Cgminer default settings are --worksize 128 and --vectors 2 so you get those free of charge
You might want to see what --worksize 64 or --worksize 256 does for your hash rate.

The stale rate of p2pool isn't comparable to other pools as p2pool uses a much higher difficulty for its shares...
Due to high value or each high-difficulty share, I recommend you use --submit-stale when mining on p2pool.
This will make cgminer submit every share it finds, even if it thinks it is already stale. This won't make a huge difference but might net you an additional share every now and then.

Intensity is a fine tuning parameter and its optimal value is dependent on the card's hash rate.
Intensity influences the size of each uninterruptible batch of calculations the GPU performs - high intensity batches of work are larger, thus taking longer to finish.
This results in occasional situations where the GPU takes a couple of seconds too long and produces a valid but already stale result wasting the computation time.
Try lowering your intensity to 8 - if the hash rate doesn't budge you made the right call and your card didn't benefit from the higher intensity.

My general observations have been that best results are achieved (using the default two threads per gpu):
  + with <200MHash/s cards: intensity 7
  + [200..400] MHash/s : intensity 8
  + >400 MHash.s : intensity 9
Intensities beyond 9 are meant only for the new 7xxx family of cards and won't do any good when used with the old, slower cards.

That!! An option designed especially for programs writing to the same file descriptor.
I totally missed it...
Instead of being moved to logfile.1, the log file will be copied over as logfile.1 and the original file will be truncated to 0 bytes.
Good find.

Did you try sudo aticonfig --initial --adapter=all?
This will re-create the X.org server configuration file (/etc/X11/xorg.conf). You might want to make a manual backup before issuing the command.

Did you execute aticonfig --adapter=all -- odgt as root?

Whatever do you mean by "it (SSH?) lost your password"

Just launch the executable from the command line terminal appending the -rescan parameter at the end.

Atta way Doc: 4.63% and 2.52% BTC blocks found! The fast-block cheat has been activated!
Just keep'em coming...

::EDIT:: use COPYTRUNCATE, as recommended by johnyj

Done! Thanks again johnyj.

Not without killing cgminer:

Another approach (more complex but avoiding the necessity of periodical cgminer restarts) is, instead of logging cgminer's output directly to the target file, run the logging through an intermediary that will follow the file name (eg. a program using fopen, fflush, and fclose for each write to the target logfile).
Piping standard output and errors with |& is your friend here, example: ls asdasdasdasd |& grep asd > ls.log


Cgminer keeps writing to the same FILE DESCRIPTOR completely ignoring the name change.

A CPU is orders of magnitude slower at mining than a lowliest hd5000 series card.

You probably shouldn't mine until you get that PSU replaced.
Never skimp on a PSU, it's the component that makes or breaks a mining rig.

If you choose to proceed with CPU mining, my recommendations are

Hmmm... that quotation of mine does seem a bit botched - let's set the scope of my agreeing with Jine from the phrase "damaging the community" onward, shall we?
By damaging the community I understand feeding of those poor, unwashed and uninformed fools still mining 24/7 at prop pools  
Like leaving the car unlocked and with keys in the ignition in a bad neighborhood...

I do agree that ABC never stole a dime from their miners, I never implied otherwise.
Let's try to be compiler-grade clear and precise here

Except you seem to have hit the reverse
Try not to post information you're not positive on, ok? Let's not pollute the forum any more than it already is.


OP, a good 750W PSU should be considered a minimum for mining with those three cards.
The 7970 itself can chew up 250 W. The 4870 is not exactly power efficient, neither is the Bulldozer CPU.

A 1000W unit would be a prudent choice if you want some flexibility with the cards - the Kingwin Lazer Platinum 1000W is 80+ Platinum certified (you can't get a more energy efficient unit) and can be had for about $180.

I take it that's not a dedicated mining rig, is it?
If it is, you should disable four cores in the BIOS as you won't be using them anyway.

Using just one GPU (e.g. the 7970) run the OCCT tool and upload the graphs it creates, that will shed some light on the voltages and thermals.
Start with a 30-minute CPU test. If the machine doesn't crash, try the GPU test. Re-try those tests for a prolonged time (like an hour).
If the machine is still stable, try the PSU test (it runs the CPU and GPU tests combined). An inadequate PSU will surely fail here.

I see where you're coming from, Kano, but please do notice ABC have repeatedly lied and misled their miners about the nature of their pool^Wproxy:
"Using the forum accounts associated with ABCPool for communication with upstream pools could confuse our customers and potentially give away details of our business model"?? WTF?
They used a sock-puppet profile to hide their presence in the topics concerning the pools they have been abusing.
It took a moderator to update the topic title with a warning and break the circle of lies.
If you don't mind your pool being dishonest with you by all means stay with them - just don't be surprised if some other, more serious, matter surfaces where they have been lying to you as well.


Yes, except the pool op is well known for using his miners' hash rate in a very cavalier fashion and denying the fact afterwards.
The only reason I migrated my miners from Eligius was your dishonesty about the whole matter, Luke - if you break the trust so easily I don't want to have anything to do with your pool.

Perhaps it's just me but I don't like being lied to.
Indeed, P2pool looks better by the hour...

Rassah, 270 MHash/s seems a tad low. May I ask what work size and vector width you are using? Are you running the default 2 threads per GPU?

Another fine-tuning suggestion: a 270 MHash/s card might be better off with intensity 8, the stale rate should drop a bit provided your pool server is fast enough.
Right now, I'm at 0,21% stales at Bitminter which is rather high - usually my stale rate is about 0,1%.

That.

Deepceleron is correct.
The 3.3V is a secondary rail used mostly by the system RAM and various motherboard components.
Kettenmonster is either horribly misinformed or trolling.

Not talking about CPU priority. I'm talking about IO priority which was a newly introduced feature in Vista.

I already voted in that poll

PPLNS is a decent system depending on the exact implementation, I'm sure you have acquainted yourself with Meni Rosenfeld's analysis of pool payout schemes.

ABCProxy deserve every bit of heat they are getting for hiding the true nature of their operation from their miners.

Best of luck to you, Jim.

What you say is true.
However, since I can't undervolt that card to achieve superior Mhash/W my next best bet is to push it as hard as possible to offset the (very significant in the case of this machine) system power usage: 8 TB of storage is quite power hungry, as is a four-core Phenom II.
Power usage rises linearly with core clock so I'm ok here, except for the temperature which isn't THAT bad. 52% fan speed is relatively quiet.

The idea of upgrading the server with mining GPUs was that the machine would pay for its energy footprint.
What I should really do, is change the case to something better ventilated, like the rv-02 I'm using for my dedicated rigs.

I take it you're running your DCIIs in a caseless rig?

If Bitlc changed their payout scheme to a non-hoppable one they would not need to take that stand...

So?
Did Simplecoin say hashkill was a live project?
Download the archive and check the last modified date of the file inside: modified 2011-09-14 20:46 (x64 version)

Doesn't Simplecoin also say "Advanced Windows & Linux (for best mining results): Download cgminer HERE"?

Mind you that the mining kernels have to be redone to accommodate the new GCN architecture. While the ancient miners should work, they won't be able to achieve high hash rates due to being optimized for VLIW4/5.

The latest 12.x driver with 2.6 SDK.
AMD drivers are apparently still being optimized to better cooperate with the GCN architecture.