.787 - 3.3%= $3.66  
1 980ti can do 230MHASH with the opensource 0.25*0.02085 = 0,0052125BTC = $3.148355
Increased Profit per card per day:
+$0,513674

I have 1.25gh  its paying .019btc a day!

But you have 980ti and 970 cards right? Then you can boost to 1.4 gh and getting  0.024btc per day with the sp-mod #3

0.05 more btc per day. 20 days to roi

Sorry for the late response sp. Are those hashrates at stock clocks? And do you hash with it?