Author

Topic: [ANN] cudaMiner & ccMiner CUDA based mining applications [Windows/Linux/MacOSX] - page 633. (Read 3426922 times)

hero member
Activity: 676
Merit: 500
where did you saw these power usage reading in hwinfo please?
legendary
Activity: 1400
Merit: 1050
I am running on script at the moment (solomining spartans) hwinfo says 188W+25W (no idea what the second number is...)
the 780ti is running at 95% tdp (temp limited... the temps are bad with skynet mod bios).
At the moment, on power consumption, the 290x wins...

But comparing to the 750ti is a bit unfair, in the sense that it is new generation against old generation.
full member
Activity: 182
Merit: 100
http://cryptomining-blog.com/1801-new-higher-performance-cgminer-3-7-3-for-heavycoin-gpu-mining/

Looks like AMD card got a boost for HVC.  Too bad the 290x power usage cant compare with the efficiency of a 750 Ti!
I didn't monitor so far the TDP of the amd (can't in MSI AB just realise I can do that with HWinfo), but with a card running at 60°C it shouldn't be high.
For some reason, everybody seems to think that the R9 290x is a very power hungry card, but it is rated 250W.
Clearly, past generation amd may have been huge electron suckers but that doesn't seem to be the case for this one
Pretty sure it can peak to 300W, and that's without overclock...
But yes, my GTX 780 has a 250W rating too...
I thought AMD has a different way to rate watts however...
sr. member
Activity: 280
Merit: 250
http://cryptomining-blog.com/1801-new-higher-performance-cgminer-3-7-3-for-heavycoin-gpu-mining/

Looks like AMD card got a boost for HVC.  Too bad the 290x power usage cant compare with the efficiency of a 750 Ti!
I didn't monitor so far the TDP of the amd (can't in MSI AB just realise I can do that with HWinfo), but with a card running at 60°C it shouldn't be high.
For some reason, everybody seems to think that the R9 290x is a very power hungry card, but it is rated 250W.
Clearly, past generation amd may have been huge electron suckers but that doesn't seem to be the case for this one


It's still one of the most power hungry cards there is


http://www.hardocp.com/article/2013/10/23/amd_radeon_r9_290x_video_card_review/15
legendary
Activity: 1400
Merit: 1050
http://cryptomining-blog.com/1801-new-higher-performance-cgminer-3-7-3-for-heavycoin-gpu-mining/

Looks like AMD card got a boost for HVC.  Too bad the 290x power usage cant compare with the efficiency of a 750 Ti!
I didn't monitor so far the TDP of the amd (can't in MSI AB just realise I can do that with HWinfo), but with a card running at 60°C it shouldn't be high.
For some reason, everybody seems to think that the R9 290x is a very power hungry card, but it is rated 250W.
Clearly, past generation amd may have been huge electron suckers but that doesn't seem to be the case for this one
newbie
Activity: 19
Merit: 0
http://cryptomining-blog.com/1801-new-higher-performance-cgminer-3-7-3-for-heavycoin-gpu-mining/

Looks like AMD card got a boost for HVC.  Too bad the 290x power usage cant compare with the efficiency of a 750 Ti!
newbie
Activity: 17
Merit: 0
Does anyone have a usable batch file for gtx 780 classy mining memory coin? Just can't get mine to work...
There are 500+ pages of good advice... unless you are more precise in what you intend to do, we can't help you...

Sorry, but can you tell me what information you need? I'm new to this..
at least what you want to mine, the configuration varies depending on what you run. (there is also a readme file which might help)
I use Z12x24 to mine at 1gh.com for memory coin. The gtx 780 was overclocked +135 MHz. I use algo=scrypt jane , no CPU parallel. This is all I can remember at the moment, I'll check later.
what is the Nfactor ? (it should be displayed when you start cudaminer).
What you can do, it to run the autotune, it works rather well with scrypt-jane.
If the Nfactor is large, it could also helpful to increase large lookup gap value (-L)
Thank you!forgot to mention, the error cudaminer gave me was "json decode fail(1):"[" or "{" expected near "<"
legendary
Activity: 1400
Merit: 1050
Does anyone have a usable batch file for gtx 780 classy mining memory coin? Just can't get mine to work...
There are 500+ pages of good advice... unless you are more precise in what you intend to do, we can't help you...

Sorry, but can you tell me what information you need? I'm new to this..
at least what you want to mine, the configuration varies depending on what you run. (there is also a readme file which might help)
I use Z12x24 to mine at 1gh.com for memory coin. The gtx 780 was overclocked +135 MHz. I use algo=scrypt jane , no CPU parallel. This is all I can remember at the moment, I'll check later.
what is the Nfactor ? (it should be displayed when you start cudaminer).
What you can do, it to run the autotune, it works rather well with scrypt-jane.
If the Nfactor is large, it could also helpful to increase large lookup gap value (-L)
newbie
Activity: 17
Merit: 0
Does anyone have a usable batch file for gtx 780 classy mining memory coin? Just can't get mine to work...
There are 500+ pages of good advice... unless you are more precise in what you intend to do, we can't help you...

Sorry, but can you tell me what information you need? I'm new to this..
at least what you want to mine, the configuration varies depending on what you run. (there is also a readme file which might help)
I use Z12x24 to mine at 1gh.com for memory coin. The gtx 780 was overclocked +135 MHz. I use algo=scrypt jane , no CPU parallel. This is all I can remember at the moment, I'll check later.
legendary
Activity: 1400
Merit: 1050
Does anyone have a usable batch file for gtx 780 classy mining memory coin? Just can't get mine to work...
There are 500+ pages of good advice... unless you are more precise in what you intend to do, we can't help you...

Sorry, but can you tell me what information you need? I'm new to this..
at least what you want to mine, the configuration varies depending on what you run. (there is also a readme file which might help)
hero member
Activity: 756
Merit: 502

hmm..is that the value in cuda_groestlcoin.cu:

const int threadsperblock = 512;

? or do i need to change it somewhere else?

We will be looking into this ourselves. I don't really want to spoon feed code modifications to users here.
Feel free to modify yourself, but you HAVE to know CUDA and watch out for the shared memory initialization
at the start of the kernel code (don't run into this initialization code with more than 256 threads)

The next release will have the change, provided that it improves hash rates.

Christian

legendary
Activity: 1400
Merit: 1050
Did I already mention this coin ? Nigeria Coin (skein algo) launched few day ago

https://bitcointalksearch.org/topic/ann090ngc-nigeriacoin-skein-coin-distribution-symmetry-mar-26-520100


Nigeria, really?

Code:

Dear CUDA Miner user,

I am Mr. paul agabi, a Lawyer by profession. I am the personal attorney to Mr. Charles (my surname) ,
a national of your country, who used to work with Chevron Oil Exploration Company in Nigeria, herein
after shall be referred to as my client.

On the 21st of April 2000, my client, his wife and their onlye child were involved in a car accident along
Lagos - Ibadan express road. All occupants of the vehicle unfortunately lost there lives.

Since then I have made several enquiries to your embassy to locate any of my clients extended relatives.
This has proved unsuccessful. I came to know about you through an enquiry I was making in the internet,
and I found out you shared the same surname with my client, which is why I have decided to contact you,
in order to assist in repatriating the money and property left behind by my client before they get confiscated
or declared unserviceable by the bank.

Particularly, in a local commercial bank here where the deceased had an account valued at about 15.5
Million Nigeria Coins, has issued me a notice to provide the next of kin or have the account confisticated
within the next ten official working days.

Since i have been unsuccessful in locating the relatives for over 2 years now I seek your consent to present
you as the next of kin of the deceased since you have the same last name so that the proceeds of this
account valued at 15.5 Million Nigeria Coins can be paid to you and then you and me can share the money,
on the ratio of 50% for me, and 50% for you.

I have all necessary legal documents that can be used to back up any claim we may make. All I require is your
honest co-operation to enable us see this business through. I guarantee that this will be executed under a
legitimate arrangement that will protect you from any breach of the law.

Please get in touch with me by email and send to me your telephone and fax numbers to enable us discuss
further about the details of this transaction.

Best regards, Mr. paul agabi.


Grin that what I thought too at first. But the launch went well with a dev. who seemed to know what he was doing and had some communication skills.
legendary
Activity: 914
Merit: 1001
5 x eVGA 750Ti SC non overclocked = 61MH

but I'm unsure how to explicitly compile it for sm_35 and 128 regcount this time (due to the new Makefile.am). Everytime I get invalid calculations for the shares. But I think compiling with standard settings is a bit slower on groestlcoin atm. Any advice?

on compute 3.0 and above we're using a block size of 768 which cannot work with more than 80 regs per thread.
it's the reason the 750Ti jumped from 9 MHash to 11 MHash... just like that Wink

ah, thanks for clarification. just two more questions:

1. under linux, after compiling all compute settings are within one executable? how can I make sure, that sm_35 is used?
2. is there a possibility to change regcount for a specific algo (in my case groestl) while having it untouched for all the others?

unser Linux it uses the compute setting matching your card best.

add a file specific rule to Makefile.am vor groestlcoin kernel. Nur IMHO you should try upping the groestl Thread block size to 768 first...

hmm..is that the value in cuda_groestlcoin.cu:

const int threadsperblock = 512;

? or do i need to change it somewhere else?
full member
Activity: 263
Merit: 100
Temporarily I created an app to auto-close any ccminer*.exe dialog window pop-ups, then the batchfile continues to auto-restarts the mining.

Last night it crashes twice for about 5hrs difference.

But too bad I can't do anything about the GPU idles (SICK?).

When a card goes idle or errors out for me usually just restarting the miner will not bring the card back up to full speed (usually just 10Khash/s or something like that).

You can reset the cards from the command line (using the DevCon [Developers Console] command available at http://support.microsoft.com/kb/311272) by disabling and enabling them again without restarting windows and it seems to always fix the problem for me.

first run 'DevCon find *' to list out all your hardware, and then find a unique part of the card ID to add to the command.

In my case '*dev_1380' identifies all my MSI 750 ti cards at once.

Quick and easy to add to your .bat loop.

devcon disable *dev_1380
devcon enable *dev_1380

Hope that helps!
Sound interesting.
Does it work with windows 8 ?

Have not tried it... I am on Win7 x64 for my mining rig, but I don't see why it wouldn't unless it is some obscure change. It is just a download .exe though you throw in your windows32 folder and run, so not hard at all for anyone to try it out.
There so many things on my computer, so if I can know if it runs before downloading it, it is always better.
And actually, it does not work.

Work for me. win 8.1 x64

750Ti -> *dev_1380
660Ti -> *dev_1183

ex.: devcon disable *dev_1380 (disable ALL 750Ti on your board)

newbie
Activity: 17
Merit: 0
Does anyone have a usable batch file for gtx 780 classy mining memory coin? Just can't get mine to work...
There are 500+ pages of good advice... unless you are more precise in what you intend to do, we can't help you...

Sorry, but can you tell me what information you need? I'm new to this..
hero member
Activity: 676
Merit: 500
The problem is that if i run 3 different for 3 different cards it is 30-40% cpu , but if i start 2 for the same two cards then the cpu goes near 100%. And it dual core cleron which is a good cpu.
hero member
Activity: 756
Merit: 502
Did I already mention this coin ? Nigeria Coin (skein algo) launched few day ago

https://bitcointalksearch.org/topic/ann090ngc-nigeriacoin-skein-coin-distribution-symmetry-mar-26-520100


Nigeria, really?

Code:

Dear CUDA Miner user,

I am Mr. paul agabi, a Lawyer by profession. I am the personal attorney to Mr. Charles (my surname) ,
a national of your country, who used to work with Chevron Oil Exploration Company in Nigeria, herein
after shall be referred to as my client.

On the 21st of April 2000, my client, his wife and their onlye child were involved in a car accident along
Lagos - Ibadan express road. All occupants of the vehicle unfortunately lost there lives.

Since then I have made several enquiries to your embassy to locate any of my clients extended relatives.
This has proved unsuccessful. I came to know about you through an enquiry I was making in the internet,
and I found out you shared the same surname with my client, which is why I have decided to contact you,
in order to assist in repatriating the money and property left behind by my client before they get confiscated
or declared unserviceable by the bank.

Particularly, in a local commercial bank here where the deceased had an account valued at about 15.5
Million Nigeria Coins, has issued me a notice to provide the next of kin or have the account confisticated
within the next ten official working days.

Since i have been unsuccessful in locating the relatives for over 2 years now I seek your consent to present
you as the next of kin of the deceased since you have the same last name so that the proceeds of this
account valued at 15.5 Million Nigeria Coins can be paid to you and then you and me can share the money,
on the ratio of 50% for me, and 50% for you.

I have all necessary legal documents that can be used to back up any claim we may make. All I require is your
honest co-operation to enable us see this business through. I guarantee that this will be executed under a
legitimate arrangement that will protect you from any breach of the law.

Please get in touch with me by email and send to me your telephone and fax numbers to enable us discuss
further about the details of this transaction.

Best regards, Mr. paul agabi.

sr. member
Activity: 350
Merit: 250
Personally i would say upgrade your cpu. If it cant handle 6 instances then i would definitely upgrade it
hero member
Activity: 676
Merit: 500
Christian
Is there a logic explanation for the fenomena.
When using risers i have 6 cards. If i open one ccminer or cudaminer for all cards , for example heavycoin, i get 55mhs for all, if i open 3 instances - one for gpu 0.1 , one for gpu 2.3 and one for 4.5 gpus. Then i get 22.5mhs each , which makes 67mhs which is 10mhs more than one ccminer for all , and if i open 2 instances for each card (when i was using 3 card on my i7 machine) i get  the same or even better results than the gpu that is directly on the pci-e x16 slot. Is there a way to achive this performance , but using only one instance of ccminer or cudaminer? Thank you in advance.

ccminer uses the thrust library for heavycoin, and thrust has some overhead. It performs memory allocations at run time, and during that time the GPU can't mine. By running two instances you would drive up the GPU utilization from e.g. 80% to 95%, getting higher throughput.

you could to the same passing the -d 0,0 flag to ccminer, however there is currently a problem where both of these GPU threads operate in the same CUDA context, step on each other's toes and hence produce invalid shares. a CCMiner v0.6 might fix that.

Christian



Thank you for your answer.
I tried that but the result is not better , for instance cudaminer on 750 ti -- running scrypt using -d 0,0 i got 135khs - 137khs about 270 for the card, but when i run 2 cmd of cudaminer for -d 0 i get 147khs -145 khs which makes 290khs for the card . and if i put the card directly on the pci-e slot it gives me 290khs. That is the way i overcome the pci-e riser khs drop issue , but on my 6 cards rig my celeron G1820 can`t cope with 12 cudaminer or even with 6 cudaminer windows. So if there is another way to overcome this issue will be superd.
 As for ccminer , yes when i put -d 0,0 sometimes work sometimes it gives the cpu validate error , and again the result is not better.
If i can help for resolving this i will be glad. Grin
Jump to: