CCminer(SP-MOD) Modded GPU kernels.

[sp_]

Someone can answer me why im not able
to run any of last release since 1.58. 59 60
It seems win 10 don't want to run them ...
I have tried to remove it from antivirus, firewall,
and windows smart screen etc ....
If anyone have idea it can help.
What you guyz doin different since release 58?
Thanks !

What is the error message:

you can try to change the register:


-   Start Menu
-   Under Search type: Regedit (Press Enter)
-   Navigate to the following directory:
o   HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Class\{4D36E968-E325-11CE-BFC1-08002BE10318}
-   You will see directories labeled “0000”, “0001” and so on that are relative to the total amount of cards.
-   Under each of those Keys (eg. 0000) create the following two Dword values, once created set their value to 1
o   DisplayLessPolicy
o   LimitVideoPresentSources
-   You can create the Dword by right clicking on the key and NEW  DWORD

https://devtalk.nvidia.com/default/topic/738070/cuda-setup-and-installation/8x-gpu-gtx-issue-under-windows/post/4208565/#4208565

[1714706286]

1714706286

[sp_]

Here's the manual script of the dga patch.
Makefile.am has two problems with the ccminer_SOURCES list.
1. The folllowing line is duplicated:
                          lyra2/lyra2REv2.cu lyra2/cuda_lyra2v2.cu \
replace one of them with:
                          lyra2/lyra2RE.cu lyra2/cuda_lyra2.cu \
2. Add the following line anywhere in the ccminer_SOURCES list except the last line of the list:
                          Algo256/cuda_bmw256.cu  Algo256/cuda_cubehash256.cu \
It compiles but I haven't tried it yet.

I have merged the pullrequest from flipclip. should work now

[chrysophylax]

Here's the manual script of the dga patch.
Makefile.am has two problems with the ccminer_SOURCES list.
1. The folllowing line is duplicated:
                          lyra2/lyra2REv2.cu lyra2/cuda_lyra2v2.cu \
replace one of them with:
                          lyra2/lyra2RE.cu lyra2/cuda_lyra2.cu \
2. Add the following line anywhere in the ccminer_SOURCES list except the last line of the list:
                          Algo256/cuda_bmw256.cu  Algo256/cuda_cubehash256.cu \
It compiles but I haven't tried it yet.

I have merged the pullrequest from flipclip. should work now

compiles now - tanx ...

will test within in a few minutes ... though its all remote - so dont have full accessibility to the miners ...

what are the best parameter settings for lyra2v2 and neoscript on commandline? ...

#crysx

[scryptr]

@CHRYSOPHYLAX--

DJM34, master of all things Vertcoin, suggests the following command lines:

ccminer.exe -a lyra2v2 -o stratum+tcp://pool.verters.com:3333 -u djm34t.user -p password  --api-bind 0  --cpu-priority 2

or, for 750ti:

ccminer.exe -a lyra2v2 -o stratum+tcp://pool.verters.com:3333 -u djm34t.user -p password  --api-bind 0  --cpu-priority 2 -d 750 -i 19

--scryptr

[chrysophylax]

@CHRYSOPHYLAX--

DJM34, master of all things Vertcoin, suggests the following command lines:

ccminer.exe -a lyra2v2 -o stratum+tcp://pool.verters.com:3333 -u djm34t.user -p password  --api-bind 0  --cpu-priority 2

or, for 750ti:

ccminer.exe -a lyra2v2 -o stratum+tcp://pool.verters.com:3333 -u djm34t.user -p password  --api-bind 0  --cpu-priority 2 -d 750 -i 19

--scryptr

you guys are awesome ...

it will be for the 750ti cards ... its all i have in the nvidia side of the farm now ...

when the donation links go live - the devs should expect to see some donation mining happening from our end on their links shortly after ... cant wait to get back into the office to get them all active ...

is there any special settings for neoscrypt that needs to go in? ...

tanx much for the help once again scryptr ...

#crysx

[pallas]

DONATION FOR GOOD WORK--

DJM34, Pallas, I flipped you each a nickle.  The overall improvement on both Lyra2v2, and on Neoscrypt, was excellent!!!

As soon as I can compile SP_ 's release dot 60 on Linux, he's up for his reward, too.  The teamwork and co-operation is really good for ccminer.  Everyone should donate....

Thanks!


--scryptr

That's why I started working on nvidia: there is community work. Once there was on AMD as well.
Thanks for the beers! :-)

[sp_]

As soon as I can compile SP_ 's release dot 60 on Linux, he's up for his reward, too.  The teamwork and co-operation is really good for ccminer.  Everyone should donate....
--scryptr

Get lates on git.
It builds now.

Beers would be great

[sp_]

At the current difficulty 1 single 750ti will earn around $1.35 per 24h with the new vertcoin algorithm.

With 100 ti's. $135 profit for 1 day of mining!

quark @ nicehash is only paying $0,394968 a day per 750ti.

Vertcoin is 340% more profitable than quark...

Go get those coins, and donate some to the hard working developers..

Happy mining!

[fenomenhaa]

At the current difficulty 1 single 750ti will earn around $1.35 per 24h with the new vertcoin algorithm.

With 100 ti's. $135 profit for 1 day of mining!

quark @ nicehash is only paying $0,394968 a day per 750ti.

Vertcoin is 340% more profitable than quark...

Go get those coins, and donate some to the hard working developers..

Happy mining!

Beers have been sent to Sp, Pallas, djm34 both:))) . Thanks's for your co-operation, .About pools, i cannot connect to givemecoins,ffpool and nicehash pools, coinotron shows the half hash, vertminers the only pool i mining without any problems right now..

[pallas]

I've been mining on give me coins successfully for some hours, but only on nvidia: on amd it is mostly rejected shares.

[fenomenhaa]

I've been mining on give me coins successfully for some hours, but only on nvidia: on amd it is mostly rejected shares.

I just tried with nvidia ,haven't been tried with amd's yet ).Give mecoins don't work on my side.changed ip and so on,strange ftc pool worked but vert pool don't wanna ) I sent a support request ,waitinig for an answer from pool..
@pallas what's the hashrate of 280x? And which miner for the new algo?

[pallas]

I've been mining on give me coins successfully for some hours, but only on nvidia: on amd it is mostly rejected shares.

I just tried with nvidia ,haven't been tried with amd's yet ).Give mecoins don't work on my side.changed ip and so on,strange ftc pool worked but vert pool don't wanna ) I sent a support request ,waitinig for an answer from pool..
@pallas what's the hashrate of 280x? And which miner for the new algo?

I didn't try the latest git commits of djm34 to see if they fix cards < 290 yet.
Hashrate should be 5-6 Mh/s.
I don't know of any other miner for lyra2rev2.

[djm34]

I've been mining on give me coins successfully for some hours, but only on nvidia: on amd it is mostly rejected shares.

I just tried with nvidia ,haven't been tried with amd's yet ).Give mecoins don't work on my side.changed ip and so on,strange ftc pool worked but vert pool don't wanna ) I sent a support request ,waitinig for an answer from pool..
@pallas what's the hashrate of 280x? And which miner for the new algo?

I didn't try the latest git commits of djm34 to see if they fix cards < 290 yet.
Hashrate should be 5-6 Mh/s.
I don't know of any other miner for lyra2rev2.

don't know either... (there is nothing "strange" in bmw256 except a few functions which were declared as static, which I removed, so I don't expect much change quite frankly... trying to fix something which works on my computer isn't easy...)

[sambiohazard]

Hi all, was just trying sp 60 release with new VTC algo

I get a solid 4160 Khash/s on 750ti with config posted by scryptr, but on 970 its gyrating wildly from 9200 khash/s to 6900 khash/s with scryptr config. power & gpu usage is very choppy on both cards.



EDIT: 970 has now settled into nice hashrate between 8900-9200 khash/s, but power & gpu usage is still choppy as hell. will this affect durability/life of the card?

[djm34]

Hi all, was just trying sp 60 release with new VTC algo

I get a solid 4160 Khash/s on 750ti with config posted by scryptr, but on 970 its gyrating wildly from 9200 khash/s to 6900 khash/s with scryptr config. power & gpu usage is very choppy on both cards.

use my version https://github.com/djm34/ccminer-lyra/releases (which actually is the official release  )

(it is probably related to either work_time somehow or rejection of dup or stuff like that, sp/tpruvot tends to use some crazy stuff there...)

[sambiohazard]

Hi all, was just trying sp 60 release with new VTC algo

I get a solid 4160 Khash/s on 750ti with config posted by scryptr, but on 970 its gyrating wildly from 9200 khash/s to 6900 khash/s with scryptr config. power & gpu usage is very choppy on both cards.

use my version https://github.com/djm34/ccminer-lyra/releases (which actually is the official release  )

(it is probably related to either work_time somehow or rejection of dup or stuff like that, sp/tpruvot tends to use some crazy stuff there...)

Are you talking about gpu usage as hashrate is now stable?

[sp_]

Hi all, was just trying sp 60 release with new VTC algo

I get a solid 4160 Khash/s on 750ti with config posted by scryptr, but on 970 its gyrating wildly from 9200 khash/s to 6900 khash/s with scryptr config. power & gpu usage is very choppy on both cards.



EDIT: 970 has now settled into nice hashrate between 8900-9200 khash/s, but power & gpu usage is still choppy as hell. will this affect durability/life of the card?

By looking at the graph it look's like the miner can be optimized.

You GPU will downclock if it goes over 70 degrees celcius.(trottle) I think this is what's happening here.. Is the hashrate stable in the first 100 accepts? (if the card is cold).


try to change the p-state in Nvidia inspector to p0.

[sambiohazard]

Hi all, was just trying sp 60 release with new VTC algo

I get a solid 4160 Khash/s on 750ti with config posted by scryptr, but on 970 its gyrating wildly from 9200 khash/s to 6900 khash/s with scryptr config. power & gpu usage is very choppy on both cards.


EDIT: 970 has now settled into nice hashrate between 8900-9200 khash/s, but power & gpu usage is still choppy as hell. will this affect durability/life of the card?

By looking at the graph it look's like the miner can be optimized.

You GPU will downclock if it goes over 70 degrees celcius.(trottle) I think this is what's happening here.. Is the hashrate stable in the first 100 accepts? (if the card is cold).


try to change the p-state in Nvidia inspector to p0.

I guess you are right about that but its stable now at 9100-9200 khash/s but gpu & power usage is still choppy. I will download & try nvidia inspector.

EDIT: Although i had no problems running other algos like scrypt-jane at 74-75 C and it wont throttle. I have seen that choppy usage come only with ccminer. on cudaminer i never have this problem.

[djm34]

Hi all, was just trying sp 60 release with new VTC algo

I get a solid 4160 Khash/s on 750ti with config posted by scryptr, but on 970 its gyrating wildly from 9200 khash/s to 6900 khash/s with scryptr config. power & gpu usage is very choppy on both cards.



EDIT: 970 has now settled into nice hashrate between 8900-9200 khash/s, but power & gpu usage is still choppy as hell. will this affect durability/life of the card?

By looking at the graph it look's like the miner can be optimized.

You GPU will downclock if it goes over 70 degrees celcius.(trottle) I think this is what's happening here.. Is the hashrate stable in the first 100 accepts? (if the card is cold).


try to change the p-state in Nvidia inspector to p0.

70° for a 970 that's kinda low... (well my 980 are running usually around 70) 68°C with this algo for the 900 serie temp limit should be around nvidia limit (82°C even though I never saw my card going that high, except when I forgot to turn on msi and its fan custom setting).

However this pattern isn't associated with throttling which has no effect on gpu usage and just decreases clock speed to match tdp requirement.  This is clearly related to ccminer asking for new block header all the time... (check dup or crap like that... which isn't useful at all in most case)

this is how it should look like:

[sambiohazard]

Is v 1.9.7.3 latest version for nvidia inspector? Any links?