[ANN] cudaMiner & ccMiner CUDA based mining applications [Windows/Linux/MacOSX]

[cbuchner1]

I get the "does not validate on CPU" messages mostly when I choose kernel launch configurations that the RAM on the card does not support.

memory usage can be estimated by calculating  blocks x warps x 4 MBytes   where your -l argument to cudaminer specifies blocks x warps

[1701752934]

1701752934

[1701752934]

1701752934

[1701752934]

1701752934

[1701752934]

1701752934

[K1773R]

I get the "does not validate on CPU" messages mostly when I choose kernel launch configurations that the RAM on the card does not support.

memory usage can be estimated by calculating  blocks x warps x 4 MBytes   where your -l argument to cudaminer specifies blocks x warps

i also get it if the con to the pool is lost for some time and afterwards being reestablished. a simple cronjob that kills and restart cudaminer every 5 minutes solves it nicely

[mhps]

i got a "result does not validate on CPU!" after starting cudaminer (after try (start) stratum_proxy) .. (coionotron)

what does that message mean?

remove the -l and -i -C parameters and use -D to find the best -l parameters for you.

[verzon]

  thx for your answers and suggestions - i´ll test it but could be tommorow when do - ´ll be back when tested to bring feedback with these workarounds

[dbabo]

Gents,
 mind to decrypt the following for me please? thank you

Code:

05/03/13 18:27:42 CTxMemPool::accept() : accepted 46d7990963 (poolsz 37)
05/03/13 18:28:00 SetBestChain: new best=69f963d82d358c88ee3d  height=346030  work=44021880709632917
05/03/13 18:28:00 ProcessBlock: ACCEPTED
05/03/13 18:28:00 force request: block 69f963d82d358c88ee3d
05/03/13 18:28:00 force request: block 69f963d82d358c88ee3d
05/03/13 18:28:00 force request: block 69f963d82d358c88ee3d
05/03/13 18:28:00 CreateNewBlock(): total size 1710
05/03/13 18:28:00 force request: block 69f963d82d358c88ee3d
05/03/13 18:28:00 Rate limit dFreeCount: 13052.3 => 13278.3
05/03/13 18:28:00 CTxMemPool::accept() : accepted f37e279f9c (poolsz 22)
05/03/13 18:28:00 getblocks -1 to 00000000000000000000 limit 500
05/03/13 18:28:00 getblocks -1 to 00000000000000000000 limit 500
05/03/13 18:28:00 getblocks -1 to 00000000000000000000 limit 500
05/03/13 18:28:01 Flushing wallet.dat
05/03/13 18:28:01 Flushed wallet.dat 30ms
05/03/13 18:28:07 getblocks -1 to 00000000000000000000 limit 500
05/03/13 18:28:12 ERROR: FetchInputs() : 0f3a6ab0fb mempool Tx prev not found 0dfc0db223
05/03/13 18:28:12 stored orphan tx 0f3a6ab0fb (mapsz 38)

[termhn]

Gents,
 mind to decrypt the following for me please? thank you

Code:

05/03/13 18:27:42 CTxMemPool::accept() : accepted 46d7990963 (poolsz 37)
05/03/13 18:28:00 SetBestChain: new best=69f963d82d358c88ee3d  height=346030  work=44021880709632917
05/03/13 18:28:00 ProcessBlock: ACCEPTED
05/03/13 18:28:00 force request: block 69f963d82d358c88ee3d
05/03/13 18:28:00 force request: block 69f963d82d358c88ee3d
05/03/13 18:28:00 force request: block 69f963d82d358c88ee3d
05/03/13 18:28:00 CreateNewBlock(): total size 1710
05/03/13 18:28:00 force request: block 69f963d82d358c88ee3d
05/03/13 18:28:00 Rate limit dFreeCount: 13052.3 => 13278.3
05/03/13 18:28:00 CTxMemPool::accept() : accepted f37e279f9c (poolsz 22)
05/03/13 18:28:00 getblocks -1 to 00000000000000000000 limit 500
05/03/13 18:28:00 getblocks -1 to 00000000000000000000 limit 500
05/03/13 18:28:00 getblocks -1 to 00000000000000000000 limit 500
05/03/13 18:28:01 Flushing wallet.dat
05/03/13 18:28:01 Flushed wallet.dat 30ms
05/03/13 18:28:07 getblocks -1 to 00000000000000000000 limit 500
05/03/13 18:28:12 ERROR: FetchInputs() : 0f3a6ab0fb mempool Tx prev not found 0dfc0db223
05/03/13 18:28:12 stored orphan tx 0f3a6ab0fb (mapsz 38)

Where is this from and what do you want to know/what do you want to fix?

[dbabo]

<gutted out to save space>
Where is this from and what do you want to know/what do you want to fix?

this is from litecoind daemon log file.
I just want to understand what is going on and what i need ( if anything) to do ?
for example:
 what does "CTxMemPool::accept() : accepted 46d7990963 (poolsz 37)" means? what is accepted and what is poolsz?
 what does "ProcessBlock: ACCEPTED" means?
 what does "force request: block" means?
 what does "getblocks -1 to 00000000000000000000 limit 500" means?
 what does "ERROR: FetchInputs() : 0f3a6ab0fb mempool Tx prev not found 0dfc0db223" means and do i need to do anything about this error?
and finally "stored orphan tx 0f3a6ab0fb (mapsz 38)?

this portion is something new - i haven't noticed before:

Code:

05/03/13 19:17:27 ProcessBlock: ACCEPTED
05/03/13 19:17:27 force request: block 8439de5375f48730c2e9
05/03/13 19:17:27 getblocks 346045 to 8439de5375f48730c2e9 limit 500
05/03/13 19:17:27   getblocks stopping at 346046 8439de5375f48730c2e9
05/03/13 19:17:27 getblocks -1 to 00000000000000000000 limit 500
05/03/13 19:17:29 Flushing wallet.dat
05/03/13 19:17:29 Flushed wallet.dat 22ms
05/03/13 19:17:30 CreateNewBlock(): total size 1225
05/03/13 19:17:33 getblocks -1 to 00000000000000000000 limit 500
05/03/13 19:17:39 ResendWalletTransactions()
05/03/13 19:17:40 Rate limit dFreeCount: 13685.4 => 13910.4

what is going on here? why is it "   getblocks stopping" ? and why esendWalletTransactions() ?
thank you!

[Cheshyr]

I get the "does not validate on CPU" messages mostly when I choose kernel launch configurations that the RAM on the card does not support.

memory usage can be estimated by calculating  blocks x warps x 4 MBytes   where your -l argument to cudaminer specifies blocks x warps

I get this on Win7 when Chrome decides to muck with something; closing Chrome followed by restarting cudaMiner solves the problem.

[GuiltySpark343]

Trying to read through all 37 pages of this thread... but I have a GTX Titan and using this to mine litecoins. Getting about 190 kh/s at stock clocks/voltages and no tweaking of launch config (just relying on auto-tune).

Does 190 kh/s sound right? Any flags to optimize a TITAN?

[cbuchner1]

Closer to 300kH/s would be more realistic for a Titan (slightly overclocked, maybe). Aren't there any Titan launch configurations listed in the Google Docs spreadsheet?

Make sure you use -i 0

[GuiltySpark343]

Checked google docs spreadsheet. No Titan configs. I will try -I and see thanks!

[ritchan]

Using CUDAMiner on a GTX470 at 770MHz here - great speeds,  180-200kH/s! Thanks so much! GPU-Z says GPU usage hovers around 90% though, is something not optimized enough? I overclocked just the VRAM by ~15%, the GPU usage did not change, so it's not a memory bottleneck.

[cbuchner1]

GPU usage may be reduced when running in interactive mode (which is now the default for display-driving cards). So use -i 0 and get another speed boost. Indeed overclocking the core clock works best, and the memory clock is not very significant.

[tidus_13]

It won't mine with stratum?

[Stoneysilence]

It won't mine with stratum?

Not without a Stratum Proxy running.  No inherent support in the software.

[GuiltySpark343]

Where are the Titan owners? Hello, we have support for Titan and I really need someone to test this. If you could post the result of a Titan autotune session with debug flags to the forum - that would be immensely helpful.

Do you still need some debugging on Titan? I have one, and am more than happy to help! Just PM me or reply to this thread.

[KnowBuddy]

i got a "result does not validate on CPU!" after starting cudaminer (after try (start) stratum_proxy) .. (coionotron)

what does that message mean?

remove the -l and -i -C parameters and use -D to find the best -l parameters for you.

Just an FYI, the -D flag just turns on verbose output (Debugging). You'll still get an autotune just by removing the -l parameter, it just won't tell you all the things it's trying until it settles on the one it thinks is best.

[verzon]

i got a "result does not validate on CPU!" after starting cudaminer (after try (start) stratum_proxy) .. (coionotron)

what does that message mean?

remove the -l and -i -C parameters and use -D to find the best -l parameters for you.

Just an FYI, the -D flag just turns on verbose output (Debugging). You'll still get an autotune just by removing the -l parameter, it just won't tell you all the things it's trying until it settles on the one it thinks is best.

ok ... what i want to have is getting it worked no matter how much is khash cause it is only a geforce 210 on a board with 4 core amd cpu.

i do not want to kill the cpu, but trying with coinotron i want to see on website coinotron in member - area some stats of my worker better than "0.0"  .. no matter how much the khash is ..

if i took the -i .. interactive .. system slow down, usage of (cpu / ) gpu high ... now i will test without interactiv giving a "6x2" ... ... ok, only want to get it woriking with showing results in my member stats on coinotron-website ... it´s about 6x2 will get some around <5 khash ... tested with some more khash wont work 100% (getting errors)

edit:
something new .. using mining_proxy .. unexpected error user authentication ...

[mhps]

if i took the -i .. interactive .. system slow down, usage of (cpu / ) gpu high ... now i will test without interactiv giving a "6x2" ... ... ok, only want to get it woriking with showing results in my member stats on coinotron-website ... it´s about 6x2 will get some around <5 khash ... tested with some more khash wont work 100% (getting errors)

edit:
something new .. using mining_proxy .. unexpected error user authentication ...

I also first tried with no -i option and random -l AxB opptions. 6x2 is a good start    Then I found if I use -D the program will doa grid search for me. Just rememer which AxB gives you the highest kh/s and use that AxB in the commandline in the future.

Note that some older mining_proxy doesn't have the -pa option that lets you choose  scrypt. better follow my instruction. the -nm and the --no-longpoll ones can be taken off.

[Chiz]

Can anyone help me. I am not using Stratum/Proxy cause I don't know how to set it up with CUDAMiner....

I've tried so many settings, but for the past day/2 days I been getting all these fails + my acceptance rate is like every 2-3 minutes, takes a good while, now I don't know if this is because of a difficulty increase or what, I'll admit, I am nooby.

http://gyazo.com/112d32b381e2146428edf6135addc3bf

Anyway just want to know is it me, or the pools.

I am building an AMD 7950 rig this weekend so hope these problems get solved otherwise I have to go to LTC :\ but I want to stick with FTC