3% faster mining with phoenix+phatk, diablo, or poclbm for everyone

[jonnynogood]

i dont know what this did but i switched from open CL to phoenix changed a bunch of flags and went from 275Mh to 290MH on each of my 4 6870 cards

thanks!

[DiabloD3]

On normal Radeons at stock clocks and voltages, and are not overheating, have a known error rate (which is something like 1 error per several hundred million instructions)

One error per several hundred million instructions... where are you getting that information from? (an honest question, I'd like to learn more)

A 58xx series, from information I can find, can do between one and four instructions per clock.  If it is clocked at 775000000 cycles per second (775 MHz), that's up to 3.1 billion instructions per second.  So according to your information there would be many errors per second.

I think I meant 1 per several hundred billion. Its in the chip specification somewhere, ask AMD.

a silicon chip is not supposed to have errors, and will only occur if: incorrect voltage, incorrect temperature range, the silicon chip itself is faulty, something extremely rare such as a cosmic ray bouncing off and [in the case of RAM] 'flipping a bit' perhaps once every few weeks

Bzzt, wrong. GPU hardware does not use the same manufacturing process CPU hardware typically does. GPUs are not mission critical hardware, and rare calculation errors are considered acceptable.

You are correct about "incorrect" voltage, however you assume that it is incorrect at all. The professional versions of these cards run at lower clock rates and lower voltages to reduce the error rate. Consumer cards are not ran at incorrect settings, they are merely ran at settings that lead to acceptable levels of errors.

[Turix]

On normal Radeons at stock clocks and voltages, and are not overheating, have a known error rate (which is something like 1 error per several hundred million instructions)

One error per several hundred million instructions... where are you getting that information from? (an honest question, I'd like to learn more)

A 58xx series, from information I can find, can do between one and four instructions per clock.  If it is clocked at 775000000 cycles per second (775 MHz), that's up to 3.1 billion instructions per second.  So according to your information there would be many errors per second.

I think I meant 1 per several hundred billion. Its in the chip specification somewhere, ask AMD.

a silicon chip is not supposed to have errors, and will only occur if: incorrect voltage, incorrect temperature range, the silicon chip itself is faulty, something extremely rare such as a cosmic ray bouncing off and [in the case of RAM] 'flipping a bit' perhaps once every few weeks

Bzzt, wrong. GPU hardware does not use the same manufacturing process CPU hardware typically does. GPUs are not mission critical hardware, and rare calculation errors are considered acceptable.

You are correct about "incorrect" voltage, however you assume that it is incorrect at all. The professional versions of these cards run at lower clock rates and lower voltages to reduce the error rate. Consumer cards are not ran at incorrect settings, they are merely ran at settings that lead to acceptable levels of errors.

This is shown up quite well if you run something like Folding or Seti on your GPU, you will notice a lot more invalid work showing up than you do using the CPU variant and has far as I am aware work for these two does not expire within any reasonable length of time unlike bitcoin so these are actually invalid results and not just "stale".

[teukon]

359 -> 367 (2.2%) on my two 5850's (900/300 1.01V).

Thank you kindly.

[DiabloD3]

On normal Radeons at stock clocks and voltages, and are not overheating, have a known error rate (which is something like 1 error per several hundred million instructions)

One error per several hundred million instructions... where are you getting that information from? (an honest question, I'd like to learn more)

A 58xx series, from information I can find, can do between one and four instructions per clock.  If it is clocked at 775000000 cycles per second (775 MHz), that's up to 3.1 billion instructions per second.  So according to your information there would be many errors per second.

I think I meant 1 per several hundred billion. Its in the chip specification somewhere, ask AMD.

a silicon chip is not supposed to have errors, and will only occur if: incorrect voltage, incorrect temperature range, the silicon chip itself is faulty, something extremely rare such as a cosmic ray bouncing off and [in the case of RAM] 'flipping a bit' perhaps once every few weeks

Bzzt, wrong. GPU hardware does not use the same manufacturing process CPU hardware typically does. GPUs are not mission critical hardware, and rare calculation errors are considered acceptable.

You are correct about "incorrect" voltage, however you assume that it is incorrect at all. The professional versions of these cards run at lower clock rates and lower voltages to reduce the error rate. Consumer cards are not ran at incorrect settings, they are merely ran at settings that lead to acceptable levels of errors.

This is shown up quite well if you run something like Folding or Seti on your GPU, you will notice a lot more invalid work showing up than you do using the CPU variant and has far as I am aware work for these two does not expire within any reasonable length of time unlike bitcoin so these are actually invalid results and not just "stale".

No, its invalid on mining as well. My miner has a HW error counter, it only ticks up when the HW produces a hash it thinks produces H == 0, but when double checked it doesn't.

[Yeti]

410 -> 420 (2.4%) on my HD5870 with the new DiabloMiner! Suh-weeet! Thanks a bunch!

[Fiyasko]

This tweek has been implemented into GUIminer.
So if you use GUIminer.
Just update.

[Alan Lupton]

280 to 289 on a stock speed 5850... will be using this on all my miners, thanks!

exact same here

[getcopy]

4850 no difference... pity... seems it's time to buy a new card...

[rethaw]

This tweek has been implemented into GUIminer.
So if you use GUIminer.
Just update.

Yes. As far as I know this has been implemented into the OpenCL kernel of all the main miners. If you want to implement this just update your miner. You can guarantee the change has been made by opening the kernel file and seeing if the Ma() line has been changed.

[datguywhowanders]

382 to 395, awesome.

[d.james]

This patch successfully pushed the difficulty up by an extra 3%

[Clavulanic]

Got a few mhash on each of my systems, except on my dual 5870 system... Didn't really see a change.

This may just be me or a coincidence, but my single card system showed bigger gains than my multiple card systems. Also my more heavily overclocked systems showed less gains than my less overclocked systems.

Thanks a bunch.

[pandemic]

I was getting 272mh/s with my 5830. After implimenting this change, I'm getting 278 or 279. Not too shabby

[Yannick]

Thanks a ton!

[newMeat1]

This actually slows down my poor little 5670       94==>92
 
What is meant by "This will ONLY WORK if you're running with BFI_INT"? Maybe that's my problem

[DiabloD3]

This actually slows down my poor little 5670       94==>92
 
What is meant by "This will ONLY WORK if you're running with BFI_INT"? Maybe that's my problem

If you're on one, you should be. OTOH, it shouldn't be slowing it down if you dont have BFI_INT enabled

[Diapolo]

All who tried this kernel, the MA-function patch from this thread is included in my modified phatk kernel.
You are able to run it with SDK 2.1, too ... so give it a try .

http://forum.bitcoin.org/index.php?topic=25860.0

YES, this seems to be an ad .

Dia

[dikidera]

This actually slows down my poor little 5670       94==>92
 
What is meant by "This will ONLY WORK if you're running with BFI_INT"? Maybe that's my problem

If you're on one, you should be. OTOH, it shouldn't be slowing it down if you dont have BFI_INT enabled

I thought only 57XX cards support bfi int

[DiabloD3]

This actually slows down my poor little 5670       94==>92
 
What is meant by "This will ONLY WORK if you're running with BFI_INT"? Maybe that's my problem

If you're on one, you should be. OTOH, it shouldn't be slowing it down if you dont have BFI_INT enabled

I thought only 57XX cards support bfi int

lol? All 5xxx and 6xxx do.