Topic: CCminer(SP-MOD) Modded NVIDIA Maxwell / Pascal kernels. - page 849. (Read 2347659 times)

Quote from: chrysophylax on December 02, 2015, 06:41:48 AM

Quote from: sp_ on December 02, 2015, 07:02:28 AM

The --benchmark --time-limit 200 isn't working as is it supposed to.
You would expect ccminer to mine for 200 secs and display the average hashrate in the whole interval, but it is not.
It will simply mine for 200 secs, and display the average hashrate for the last second or so in the 200sec interval. Since the hashrate is fluctating alot, this will not give a correct output.
If somebody could correct this and create a average function of all the outputs to the commandline I could merge your changes into the main branch. Should be a 15min-30min job for a experienced C coder.

also sp - as i mentioned in the nicehash thread - would it be possible to add an eth kernel in ccminer-spmod? ...
even if its not an optimized one? ...
#crysx

Yes, I am planning to add genoils kernal, but I don't think I will add the code that generate the Dag file, so you need to generate this file with the original miner.

what would that do - if you did add the code? ...
that just sounds quite fiddly to me mate ...
#crysx

You just generate the Dag file, and this file will be valid for a while until the blocksize changes. So you can mine for weeks without changing the dagger.

A SIGNIFICANT ADDITION! --

Worth donations from all users. I hope to see more about this soon! --scryptr

ill always donate anyway ...

but yes - agreed ... i still think it would be worth it to have the whole thing in ONE miner ...

it always seems to be the same situation for me ... where if you had one or two machines mining - that would be easy to do - but a farm is more trouble than its worth ... it really is ...

one miner that does it all ... thats my vote ...

#crysx

scryptr

legendary

Activity: 1797

Merit: 1028

Quote from: chrysophylax on December 02, 2015, 06:51:39 AM

Quote from: chrysophylax on December 02, 2015, 06:41:48 AM

Quote from: chrysophylax on December 02, 2015, 06:51:39 AM

The --benchmark --time-limit 200 isn't working as is it supposed to.
You would expect ccminer to mine for 200 secs and display the average hashrate in the whole interval, but it is not.
It will simply mine for 200 secs, and display the average hashrate for the last second or so in the 200sec interval. Since the hashrate is fluctating alot, this will not give a correct output.
If somebody could correct this and create a average function of all the outputs to the commandline I could merge your changes into the main branch. Should be a 15min-30min job for a experienced C coder.

also sp - as i mentioned in the nicehash thread - would it be possible to add an eth kernel in ccminer-spmod? ...
even if its not an optimized one? ...
#crysx

Yes, I am planning to add genoils kernal, but I don't think I will add the code that generate the Dag file, so you need to generate this file with the original miner.

what would that do - if you did add the code? ...
that just sounds quite fiddly to me mate ...
#crysx

You just generate the Dag file, and this file will be valid for a while until the blocksize changes. So you can mine for weeks without changing the dagger.

A SIGNIFICANT ADDITION! --

Worth donations from all users. I hope to see more about this soon! --scryptr

sp_

legendary

Activity: 2954

Merit: 1087

Team Black developer

Quote from: chrysophylax on December 02, 2015, 06:41:48 AM

Quote from: chrysophylax on December 02, 2015, 06:51:39 AM

The --benchmark --time-limit 200 isn't working as is it supposed to.
You would expect ccminer to mine for 200 secs and display the average hashrate in the whole interval, but it is not.
It will simply mine for 200 secs, and display the average hashrate for the last second or so in the 200sec interval. Since the hashrate is fluctating alot, this will not give a correct output.
If somebody could correct this and create a average function of all the outputs to the commandline I could merge your changes into the main branch. Should be a 15min-30min job for a experienced C coder.

also sp - as i mentioned in the nicehash thread - would it be possible to add an eth kernel in ccminer-spmod? ...
even if its not an optimized one? ...
#crysx

Yes, I am planning to add genoils kernal, but I don't think I will add the code that generate the Dag file, so you need to generate this file with the original miner.

what would that do - if you did add the code? ...
that just sounds quite fiddly to me mate ...
#crysx

You just generate the Dag file, and this file will be valid for a while until the blocksize changes. So you can mine for weeks without changing the dagger.

pallas

legendary

Activity: 2716

Merit: 1094

Black Belt Developer

Quote from: chrysophylax on December 02, 2015, 06:41:48 AM

The --benchmark --time-limit 200 isn't working as is it supposed to.
You would expect ccminer to mine for 200 secs and display the average hashrate in the whole interval, but it is not.
It will simply mine for 200 secs, and display the average hashrate for the last second or so in the 200sec interval. Since the hashrate is fluctating alot, this will not give a correct output.
If somebody could correct this and create a average function of all the outputs to the commandline I could merge your changes into the main branch. Should be a 15min-30min job for a experienced C coder.

also sp - as i mentioned in the nicehash thread - would it be possible to add an eth kernel in ccminer-spmod? ...
even if its not an optimized one? ...
#crysx

Yes, I am planning to add genoils kernal, but I don't think I will add the code that generate the Dag file, so you need to generate this file with the original miner.

what would that do - if you did add the code? ...

that just sounds quite fiddly to me mate ...

#crysx

I think you'd need to periodically run the eth miner to generate the dag files, so ccminer can use them.

chrysophylax

legendary

Activity: 2940

Merit: 1091

--- ChainWorks Industries ---

Quote from: chrysophylax on December 02, 2015, 06:41:48 AM

Quote from: chrysophylax on December 02, 2015, 06:41:48 AM

The --benchmark --time-limit 200 isn't working as is it supposed to.
You would expect ccminer to mine for 200 secs and display the average hashrate in the whole interval, but it is not.
It will simply mine for 200 secs, and display the average hashrate for the last second or so in the 200sec interval. Since the hashrate is fluctating alot, this will not give a correct output.
If somebody could correct this and create a average function of all the outputs to the commandline I could merge your changes into the main branch. Should be a 15min-30min job for a experienced C coder.

also sp - as i mentioned in the nicehash thread - would it be possible to add an eth kernel in ccminer-spmod? ...
even if its not an optimized one? ...
#crysx

Yes, I am planning to add genoils kernal, but I don't think I will add the code that generate the Dag file, so you need to generate this file with the original miner.

what would that do - if you did add the code? ...

that just sounds quite fiddly to me mate ...

#crysx

sp_

legendary

Activity: 2954

Merit: 1087

Team Black developer

The --benchmark --time-limit 200 isn't working as is it supposed to.
You would expect ccminer to mine for 200 secs and display the average hashrate in the whole interval, but it is not.
It will simply mine for 200 secs, and display the average hashrate for the last second or so in the 200sec interval. Since the hashrate is fluctating alot, this will not give a correct output.
If somebody could correct this and create a average function of all the outputs to the commandline I could merge your changes into the main branch. Should be a 15min-30min job for a experienced C coder.

also sp - as i mentioned in the nicehash thread - would it be possible to add an eth kernel in ccminer-spmod? ...
even if its not an optimized one? ...
#crysx

Yes, I am planning to add genoils kernal, but I don't think I will add the code that generate the Dag file, so you need to generate this file with the original miner.

chrysophylax

legendary

Activity: 2940

Merit: 1091

--- ChainWorks Industries ---

Quote from: sp_ on December 02, 2015, 04:09:54 AM

The --benchmark --time-limit 200 isn't working as is it supposed to.

You would expect ccminer to mine for 200 secs and display the average hashrate in the whole interval, but it is not.

It will simply mine for 200 secs, and display the average hashrate for the last second or so in the 200sec interval. Since the hashrate is fluctating alot, this will not give a correct output.
If somebody could correct this and create a average function of all the outputs to the commandline I could merge your changes into the main branch. Should be a 15min-30min job for a experienced C coder.

also sp - as i mentioned in the nicehash thread - would it be possible to add an eth kernel in ccminer-spmod? ...

even if its not an optimized one? ...

#crysx

sp_

legendary

Activity: 2954

Merit: 1087

Team Black developer

The --benchmark --time-limit 200 isn't working as is it supposed to.

You would expect ccminer to mine for 200 secs and display the average hashrate in the whole interval, but it is not.

It will simply mine for 200 secs, and display the average hashrate for the last second or so in the 200sec interval. Since the hashrate is fluctating alot, this will not give a correct output.
If somebody could correct this and create a average function of all the outputs to the commandline I could merge your changes into the main branch. Should be a 15min-30min job for a experienced C coder.

pallas

legendary

Activity: 2716

Merit: 1094

Black Belt Developer

maybe -q suppresses the output.

impulse2000

newbie

Activity: 26

Merit: 0

Quote from: impulse2000 on December 02, 2015, 04:03:30 AM

Quote from: bathrobehero on December 02, 2015, 03:25:21 AM

Quote from: thefix on December 02, 2015, 01:06:38 AM

Quote from: AzzAz on December 01, 2015, 08:56:00 PM

How much hash does a 750Ti get on Lyra2REv2?

I get between 5000 and 5500 depending on the card and overclock

Huh, I only get about 4.6 Mhs with about +100 Mhz.

Test machine agrees with you, running sp_'s git.

Lyra2v2 GTX750Ti +185/+185
[2015-12-02 12:00:21] Stratum difficulty set to 4
[2015-12-02 12:00:25] accepted: 41/41 (100.00%), 5176 kH/s (yes!!!)
[2015-12-02 12:00:27] accepted: 42/42 (100.00%), 5175 kH/s (yes!!!)
[2015-12-02 12:00:31] accepted: 43/43 (100.00%), 5175 kH/s (yes!!!)
[2015-12-02 12:00:51] accepted: 44/44 (100.00%), 5175 kH/s (yes!!!)
[2015-12-02 12:01:08] accepted: 45/45 (100.00%), 5175 kH/s (yes!!!)

The sp-modded lyra2v2 on the 750ti is almost as fast as the DJM34 opensource kernal on the 960.

~~No Linux~~

My code will almost turn your 750ti into a gtx 960.

I haven't submitted anything to github in almost 2 weeks now. I am taking a break.

Make some BAT file

Code:

ccminer69.exe --no-color --benchmark --time-limit 200 -a lyra2v2>>bench69.txt
ccminer69.exe -q --no-color --benchmark --time-limit 200 -a quark >> bench69.txt
ccminer69.exe -q --no-color --benchmark --time-limit 200 -a qubit >> bench69.txt
ccminer69.exe -q --no-color --benchmark --time-limit 200 -a x11 >> bench69.txt
ccminer69.exe -q --no-color --benchmark --time-limit 200 -a x13 >> bench69.txt
ccminer69.exe -q --no-color --benchmark --time-limit 200 -a x15 >> bench69.txt
ccminer69.exe -q --no-color --benchmark --time-limit 200 -a neoscrypt >> bench69.txt
ccminer69.exe -q --no-color --benchmark --time-limit 200 -a groestl >> bench69.txt
ccminer69.exe -q --no-color --benchmark --time-limit 200 -a keccak >> bench69.txt
ccminer69.exe -q --no-color --benchmark --time-limit 200 -a bitcoin >> bench69.txt
ccminer69.exe -q --no-color --benchmark --time-limit 200 -a whirlpoolx >> bench69.txt
ccminer69.exe -q --no-color --benchmark --time-limit 200 -a nist5 >> bench69.txt

ccminer74.exe -q --no-color --benchmark --time-limit 200 -a lyra2v2 >> bench74.txt
ccminer74.exe -q --no-color --benchmark --time-limit 200 -a quark >> bench74.txt
ccminer74.exe -q --no-color --benchmark --time-limit 200 -a qubit >> bench74.txt
ccminer74.exe -q --no-color --benchmark --time-limit 200 -a x11 >> bench74.txt
ccminer74.exe -q --no-color --benchmark --time-limit 200 -a x13 >> bench74.txt
ccminer74.exe -q --no-color --benchmark --time-limit 200 -a x15 >> bench74.txt
ccminer74.exe -q --no-color --benchmark --time-limit 200 -a neoscrypt >> bench74.txt
ccminer74.exe -q --no-color --benchmark --time-limit 200 -a groestl >> bench74.txt
ccminer74.exe -q --no-color --benchmark --time-limit 200 -a keccak >> bench74.txt
ccminer74.exe -q --no-color --benchmark --time-limit 200 -a bitcoin >> bench74.txt
ccminer74.exe -q --no-color --benchmark --time-limit 200 -a whirlpoolx >> bench74.txt
ccminer74.exe -q --no-color --benchmark --time-limit 200 -a nist5 >> bench74.txt

But in TXT file only header:

Code:

*** ccminer 1.5.69-git(SP-MOD) for nVidia GPUs by sp-hash@github ***
Built with VC++ 2013 and nVidia CUDA SDK 6.5

Based on pooler cpuminer 2.3.2 and the tpruvot@github fork
CUDA support by Christian Buchner, Christian H. and DJM34
Includes optimizations implemented by sp, klaust, tpruvot, tsiv and pallas.

Whatsup?

Cryptozillah

hero member

Activity: 687

Merit: 502

Quote from: joblo on December 01, 2015, 06:20:38 PM

Quote from: Fuzzbawls on December 01, 2015, 06:04:23 PM

Quote from: Cryptozillah on December 01, 2015, 05:46:41 PM

I have been using the nicehash miner for the last few days.
Not sure what is most profitable because the 980 rig is going x13/neoscrypt and the 750Ti rigs are going lyra2rev2.
Why are they choosing different algos ? If the miner itself should choose the most profitable algo all my miners should mine the same ?

power-to-hashrate ratio is likely different? Haven't really looked at the nicehash miner myself but that would make sense.

It's a fact that lyra2v2 hashes better on 750ti than 9xx when compared to other algos, so there are times when lyra2v2 is
most profitable with 750ti but another algo is highest with 9xx GPUs. Maybe the nicehash miner knows this and adjusts
its profit switching accordingly. You should post your question in the nicehash thread for a definitive answer.

that is correct, that's why you have "benchmark" - to test your rig and it mines accordingly. That is why for nicehash miner you should build rig of 1 type card.

Ok, i have been building my rigs like that so i should be ok then. Thanks

chrysophylax

legendary

Activity: 2940

Merit: 1091

--- ChainWorks Industries ---

Quote from: pallas on December 02, 2015, 04:14:36 AM

Quote from: Al_may on December 01, 2015, 10:48:27 PM

Quote from: hammer24p on December 01, 2015, 08:55:46 PM

hashpower connection refuse what that mien

it means that the data packet got corrupt some where between you and the pool.
with all of the bad weather that is going on in the usa right a lot of pools are having internet problems

hashpower is having problems: it happens periodically because the software is unstable and there is no constant maintenance going on (the same for ffpool).

griimp would be completely stable AND ongoing Wink

...

just sayin ... hehehe ...

#crysx

pallas

legendary

Activity: 2716

Merit: 1094

Black Belt Developer

Quote from: Al_may on December 01, 2015, 10:48:27 PM

Quote from: hammer24p on December 01, 2015, 08:55:46 PM

hashpower connection refuse what that mien

it means that the data packet got corrupt some where between you and the pool.
with all of the bad weather that is going on in the usa right a lot of pools are having internet problems

hashpower is having problems: it happens periodically because the software is unstable and there is no constant maintenance going on (the same for ffpool).

sp_

legendary

Activity: 2954

Merit: 1087

Team Black developer

Quote from: impulse2000 on December 02, 2015, 04:03:30 AM

Quote from: bathrobehero on December 02, 2015, 03:25:21 AM

Quote from: thefix on December 02, 2015, 01:06:38 AM

How much hash does a 750Ti get on Lyra2REv2?

I get between 5000 and 5500 depending on the card and overclock

Huh, I only get about 4.6 Mhs with about +100 Mhz.

Test machine agrees with you, running sp_'s git.

Lyra2v2 GTX750Ti +185/+185
[2015-12-02 12:00:21] Stratum difficulty set to 4
[2015-12-02 12:00:25] accepted: 41/41 (100.00%), 5176 kH/s (yes!!!)
[2015-12-02 12:00:27] accepted: 42/42 (100.00%), 5175 kH/s (yes!!!)
[2015-12-02 12:00:31] accepted: 43/43 (100.00%), 5175 kH/s (yes!!!)
[2015-12-02 12:00:51] accepted: 44/44 (100.00%), 5175 kH/s (yes!!!)
[2015-12-02 12:01:08] accepted: 45/45 (100.00%), 5175 kH/s (yes!!!)

The sp-modded lyra2v2 on the 750ti is almost as fast as the DJM34 opensource kernal on the 960.

~~No Linux~~

My code will almost turn your 750ti into a gtx 960.

I haven't submitted anything to github in almost 2 weeks now. I am taking a break.

impulse2000

newbie

Activity: 26

Merit: 0

Quote from: bathrobehero on December 02, 2015, 03:25:21 AM

Quote from: thefix on December 02, 2015, 01:06:38 AM