python OpenCL bitcoin miner

[OVerLoRDI]

I'm getting 340000 khash/s on each of my 6970s.  Running Stream SDK 2.3... I will see about downgrading in the near future since I have read 2.2 is faster.

My main question is each of the instances of poclbm is maxing out a core.  I have noticed if I max out all 4 cores with prime95, my khash/s does not increase, so it seems the miner is just wasting cpu cycles?  Is there something I can do to address this or is it a problem with the program?

Win 7 64bit
Cats 11.1a
AMD Stream 2.3
Dual 6970s

[Cablesaurus]

Can anyone confirm they're still finding blocks solo mining with this miner? Just a sanity check, as I haven't found anything solo mining in quite some time @ 625,000khash/s. I've found no blocks with the current version at all as of yet.

[Cryptoman]

I upgraded about 2 Ghash/s worth of miners a couple days ago with the latest version and haven't found anything since either.  Maybe it's just a spell of bad luck.  I'll update the thread if anything manifests.

[Cablesaurus]

I am by no means qualified in the ins and outs of mining programming for Bitcoin, but empirical observation really seems to me like something is vastly less efficient and/or broke from 2 or 3 versions prior to the current. Any discussion from m0mchil or other miners would please be welcome.

[m0mchil]

I'm getting 340000 khash/s on each of my 6970s.  Running Stream SDK 2.3... I will see about downgrading in the near future since I have read 2.2 is faster.

My main question is each of the instances of poclbm is maxing out a core.  I have noticed if I max out all 4 cores with prime95, my khash/s does not increase, so it seems the miner is just wasting cpu cycles?  Is there something I can do to address this or is it a problem with the program?

Win 7 64bit
Cats 11.1a
AMD Stream 2.3
Dual 6970s

I personally run it on Win 7 64bit, but with cat 10.12 and Stream SDK 2.2 and CPU utilization is pretty low, actually close to 0% Could you please try with Stream 2.2 and post results?

Can anyone confirm they're still finding blocks solo mining with this miner?

There was a version which had a problem, generally not calculating correct hashes above first 32 bits. But with 20110204 I was able to find blocks on test net (which had significantly higher than 32 bits difficulty by the time).

Do you at least get 'invalid or stale' message? This should show up from time to time even if miner is broken. Bitcoin will reject the solution. But I really can't imagine other reason for not seeing anything, not even 'invalid' message - the one and only reason to not see any message for a long time is difficulty.

[aistto]

Hello, I currently have 2x6950 in crossfire, i7 2600k, 11.1a drivers, ATi Stream 2.3.

When I try to run it with the following parems:
poclbm.exe --user=x --pass=y --host=http://mining.bitcoin.cz:8332 -d 0

http://

•      Cayman
[1]     Cayman
[2]            Intel(R) Core(TM) i7-2600K CPU @ 3.40GHz

It gives me this error:

InvalidURL: nonnumeric port: ''
Unexpected error:
Traceback (most recent call last):
  File "BitcoinMiner.pyc", line 171, in mine
  File "BitcoinMiner.pyc", line 157, in getwork
  File "jsonrpc\proxy.pyc", line 42, in __call__
  File "urllib.pyc", line 89, in urlopen
  File "urllib.pyc", line 208, in open
  File "urllib.pyc", line 334, in open_http
  File "httplib.pyc", line 1008, in __init__
  File "httplib.pyc", line 656, in __init__
  File "httplib.pyc", line 673, in _set_hostport

Any help? Thanks!

Hi,
get rid of "http://:8332" from the host line, and put the port number after the "--port=" switch.

Same problem:( i have winXP sp3, ati catalyst 11.1, ati stream sdk 2.3, hd4830

my promt

poclbm.exe -u USER --pass=PASS --host=http://mining.bitcoin.cz --port=8332 -d 0
Unexpected error:
Traceback (most recent call last):

 Compilation terminated.
 File "BitcoinMiner.pyc", line 175, in mine
  File "BitcoinMiner.pyc", line 161, in getwork
  File "jsonrpc\proxy.pyc", line 42, in __call__
  File "urllib.pyc", line 89, in urlopen
  File "urllib.pyc", line 208, in open
  File "urllib.pyc", line 334, in open_http
  File "httplib.pyc", line 1008, in __init__
  File "httplib.pyc", line 656, in __init__
  File "httplib.pyc", line 673, in _set_hostport
InvalidURL: nonnumeric port: ''

[Tolsi]

My best settings:
Windows 7 x64, Stream 2.3, AMD HD6870 (970/1100)
with "-v -w 128 -f 60" - ~247 khash/s
with "-v -w 128 -f 0" - ~250 khash/s, desktop is lagging

[kw71]

Is anyone else having trouble, with not finding any blocks at all with the last 2 versions of this miner post 2/1/11?

I've mined for well over 5 or 6 days with nothing and I'm running about 630khash between the 2 cores on my 5970. I'm just using -v -w 128 -f 120........... I haven't found a block in what seems like far too long.

It's only working with low difficulty like testnet or mining pool work.

Sabotaged?

[Quantumboredom]

I've gotten successful generations using version 20110204. Yesterday and three days ago.

This is on Win 7 with HD 5870 using catalyst 11.1 (SDK 2.3)

[ribuck]

I've had one generation with 20110204 on Fedora 14 Linux, with a 5870 card on SDK 2.1.

[Cablesaurus]

There was a version which had a problem, generally not calculating correct hashes above first 32 bits. But with 20110204 I was able to find blocks on test net (which had significantly higher than 32 bits difficulty by the time).

Do you at least get 'invalid or stale' message? This should show up from time to time even if miner is broken. Bitcoin will reject the solution. But I really can't imagine other reason for not seeing anything, not even 'invalid' message - the one and only reason to not see any message for a long time is difficulty.

I haven't ran it 24/7 due to workstation limitations, Ihave to reboot at least once a day.

It's just that I ran for like 5 or 6 days for like 22 out of 24h in the day and got nothing.

I'm using Win7 64 with latest Cats and Stream 2.3.

It's not so much that I'm complaining or surprised. I just find it REALLY odd that the first day I mined Bitcoin on like 1/22 or so, I found 4 blocks. Then after that, (once a version was upgraded) I found nothing...

Is there a way you can link the version from 2 or 3 revisions ago I can test with? It's the one from late Janurary

[ribuck]

I just find it REALLY odd that the first day I mined Bitcoin on like 1/22 or so, I found 4 blocks. Then after that, (once a version was upgraded) I found nothing...

Generation is hugely variable. I have seen statistics for seven similar generators. In the same timespan they generated 1, 2, 4, 0, 7, 0 and 0 blocks. Go figure!

[xenon481]

It's not so much that I'm complaining or surprised. I just find it REALLY odd that the first day I mined Bitcoin on like 1/22 or so, I found 4 blocks. Then after that, (once a version was upgraded) I found nothing...

Variance.

I was watching Slush's pool yesterday while it was running ~60GHash/sec. At that rate, the pool should have been solving a block every ~30 minutes and should have had a 95% chance to have found one within ~1.5hrs. But, there was one that took over 5hrs to solve.

If you average 625MHash/sec solo mining running 24/7, (at the current difficulty) you should average ~2 days per block. But the variance is a killer. That 5hrs solve for Slush's pool yesterday? That's 10x longer than the average. For you, that level of variance would have taken ~20 days.

[ribuck]

Just look at Bitcoin Monitor and see the variance in block generation.

[Cablesaurus]

It's not so much that I'm complaining or surprised. I just find it REALLY odd that the first day I mined Bitcoin on like 1/22 or so, I found 4 blocks. Then after that, (once a version was upgraded) I found nothing...

Variance.

I was watching Slush's pool yesterday while it was running ~60GHash/sec. At that rate, the pool should have been solving a block every ~30 minutes and should have had a 95% chance to have found one within ~1.5hrs. But, there was one that took over 5hrs to solve.

If you average 625MHash/sec solo mining running 24/7, (at the current difficulty) you should average ~2 days per block. But the variance is a killer. That 5hrs solve for Slush's pool yesterday? That's 10x longer than the average. For you, that level of variance would have taken ~20 days.

Okay, is there a way to get around this in solo mining?

Changing the getwork rate or anything? Just wondering if theres anyway to dump the workload if it's been working on a block for 'too long'

Not sure if that'd even make sense, but it seems like it would be a good idea

[OVerLoRDI]

I'll test 2.2 and see if it fixes the cpu usage problem soon.

Another question, how often do you guys get invalid or stale blocks?  I'm hooked up with a mining pool.  It seems to me I get 1/25 blocks as stale or invalid.  Is this normal? or should I look into stability issues with my video cards?  They are not overclocked and I haven't had any issues with them gaming.

[Cryptoman]

I upgraded about 2 Ghash/s worth of miners a couple days ago with the latest version and haven't found anything since either.  Maybe it's just a spell of bad luck.  I'll update the thread if anything manifests.

Finally got a block, on my slowest GPU, a 5770 running about 190 Mhash/s.  This is the first block in three days.  The 95th percentile for my cluster is about 2 days.  Has anyone kept records and compared ideal generation rate to actual rate over a longer period of time?  I'm going to start doing so.

[Fractality]

I received a message "accepted" a couple of hours ago. Does that mean I found a block? Nothing has appeared in my wallet, though. Unfortunately I rebooted the PC by now, so I can't remember the value associated with the message - unless it is in some log file?

[marcus_of_augustus]

Hi xenon481,

have you got a miner running on the xenon XCGPU processor?

If so what kind of Mhash/s are you seeing?

thnx.

[yunk3r]

when i run the poclbm.exe iam able to get to the point when it shows my khash/s which is 308775, but in bitcoin it says only 10762khash/s is this how it is supposed to look.