search

ronnieman / pyrit

Automatically exported from code.google.com/p/pyrit
0 stars 0 forks source link

No commands function on GTX 470 #153

Closed GoogleCodeExporter closed 8 years ago

GoogleCodeExporter commented 8 years ago 
What steps will reproduce the problem?
1. pyrit list_cores
2. pyrit benchmark
3.

What is the expected output? What do you see instead?

Expected output is obvious.

----------------------------------------------------------------------
list_cores

Traceback (most recent call last):
  File "/usr/local/bin/pyrit", line 6, in <module>
    pyrit_cli.Pyrit_CLI().initFromArgv()
  File "/usr/local/lib/python2.6/dist-packages/pyrit_cli.py", line 110, in
initFromArgv
    func(self, **options)
  File "/usr/local/lib/python2.6/dist-packages/pyrit_cli.py", line 255, in
list_cores
    with cpyrit.cpyrit.CPyrit() as cp:
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 392,
in __init__
    self.cores.append(CUDACore(queue=self, dev_idx=dev_idx))
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 181,
in __init__
    _cpyrit_cuda.CUDADevice.__init__(self, dev_idx)
SystemError: CUDA_ERROR_INVALID SOURCE

----------------------------------------------------------------------
benchmark

Traceback (most recent call last):
  File "/usr/local/bin/pyrit", line 6, in <module>
    pyrit_cli.Pyrit_CLI().initFromArgv()
  File "/usr/local/lib/python2.6/dist-packages/pyrit_cli.py", line 110, in
initFromArgv
    func(self, **options)
  File "/usr/local/lib/python2.6/dist-packages/pyrit_cli.py", line 897, in
benchmark
    with cpyrit.cpyrit.CPyrit() as cp:
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 392,
in __init__
    self.cores.append(CUDACore(queue=self, dev_idx=dev_idx))
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 181,
in __init__
    _cpyrit_cuda.CUDADevice.__init__(self, dev_idx)
SystemError: CUDA_ERROR_INVALID SOURCE

----------------------------------------------------------------------

What version of the product are you using? On what operating system?
just retrieved from svn this evening
ubuntu 9.10

Please provide any additional information below.
I have both a GTX470 and GT8800, and I have had the 8800 working by itself,
but I haven't been able to get them working in tandem, nor have I been able
to get the 470 to work solo.

The past two nvidia driver releases supported the 400 series cards, but
they both provide me the errors above. Everything else is up to date as far
as I am aware. Thanks in advance for any assistance.

Original issue reported on code.google.com by tdin...@gmail.com on 30 Apr 2010 at 4:59

GoogleCodeExporter commented 8 years ago 
I would guess that you've had installed cpyrit_cuda before you had 470.
Try to do
1. Install latest drivers & cuda toolkit
2. remove _cpyrit_cudakernel.cubin.h ( cpyrit_cuda directory )
3. remove build subdirectory from cpyrit_cuda
4. compile & install

If it will work please post benchmark results :). Also if you can from opencl 
version 
too :).

Original comment by hazema...@gmail.com on 30 Apr 2010 at 9:34

GoogleCodeExporter commented 8 years ago 
This was from a clean install, and I do think that I installed all my drivers, 
the toolkit, and the sdk before 
doing anything else, but I went ahead and followed your suggestion. I still 
receive the same output when i 
attempt to do anything with pyrit.

After I compile and install pyrit, I can run list_cores or benchmark, and it 
behaves as expected. As soon 
as I add cpyrit_cuda, thats when I get the errors as listed above.

Original comment by tdin...@gmail.com on 30 Apr 2010 at 3:23

GoogleCodeExporter commented 8 years ago 
Hopefully this can be solved soon as I have a superclocked 480 arriving in a 
couple
hours and I would like to bench it.

Original comment by sprice82 on 30 Apr 2010 at 3:36

GoogleCodeExporter commented 8 years ago 
Well my 480 does the same thing, this sucks. When I get home later tonight I 
will
take a look at it and see what I find.

Original comment by sprice82 on 1 May 2010 at 12:47

GoogleCodeExporter commented 8 years ago 
Just to ring in on this, I've been working on this same issue with my 480 for 
the past 
day or so. I haven't got anywhere with it. My 480 gives the exact same output 
as 
above. Nvidia driver is 195.36.24 and pyrit and cpyrit_cuda were both from 
yesterdays 
svn.

Let me know if you need me to provide any particular information.

Original comment by ryan...@gmail.com on 1 May 2010 at 3:44

GoogleCodeExporter commented 8 years ago 
Perhaps the following will be of some use. From the NVIDIA CUDA Fermi 
Compatibility Guide:

How can I confirm that my application is ready to run on Fermi?
Answer: CUDA applications built using the CUDA Toolkit versions 2.1 through 2.3 
are compatible with Fermi 
as long as they are built to include PTX versions of their kernels. NVIDIA 
Driver versions 195.xx or newer allow 
the application to use the PTX JIT code path. To test that PTX JIT is working 
for your application, you can do 
the following:
 Go to the NVIDIA website, and install the latest R195 driver. 
 Set the system environment flag CUDA_FORCE_PTX_JIT=1 
 Launch your application.

1   Fermi Compatibility Guide for CUDA Applications
When starting a CUDA application for the first time with the above environment 
flag, the CUDA driver will JIT 
compile the PTX for each CUDA kernel that is used into native CUBIN code. The 
generated CUBIN for the target 
GPU architecture is cached by the CUDA driver. This cache persists across 
system shutdown/restart events.
If this test passes, then your application is ready for Fermi.

Guide is found here: http://developer.nvidia.com/object/gpucomputing.html

Does cpyrit_cuda include support for the PTX versions?

Original comment by ryan...@gmail.com on 2 May 2010 at 3:41

GoogleCodeExporter commented 8 years ago 
If there is a problem it's probably line 81 in setup.py

subprocess.check_call(NVCC + ' --host-compilation C -Xptxas "-v" '\
                        '-Xcompiler "-fPIC" --cubin ' \
                        './_cpyrit_cudakernel.cu', shell=True)
f = open("_cpyrit_cudakernel.cubin", "rb")

Probably some flags should be changed ( I guess --cubin :) ). And output file 
name
from nvcc compiler can change ( _cpyrit_cudakernel.cubin ).

I don't have nvidia card myself ( so can't test it ). But here it goes 

subprocess.check_call(NVCC + ' --host-compilation C -Xptxas "-v" '\
                        '-Xcompiler "-fPIC" -ptx ' \
                        '-o _cpyrit_cudakernel.ptx ./_cpyrit_cudakernel.cu', shell=True)
f = open("_cpyrit_cudakernel.ptx", "rb")

Original comment by hazema...@gmail.com on 2 May 2010 at 7:37

GoogleCodeExporter commented 8 years ago 
[deleted comment]
GoogleCodeExporter commented 8 years ago 
After exporting CUDA_FORCE_PTX_JIT=1 - the output changes (pyrit was rebuilded 
after
export):

pyrit list_cores
Pyrit 0.3.1-dev (svn r238) (C) 2008-2010 Lukas Lueg http://pyrit.googlecode.com
This code is distributed under the GNU General Public License v3+

Traceback (most recent call last):
  File "/usr/bin/pyrit", line 6, in <module>
    pyrit_cli.Pyrit_CLI().initFromArgv()
  File "/usr/lib/python2.6/site-packages/pyrit_cli.py", line 110, in initFromArgv
    func(self, **options)
  File "/usr/lib/python2.6/site-packages/pyrit_cli.py", line 243, in list_cores
    with cpyrit.cpyrit.CPyrit() as cp:
  File "/usr/lib/python2.6/site-packages/cpyrit/cpyrit.py", line 368, in __init__
    self.cores.append(CUDACore(queue=self, dev_idx=dev_idx))
  File "/usr/lib/python2.6/site-packages/cpyrit/cpyrit.py", line 181, in __init__
    _cpyrit_cuda.CUDADevice.__init__(self, dev_idx)
SystemError: CUDA_ERROR_NO_BINARY_FOR_GPU

When exporting CUDA_FORCE_PTX_JIT=0, pyrit reports CUDA_ERROR_INVALID SOURCE 
again ..
(GTX 480 && GTX 280 inside, drivers 195.36,24, gentoo)

Original comment by collib...@gmail.com on 2 May 2010 at 12:44

GoogleCodeExporter commented 8 years ago 
thx hazeman, it works ;) ..

Pyrit 0.3.1-dev (svn r253) (C) 2008-2010 Lukas Lueg http://pyrit.googlecode.com
This code is distributed under the GNU General Public License v3+

Running benchmark (41273.7 PMKs/s)... \ 

Computed 41367.20 PMKs/s total.
#1: 'CUDA-Device #1 'GeForce GTX 480'': 27969.5 PMKs/s (RTT 1.4)
#2: 'CUDA-Device #2 'GeForce GTX 280'': 12158.4 PMKs/s (RTT 2.9)
#3: 'CPU-Core (SSE2)': 599.6 PMKs/s (RTT 3.0)
#4: 'CPU-Core (SSE2)': 592.9 PMKs/s (RTT 3.0)

on Intel(R) Core(TM)2 Extreme CPU X9650  @ 3.00GHz, MB DX48BT2, GPU's not
owerclocked, GTX280 on 4x PCI-e

Original comment by collib...@gmail.com on 2 May 2010 at 2:27

GoogleCodeExporter commented 8 years ago 
Good to hear it :). But we will have to wait for Lukas to make the change in 
svn.
He's busy lately so it will probably take some time.

But what's surprising is the 480 performance. ATI 5870 is doing 83000 ( latest 
CAL++
core ).

Original comment by hazema...@gmail.com on 2 May 2010 at 3:32

GoogleCodeExporter commented 8 years ago 
Well the good news is that your modification in setup.py worked, hazeman. 
cpyrit_cuda 
now builds properly and 'pyrit list_cores' lists the CUDA-Device GTX 480 
properly. 
However, when I try running a 'pyrit benchmark' it dies with the following:

Calibrating... Exception in thread CUDA-Device #1 'GeForce GTX 480':
Traceback (most recent call last):
  File "/usr/lib/python2.6/threading.py", line 525, in __bootstrap_inner
    self.run()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 111, in run
    self._testComputeFunction(101)
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 100, in 
_testComputeFunction
    self.solve(Core.TV_ESSID, [Core.TV_PW] * i))):
SystemError: CUDA_ERROR_LAUNCH_FAILED

Traceback (most recent call last):
  File "/usr/local/bin/pyrit", line 6, in <module>
    pyrit_cli.Pyrit_CLI().initFromArgv()
  File "/usr/local/lib/python2.6/dist-packages/pyrit_cli.py", line 110, in 
initFromArgv
    func(self, **options)
  File "/usr/local/lib/python2.6/dist-packages/pyrit_cli.py", line 903, in benchmark
    cp.enqueue('foo', ['barbarbar'] * 500)
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 514, in 
enqueue
    self._check_cores()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 440, in 
_check_cores
    raise SystemError("The core '%s' has died unexpectedly" % core)
SystemError: The core 'CUDA-Device #1 'GeForce GTX 480'' has died unexpectedly

Original comment by ryan...@gmail.com on 2 May 2010 at 4:49

GoogleCodeExporter commented 8 years ago 
Got mine working as well, thanks hazeman! Here's the benchmark if anybody is 
looking
at price vs performance in deciding to buy a 470 or 480. By the way, is the 
83000
benchmark you were referencing a single card? If so, that seems just insane...

Pyrit 0.3.1-dev (svn r253) (C) 2008-2010 Lukas Lueg http://pyrit.googlecode.com
This code is distributed under the GNU General Public License v3+

Running benchmark (29798.9 PMKs/s)... / 

Computed 30782.68 PMKs/s total.
#1: 'CUDA-Device #1 'GeForce GTX 470'': 23234.6 PMKs/s (RTT 1.7)
#2: 'CUDA-Device #2 'GeForce 8800 GT'': 5971.0 PMKs/s (RTT 3.0)
#3: 'CPU-Core (SSE2)': 712.2 PMKs/s (RTT 3.0)
#4: 'CPU-Core (SSE2)': 683.6 PMKs/s (RTT 3.0)

Original comment by tdin...@gmail.com on 3 May 2010 at 5:51

GoogleCodeExporter commented 8 years ago 
Yep single card - singe core :). Dual core 5970 is doing 145K.
To tell you the truth you could probably add 15% to 4xx results for fair 
comparison.
This new CAL++ core is using LowLatencyCore class which could help also cuda 
core.

Original comment by hazema...@gmail.com on 3 May 2010 at 6:00

GoogleCodeExporter commented 8 years ago 
I've increased the maximum buffer-size for the cuda-core. It should give you
rtt-values of close to 3.0 again and slightly increase performance.

Original comment by lukas.l...@gmail.com on 3 May 2010 at 6:19

GoogleCodeExporter commented 8 years ago 
yes lukas, performance slightly goes up ;)

Pyrit 0.3.1-dev (svn r253) (C) 2008-2010 Lukas Lueg http://pyrit.googlecode.com
This code is distributed under the GNU General Public License v3+

Running benchmark (43482.3 PMKs/s)... | 

Computed 43494.89 PMKs/s total.
#1: 'CUDA-Device #1 'GeForce GTX 480'': 28637.6 PMKs/s (RTT 2.7)
#2: 'CUDA-Device #2 'GeForce GTX 280'': 12167.4 PMKs/s (RTT 2.9)
#3: 'CPU-Core (SSE2)': 720.7 PMKs/s (RTT 3.0)
#4: 'CPU-Core (SSE2)': 716.6 PMKs/s (RTT 3.0)

5870 has really nice performance here .. of course, fermi has different GPU 
staff
"geometry" .. may be after some time we will see ~45k on GTX480 ;) ..

Original comment by collib...@gmail.com on 3 May 2010 at 1:01

GoogleCodeExporter commented 8 years ago 
I'm afraid it's small chance for 480 to achieve 45k. In theory it can be as 
much as 
480*1401000000/(4*4096*1311)=31308 ( 480(cuda
cores)*1401mhz(clock)/(4*4096*1311-number of operations per pmk ) 
If the 480 would have cyclic rotate it would be 
480*1401000000/(4*4096*863)=47560 -
but looking at the ptx instruction set I don't see cyclic shift :(.
For 5870 theoretical max speed is 96K ( 1600*850mhz/(4*4096*863) ).

Original comment by hazema...@gmail.com on 3 May 2010 at 3:21

GoogleCodeExporter commented 8 years ago 
Is the new svn (254) supposed to fix this, because I am still getting the same 
error.

Original comment by sprice82 on 3 May 2010 at 5:49

GoogleCodeExporter commented 8 years ago 
sprice: I don't believe it was. After doing an svn update, you will have to 
re-fix 
your setup.py the way hazeman mentioned in Comment 7. This will allow you to 
properly 
build cpyrit_cuda.

That seems to have been enough for the others, but let us know if it works for 
you 
too. It got me past the build and getting the card recognized by pyrit, but now 
benchmark and selftest are failing with new errors such as the ones I posted 
above.

Original comment by ryan...@gmail.com on 3 May 2010 at 5:56

GoogleCodeExporter commented 8 years ago 
I have made your change several times and it hasn't solved the problem, I am not
entirely sure why. I may just wait for the svn that actually fixes it.

Original comment by sprice82 on 3 May 2010 at 6:25

GoogleCodeExporter commented 8 years ago 
What error do you get after making the changes hazeman mentioned? Are you sure 
that 
the indention is the same as the original? That may seem funny, but python 
complained 
about it the first time I did it. Can you post what error you are getting, what 
driver 
version, and what CUDA toolkit version?

Original comment by ryan...@gmail.com on 3 May 2010 at 6:34

GoogleCodeExporter commented 8 years ago 
Just to update this ticket, I have a clean install of Linux Mint 9 x64 with a 
GTX480. I just did a clean new install of the latest NVIDIA cuda dev driver 
(256.35) and the CUDA toolkit 3.1. Pulled down the latest svn of pyrit and did 
the build/install of pyrit and cpyrit_cuda. Afterward, running anything in 
pyrit fails with the errors in the original post here. 

I did the fix provided by hazeman in post #7, removed all the files from the 
initial build/install of cpyrit_cuda, rebuilt it, and reinstalled it. Pyrit 
list_cores now lists my GTX480 (instead of failing like before). However, pyrit 
benchmark is still failing with the following errors:

vidkun@tyr ~/pyrit/cpyrit_cuda $ pyrit benchmark
Pyrit 0.3.1-dev (svn r265) (C) 2008-2010 Lukas Lueg http://pyrit.googlecode.com
This code is distributed under the GNU General Public License v3+

Calibrating... Exception in thread CUDA-Device #1 'GeForce GTX 480':
Traceback (most recent call last):
  File "/usr/lib/python2.6/threading.py", line 532, in __bootstrap_inner
    self.run()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 112, in run
    self._testComputeFunction(101)
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 101, in _testComputeFunction
    self.solve(Core.TV_ESSID, [Core.TV_PW] * i))):
SystemError: CUDA_ERROR_LAUNCH_FAILED

Exception in thread CPU-Core (SSE2):
Traceback (most recent call last):
  File "/usr/lib/python2.6/threading.py", line 532, in __bootstrap_inner
    self.run()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 109, in run
    essid, pwlist = self.queue._gather(self.buffersize, timeout=0.5)
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 640, in _gather
    self._check_cores()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 498, in _check_cores
    raise SystemError("The core '%s' has died unexpectedly" % core)
SystemError: The core 'CUDA-Device #1 'GeForce GTX 480'' has died unexpectedly

Exception in thread CPU-Core (SSE2):
Traceback (most recent call last):
  File "/usr/lib/python2.6/threading.py", line 532, in __bootstrap_inner
    self.run()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 109, in run
    essid, pwlist = self.queue._gather(self.buffersize, timeout=0.5)
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 640, in _gather
    self._check_cores()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 498, in _check_cores
    raise SystemError("The core '%s' has died unexpectedly" % core)
SystemError: The core 'CUDA-Device #1 'GeForce GTX 480'' has died unexpectedly

Exception in thread CPU-Core (SSE2):
Traceback (most recent call last):
  File "/usr/lib/python2.6/threading.py", line 532, in __bootstrap_inner
    self.run()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 109, in run
    essid, pwlist = self.queue._gather(self.buffersize, timeout=0.5)
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 640, in _gather
    self._check_cores()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 498, in _check_cores
    raise SystemError("The core '%s' has died unexpectedly" % core)
SystemError: The core 'CUDA-Device #1 'GeForce GTX 480'' has died unexpectedly

Exception in thread CPU-Core (SSE2):
Traceback (most recent call last):
  File "/usr/lib/python2.6/threading.py", line 532, in __bootstrap_inner
    self.run()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 109, in run
    essid, pwlist = self.queue._gather(self.buffersize, timeout=0.5)
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 640, in _gather
    self._check_cores()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 498, in _check_cores
    raise SystemError("The core '%s' has died unexpectedly" % core)
SystemError: The core 'CUDA-Device #1 'GeForce GTX 480'' has died unexpectedly

Exception in thread CPU-Core (SSE2):
Traceback (most recent call last):
  File "/usr/lib/python2.6/threading.py", line 532, in __bootstrap_inner
    self.run()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 109, in run
    essid, pwlist = self.queue._gather(self.buffersize, timeout=0.5)
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 640, in _gather
    self._check_cores()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 498, in _check_cores
    raise SystemError("The core '%s' has died unexpectedly" % core)
SystemError: The core 'CUDA-Device #1 'GeForce GTX 480'' has died unexpectedly

Exception in thread CPU-Core (SSE2):
Traceback (most recent call last):
  File "/usr/lib/python2.6/threading.py", line 532, in __bootstrap_inner
    self.run()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 109, in run
    essid, pwlist = self.queue._gather(self.buffersize, timeout=0.5)
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 640, in _gather
    self._check_cores()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 498, in _check_cores
    raise SystemError("The core '%s' has died unexpectedly" % core)
SystemError: The core 'CUDA-Device #1 'GeForce GTX 480'' has died unexpectedly

Exception in thread CPU-Core (SSE2):
Traceback (most recent call last):
  File "/usr/lib/python2.6/threading.py", line 532, in __bootstrap_inner
    self.run()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 109, in run
    essid, pwlist = self.queue._gather(self.buffersize, timeout=0.5)
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 640, in _gather
    self._check_cores()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 498, in _check_cores
    raise SystemError("The core '%s' has died unexpectedly" % core)
SystemError: The core 'CUDA-Device #1 'GeForce GTX 480'' has died unexpectedly

Traceback (most recent call last):
  File "/usr/local/bin/pyrit", line 6, in <module>
    pyrit_cli.Pyrit_CLI().initFromArgv()
  File "/usr/local/lib/python2.6/dist-packages/pyrit_cli.py", line 110, in initFromArgv
    func(self, **options)
  File "/usr/local/lib/python2.6/dist-packages/pyrit_cli.py", line 903, in benchmark
    cp.enqueue('foo', ['barbarbar'] * 500)
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 572, in enqueue
    self._check_cores()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 498, in _check_cores
    raise SystemError("The core '%s' has died unexpectedly" % core)
SystemError: The core 'CUDA-Device #1 'GeForce GTX 480'' has died unexpectedly

What else can I provide to help get to the bottom of what is causing this error?

Original comment by ryan...@gmail.com on 8 Jul 2010 at 2:58

GoogleCodeExporter commented 8 years ago 
I tried nvidia driver 195.36.31 and cudatoolkit 2.3 with the changes proposed 
by hazeman11 in comment 7, but I still get the same result as ryanlev in 
comment 22. Its clean build with version r268 pyrit.

Original comment by huntthe...@gmail.com on 16 Jul 2010 at 12:29

GoogleCodeExporter commented 8 years ago 
I can confirm the issues that posts 22 and 23 are having. A totally clean 
build, running the 3.1 cuda toolkit, with 256.35 drivers with post 7 fix 
applied. It just dies unexpectedly... Also the Binary to LargeBinary warning 
that was mentioned in other posts is still kicking around.

If anyone has any ideas for a work around it would be greatly appreciated (for 
the 480 dieing that is).

jake@sniper:~/bomba/FileServer/VI/pyrit/pyrit-read-only/cpyrit_cuda$ pyrit 
list_cores
/usr/local/lib/python2.6/dist-packages/cpyrit/storage.py:743: 
SADeprecationWarning: The Binary type has been renamed to LargeBinary.
  sql.Column('essid', sql.Binary(32), nullable=False),
/usr/local/lib/python2.6/dist-packages/cpyrit/storage.py:753: 
SADeprecationWarning: The Binary type has been renamed to LargeBinary.
  sql.Column('collection_buffer', sql.Binary(2**24-1), \
/usr/local/lib/python2.6/dist-packages/cpyrit/storage.py:765: 
SADeprecationWarning: The Binary type has been renamed to LargeBinary.
  sql.Column('results_buffer', sql.Binary(2**24 - 1), \
Pyrit 0.3.1-dev (svn r269) (C) 2008-2010 Lukas Lueg http://pyrit.googlecode.com
This code is distributed under the GNU General Public License v3+

The following cores seem available...
#1:  'CUDA-Device #1 'GeForce GTX 480''
#2:  'CUDA-Device #2 'GeForce GTX 480''
#3:  'CPU-Core (SSE2)'
#4:  'CPU-Core (SSE2)'
#5:  'CPU-Core (SSE2)'
#6:  'CPU-Core (SSE2)'
#7:  'CPU-Core (SSE2)'
#8:  'CPU-Core (SSE2)'
jake@sniper:~/bomba/FileServer/VI/pyrit/pyrit-read-only/cpyrit_cuda$ pyrit 
benchmark
/usr/local/lib/python2.6/dist-packages/cpyrit/storage.py:743: 
SADeprecationWarning: The Binary type has been renamed to LargeBinary.
  sql.Column('essid', sql.Binary(32), nullable=False),
/usr/local/lib/python2.6/dist-packages/cpyrit/storage.py:753: 
SADeprecationWarning: The Binary type has been renamed to LargeBinary.
  sql.Column('collection_buffer', sql.Binary(2**24-1), \
/usr/local/lib/python2.6/dist-packages/cpyrit/storage.py:765: 
SADeprecationWarning: The Binary type has been renamed to LargeBinary.
  sql.Column('results_buffer', sql.Binary(2**24 - 1), \
Pyrit 0.3.1-dev (svn r269) (C) 2008-2010 Lukas Lueg http://pyrit.googlecode.com
This code is distributed under the GNU General Public License v3+

Calibrating... Exception in thread CUDA-Device #2 'GeForce GTX 480':
Traceback (most recent call last):
  File "/usr/lib/python2.6/threading.py", line 532, in __bootstrap_inner
    self.run()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 112, in run
    self._testComputeFunction(101)
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 101, in _testComputeFunction
    self.solve(Core.TV_ESSID, [Core.TV_PW] * i))):
SystemError: CUDA_ERROR_LAUNCH_FAILED

Exception in thread CPU-Core (SSE2):
Traceback (most recent call last):
  File "/usr/lib/python2.6/threading.py", line 532, in __bootstrap_inner
    self.run()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 109, in run
    essid, pwlist = self.queue._gather(self.buffersize, timeout=0.5)
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 640, in _gather
    self._check_cores()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 498, in _check_cores
    raise SystemError("The core '%s' has died unexpectedly" % core)
SystemError: The core 'CUDA-Device #2 'GeForce GTX 480'' has died unexpectedly

Exception in thread CPU-Core (SSE2):
Traceback (most recent call last):
  File "/usr/lib/python2.6/threading.py", line 532, in __bootstrap_inner
    self.run()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 109, in run
    essid, pwlist = self.queue._gather(self.buffersize, timeout=0.5)
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 640, in _gather
    self._check_cores()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 498, in _check_cores
    raise SystemError("The core '%s' has died unexpectedly" % core)
SystemError: The core 'CUDA-Device #2 'GeForce GTX 480'' has died unexpectedly

Exception in thread CPU-Core (SSE2):
Traceback (most recent call last):
  File "/usr/lib/python2.6/threading.py", line 532, in __bootstrap_inner
    self.run()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 109, in run
    essid, pwlist = self.queue._gather(self.buffersize, timeout=0.5)
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 640, in _gather
    self._check_cores()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 498, in _check_cores
    raise SystemError("The core '%s' has died unexpectedly" % core)
SystemError: The core 'CUDA-Device #2 'GeForce GTX 480'' has died unexpectedly

Exception in thread CPU-Core (SSE2):
Traceback (most recent call last):
  File "/usr/lib/python2.6/threading.py", line 532, in __bootstrap_inner
    self.run()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 109, in run
    essid, pwlist = self.queue._gather(self.buffersize, timeout=0.5)
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 640, in _gather
    self._check_cores()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 498, in _check_cores
    raise SystemError("The core '%s' has died unexpectedly" % core)
SystemError: The core 'CUDA-Device #2 'GeForce GTX 480'' has died unexpectedly

Exception in thread CPU-Core (SSE2):
Traceback (most recent call last):
  File "/usr/lib/python2.6/threading.py", line 532, in __bootstrap_inner
    self.run()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 109, in run
    essid, pwlist = self.queue._gather(self.buffersize, timeout=0.5)
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 640, in _gather
    self._check_cores()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 498, in _check_cores
    raise SystemError("The core '%s' has died unexpectedly" % core)
SystemError: The core 'CUDA-Device #2 'GeForce GTX 480'' has died unexpectedly

Exception in thread CUDA-Device #1 'GeForce GTX 480':
Traceback (most recent call last):
  File "/usr/lib/python2.6/threading.py", line 532, in __bootstrap_inner
    self.run()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 109, in run
    essid, pwlist = self.queue._gather(self.buffersize, timeout=0.5)
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 640, in _gather
    self._check_cores()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 498, in _check_cores
    raise SystemError("The core '%s' has died unexpectedly" % core)
SystemError: The core 'CUDA-Device #2 'GeForce GTX 480'' has died unexpectedly

Exception in thread CPU-Core (SSE2):
Traceback (most recent call last):
  File "/usr/lib/python2.6/threading.py", line 532, in __bootstrap_inner
    self.run()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 109, in run
    essid, pwlist = self.queue._gather(self.buffersize, timeout=0.5)
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 640, in _gather
    self._check_cores()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 498, in _check_cores
    raise SystemError("The core '%s' has died unexpectedly" % core)
SystemError: The core 'CUDA-Device #1 'GeForce GTX 480'' has died unexpectedly

Traceback (most recent call last):
  File "/usr/local/bin/pyrit", line 6, in <module>
    pyrit_cli.Pyrit_CLI().initFromArgv()
  File "/usr/local/lib/python2.6/dist-packages/pyrit_cli.py", line 110, in initFromArgv
    func(self, **options)
  File "/usr/local/lib/python2.6/dist-packages/pyrit_cli.py", line 903, in benchmark
    cp.enqueue('foo', ['barbarbar'] * 500)
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 572, in enqueue
    self._check_cores()
  File "/usr/local/lib/python2.6/dist-packages/cpyrit/cpyrit.py", line 498, in _check_cores
    raise SystemError("The core '%s' has died unexpectedly" % core)
SystemError: The core 'CUDA-Device #1 'GeForce GTX 480'' has died unexpectedly

Original comment by j...@ris-net.net on 21 Jul 2010 at 2:24

GoogleCodeExporter commented 8 years ago 
Thx Hazeman it worked great on my gtx470! 

Running benchmark (25236.5 PMKs/s)... /

Computed 25236.46 PMKs/s total.
#1: 'CUDA-Device #1 'GeForce GTX 470'': 23410.3 PMKs/s (RTT 2.8)
#2: 'CPU-Core (SSE2)': 548.3 PMKs/s (RTT 3.0)
#3: 'CPU-Core (SSE2)': 540.6 PMKs/s (RTT 3.0)
#4: 'CPU-Core (SSE2)': 535.8 PMKs/s (RTT 3.0)
#5: 'CPU-Core (SSE2)': 551.0 PMKs/s (RTT 3.0)
#6: 'CPU-Core (SSE2)': 524.9 PMKs/s (RTT 3.0)
#7: 'CPU-Core (SSE2)': 539.2 PMKs/s (RTT 3.0)
#8: 'CPU-Core (SSE2)': 580.1 PMKs/s (RTT 3.1)

Original comment by nrc...@gmail.com on 27 Jul 2010 at 7:08

GoogleCodeExporter commented 8 years ago 
So it looks like the issues are, so far, limited to the 480 cards... Is there 
anything particularly different between the two models? 

Any other updates on this issue?

Original comment by ryan...@gmail.com on 2 Aug 2010 at 4:26

GoogleCodeExporter commented 8 years ago 
can anyone provide me with ssh-access to a machine running such a gpu ?

Original comment by lukas.l...@gmail.com on 2 Aug 2010 at 4:32

GoogleCodeExporter commented 8 years ago 
As far as I know the 480 just has more cores. I will work on getting you SSH 
access to the machine later today if you want.

Original comment by sprice82 on 2 Aug 2010 at 5:38

GoogleCodeExporter commented 8 years ago 
@28, please get in contact with me by lukas.lueg@googlemail.com then...

Original comment by lukas.l...@gmail.com on 2 Aug 2010 at 5:51

GoogleCodeExporter commented 8 years ago 
lukas,

I can't get you SSH access, but I'm happy to send you output of any 
commands/info you need. If so, just email me with the info you need.

Original comment by ryan...@gmail.com on 5 Aug 2010 at 12:00

GoogleCodeExporter commented 8 years ago 
I had to relocate some of my servers, so it will be a couple more days before I 
can get you access, sorry

Original comment by sprice82 on 5 Aug 2010 at 12:51

GoogleCodeExporter commented 8 years ago 
I just sent you an email.. im sure I can work out access to a 480 SLI machine 
as soon as you want it.

Original comment by j...@ris-net.net on 6 Aug 2010 at 2:40

GoogleCodeExporter commented 8 years ago 
As of svn-r277 the device-code is compiled to PTX-code instead of going 
directly to CUBIN. Please try the most recent svn-version...

Original comment by lukas.l...@gmail.com on 7 Aug 2010 at 6:21

GoogleCodeExporter commented 8 years ago 
You are the man! I just svn updated, did a clean build of pyrit and 
cpyrit_cuda. Both list_cores and benchmark commands are running fine so far. I 
will do some more testing in a little while when I have some more time.

Thanks again!

Original comment by ryan...@gmail.com on 7 Aug 2010 at 7:05

GoogleCodeExporter commented 8 years ago 
@ryanlev, are you on 64bit or 32bit ?

Original comment by lukas.l...@gmail.com on 7 Aug 2010 at 7:10

GoogleCodeExporter commented 8 years ago 
64bit Mint 9 Kernel 2.6.35 and nvidia drivers 256.44

Original comment by ryan...@gmail.com on 7 Aug 2010 at 7:23

GoogleCodeExporter commented 8 years ago

Original comment by lukas.l...@gmail.com on 9 Aug 2010 at 6:10

GoogleCodeExporter commented 8 years ago 
I had the same issue. Using the latest SVN version solved it. I have incredible 
results with 8 C2050 Tesla boards :

Pyrit 0.3.0 (C) 2008-2010 Lukas Lueg http://pyrit.googlecode.com
This code is distributed under the GNU General Public License v3+

Running benchmark (174037.5 PMKs/s)... | 

Computed 174349.13 PMKs/s total.
#1: 'CUDA-Device #1 'Tesla C2050'': 21719.6 PMKs/s (RTT 1.6)
#2: 'CUDA-Device #2 'Tesla C2050'': 21437.7 PMKs/s (RTT 1.5)
#3: 'CUDA-Device #3 'Tesla C2050'': 21501.0 PMKs/s (RTT 1.5)
#4: 'CUDA-Device #4 'Tesla C2050'': 21698.4 PMKs/s (RTT 1.6)
#5: 'CUDA-Device #5 'Tesla C2050'': 21745.1 PMKs/s (RTT 1.4)
#6: 'CUDA-Device #6 'Tesla C2050'': 21957.6 PMKs/s (RTT 1.7)
#7: 'CUDA-Device #7 'Tesla C2050'': 21443.8 PMKs/s (RTT 1.4)
#8: 'CUDA-Device #8 'Tesla C2050'': 21755.8 PMKs/s (RTT 1.6)
#9: 'CPU-Core (SSE2)': 762.4 PMKs/s (RTT 2.9)
#10: 'CPU-Core (SSE2)': 771.6 PMKs/s (RTT 2.9)
#11: 'CPU-Core (SSE2)': 772.8 PMKs/s (RTT 2.9)
#12: 'CPU-Core (SSE2)': 767.2 PMKs/s (RTT 2.9)
#13: 'Network-Clients': 0.0 PMKs/s (RTT 0.0)

Original comment by JeanDami...@gmail.com on 13 Oct 2010 at 3:03

GoogleCodeExporter commented 8 years ago 
how the hell can you afford 8 c2050 boards? :)

Original comment by lukas.l...@gmail.com on 14 Oct 2010 at 10:33

GoogleCodeExporter commented 8 years ago 
We're currently assessing the power of GPU optimized applications (to do 
bio-info at the end). I've choosen to use pyrit to benchmark the different 
hardware configuration easily. This server has been provided by an NVIDIA 
partner.

Original comment by JeanDami...@gmail.com on 14 Oct 2010 at 2:38

GoogleCodeExporter commented 8 years ago 
Pogolotti,
Of course, this think is for sure valid only talking about pyrit, but take in 
mind that a single ATI 5870 is able to run about 48000-83000 PMK/s (if you use 
openCL or calpp). So, maybe it will be good for you to test also ATI hardware, 
it could be cheaper to buy and less expencive under electrical power viewpoint.

Original comment by pyrit.lo...@gmail.com on 17 Oct 2010 at 1:45

GoogleCodeExporter commented 8 years ago 
Was the result for 8 teslas obtained with pyrit benchmark_long ? 

Results returned by 'pyrit benchmark' are really inaccurate.

Original comment by hazema...@gmail.com on 19 Oct 2010 at 12:17

GoogleCodeExporter commented 8 years ago 
Pogolotti, don't trust benchmark/benchmark_long, try also a REAL test as: 

pyrit -e YOUR_ESSID -r HANDSHAKE.cap -i PASSWORD_LIST.txt attack_passthrough

because so probably you will have bad surprise: pyrit has issue in support 
multiple GPUS, have a look to issue 173.

Original comment by pyrit.lo...@gmail.com on 19 Oct 2010 at 9:33