python OpenCL bitcoin miner

[FnuGk]

I just installed the 270.61 nvidia drivers, and it seems to make the drivers crash each time I exit poclbm by clicking the x, if I ctrl-c they seem to exit fine most of the time. Nothing serious yet, but I keep getting these in my system event log:

"Display driver nvlddmkm stopped responding and has successfully recovered."

https://i.imgur.com/VrnJn.png

have you done any overclocking?

[drgr33n]

No. The kernel computes correctly only the 4 first bytes. It's confusing, because there is a code in BitcoinMiner.cl BelowOrEquals() which checks 8 bytes - this produces better assembler for some reason, at least in my setup. It can be replaced with 'if (H == 0)' (but it was slower). That's exactly why the targets are hard coded to difficulty of 1 (00000000 FFFF0000).

If you mean that this:

Code:

#ifdef VECTORS
	if (belowOrEquals(H.x, targetH, G.x, targetG))
	{
		output[OUTPUT_SIZE] = output[nonce.x & OUTPUT_MASK] = nonce.x;
	}
	else if (belowOrEquals(H.y, targetH, G.y, targetG))
	{
		output[OUTPUT_SIZE] = output[nonce.y & OUTPUT_MASK] = nonce.y;
	}
#else
	if (belowOrEquals(H, targetH, G, targetG))
	{
		output[OUTPUT_SIZE] = output[nonce & OUTPUT_MASK] = nonce;
	}
#endif
}

is faster than this:

Code:

#ifdef VECTORS
	if (H.x == 0)
	{
		output[OUTPUT_SIZE] = output[nonce.x & OUTPUT_MASK] = nonce.x;
	}
	else if (H.y == 0)
	{
		output[OUTPUT_SIZE] = output[nonce.y & OUTPUT_MASK] = nonce.y;
	}
#else
	if (H == 0)
	{
		output[OUTPUT_SIZE] = output[nonce & OUTPUT_MASK] = nonce;
	}
#endif

then I got very different results in my tests on Windows 7 x64 & HD5850.

With belowEquals I get ~270Mhps
With the H == 0 version I get ~275Mhps

That's a 1.8% speed increase.

Actually you are right, but in a different way - because of this I should use hard coded kernel target of 00000000 FFFFFFFF in order to not lose 1 thousandth of a percent of all valid difficulty=1 candidates. I'll do this with the next release.

I'd just stop the target from being sent to the GPU and do that H==0 thing. I see no logical reason that it would be slower. There might be illogical reasons, though :-)

Why you check for 'lower' with a 'greater than' operator? Where did the '0x80000000' came from?

I guess I didn't explain that part that well. I chose 0x80000000 to make it easy to determine if bytes 4-7 were actually calculated correctly. I edited BitcoinMiner.cl to only accept hashes below 0x80000000. In the Python code, "> 0x80000000" (should've been >= though) is checking if that part of the hash is over 0x80000000 and warns the user if it is, i.e. if the GPU calculated the hash wrong. But none of that matters since it wasn't even supposed to calculate bytes 4-7 right.

mOmchill,

Is this H==0 mod. going to go into official version of poclbm miner on github  ?

Cheers.

I just tried this and I'm too getting around 1.8% increase in speed.

debian sid x86_64
ATI APP v2.4
11.3 catalyst
pyopencl git
poclbm git
XFX ATI Radeon HD5830 xxx

with the git version I'm getting around 240000
adding the above I'm getting 245000

[djex]

Code:

#ifdef VECTORS
	if (H.x == 0)
	{
		output[OUTPUT_SIZE] = output[nonce.x & OUTPUT_MASK] = nonce.x;
	}
	else if (H.y == 0)
	{
		output[OUTPUT_SIZE] = output[nonce.y & OUTPUT_MASK] = nonce.y;
	}
#else
	if (H == 0)
	{
		output[OUTPUT_SIZE] = output[nonce & OUTPUT_MASK] = nonce;
	}
#endif

I can confirm that this code does give a faster hash rate.

Windows 7 64bit
XFX ATI HD 5850
SDK 2.1
11.3 Catalyst

With the original code I get ~249Mhash/s

With the new code suggested above I get ~251Mhash/s

Update: I noticed that this new code is temperature dependent with my setup. I turned my fan up from 50% to 60% and got 255Mhash/s.

At 50% fan speed I get 74*C
At 60% fan speed I get 66*C

If I use the old code turning the fan up has no effect. I'm not sure why this is.

[nster]

I go from 233 to 238Mh/s with that code, but fan speed has no effect in any way on both codes

[qed]

The if (H == 0) modification fixed my crossfire issue. Now all the GPUs are used at 97%-99%, before i had 1 GPU at 99% one at around 94% and one at around 85%.

[marcus_of_augustus]

The if (H == 0) modification fixed my crossfire issue. Now all the GPUs are used at 97%-99%, before i had 1 GPU at 99% one at around 94% and one at around 85%.

Wow, really? How could that be I wonder?

Did you do anything else besides pasting in the above section of code?

[nster]

hmmm imana try and see if it fixed my too, brb

[nster]

It MAY have lowered the amplitude of the problem, but not by much, it still requires -f 60 for it to be somewhat stable

[qed]

The if (H == 0) modification fixed my crossfire issue. Now all the GPUs are used at 97%-99%, before i had 1 GPU at 99% one at around 94% and one at around 85%.

Wow, really? How could that be I wonder?

Did you do anything else besides pasting in the above section of code?

I just made all the 3 changes posted, that's it. I'm using 3 hd 6950 on windows 7 64 bit.

[LobsterMan]

I just installed the 270.61 nvidia drivers, and it seems to make the drivers crash each time I exit poclbm by clicking the x, if I ctrl-c they seem to exit fine most of the time. Nothing serious yet, but I keep getting these in my system event log:

"Display driver nvlddmkm stopped responding and has successfully recovered."

have you done any overclocking?

No 

[dishwara]

I just installed the 270.61 nvidia drivers, and it seems to make the drivers crash each time I exit poclbm by clicking the x, if I ctrl-c they seem to exit fine most of the time. Nothing serious yet, but I keep getting these in my system event log:

"Display driver nvlddmkm stopped responding and has successfully recovered."

have you done any overclocking?

No 

you using flag -f1 or fxxx?

Completely remove driver using driver sweeper & install again with CUDA.

[kindle]

Just curious did anyone manage to mine a successful block with the (H == 0)  mod?

[nster]

Just curious did anyone manage to mine a successful block with the (H == 0)  mod?

deepbit's pool is seeing the right hashrate so I am being paid in that way :p

[marcus_of_augustus]

The if (H == 0) modification fixed my crossfire issue. Now all the GPUs are used at 97%-99%, before i had 1 GPU at 99% one at around 94% and one at around 85%.

Wow, really? How could that be I wonder?

Did you do anything else besides pasting in the above section of code?

I just made all the 3 changes posted, that's it. I'm using 3 hd 6950 on windows 7 64 bit.

Besides pasting in the
#ifdef VECTORS
   if (H.x == 0)
-------
loop, what other 2 changes did you make exactly?

[nster]

The if (H == 0) modification fixed my crossfire issue. Now all the GPUs are used at 97%-99%, before i had 1 GPU at 99% one at around 94% and one at around 85%.

Wow, really? How could that be I wonder?

Did you do anything else besides pasting in the above section of code?

I just made all the 3 changes posted, that's it. I'm using 3 hd 6950 on windows 7 64 bit.

Besides pasting in the
#ifdef VECTORS
   if (H.x == 0)
-------
loop, what other 2 changes did you make exactly?

H.y == 0
H == 0
 

I believe he didn't paste but just manually change the 3 things (H.x == 0 and the 2 above)

[marcus_of_augustus]

Okay, I tested the H == 0 mod on linux ubuntblow 10.10, HD5970 and it is 1-2% SLOWER. (Wonder why it is faster on Win?)

... poclbm fork for Win users ... ?

[nster]

yup I am on win 7 64 bit

[drgr33n]

Okay, I tested the H == 0 mod on linux ubuntblow 10.10, HD5970 and it is 1-2% SLOWER. (Wonder why it is faster on Win?)

... poclbm fork for Win users ... ?

It's not just windows. Its running around 5000kh/s faster on my debian server.

[djex]

Okay, I tested the H == 0 mod on linux ubuntblow 10.10, HD5970 and it is 1-2% SLOWER. (Wonder why it is faster on Win?)

... poclbm fork for Win users ... ?

I have a Ubuntu 10.10 box and it gets about 1.1Mh/s faster with this mod.

[eleuthria]

I made the modification and my 5870s went from 363 mHash/sec and 361 mHash/sec to 370 even on both.  Ubuntu 10.10.