Hey all, maybe someone can look at my computations and let me know if I am doing something wrong, or what I can do to fix this, everytime I do a GPU Grid WU as of late, they all have computation errors.

Used to have a 8600gt Nvidia card in my AMD Dual core,and it was happening, and I now even upgraded to a GTX 260, newest drivers, current boinc, and they are still failing...any suggestions, ideas?

Thanks in advance...

---

try installing driver 185.85... 186.xx has given me trouble.... try that...

---

I recommend Secunia PSI: http://secunia.com/vulnerability_scanning/personal/

Finished 2 more work units, both computation errors...any other suggestions, what can be going wrong...I have checked my drivers, I have a GTX 250 why is it always failing?

Thanks to everyone in advance for your time...

ideas?

medictom

---

On Windows 32bit, it seems that you need drivers 181.xx.
The new drivers don't work.
gdf

Frustrated. Two more computation errors...

---

Still no success. Using driver 181.22 and still failing all wus. Same error -
7/24/2009 6:02:48 AM GPUGRID Computation for task 8-KASHIF_HIVPR_sub_so_ba1-9-100-RND9354_0 finished
7/24/2009 6:02:48 AM GPUGRID Output file 8-KASHIF_HIVPR_sub_so_ba1-9-100-RND9354_0_1 for task 8-KASHIF_HIVPR_sub_so_ba1-9-100-RND9354_0 absent
7/24/2009 6:02:48 AM GPUGRID Output file 8-KASHIF_HIVPR_sub_so_ba1-9-100-RND9354_0_2 for task 8-KASHIF_HIVPR_sub_so_ba1-9-100-RND9354_0 absent
7/24/2009 6:02:48 AM GPUGRID Output file 8-KASHIF_HIVPR_sub_so_ba1-9-100-RND9354_0_3 for task 8-KASHIF_HIVPR_sub_so_ba1-9-100-RND9354_0 absent

Setting NNT until this is sorted out. (Windows XP 32 bit, GTX 260 Core 216)

Frustrated. Two more computation errors...

190 works fine for me in a GTX260 in a Q6600 with Win XP 32 bits.

I think, could be your BOINC version... yo're using 6.6.36.

Do you remember what BOINC version you used when all worked fine?

Mine: 260 (192) 190 drivers qx9650 x48 vista64

I get driver nvlddmkm failed and recovered which in turn causes computation error running stock clocks.

Other machine on xp and 98gtx and 182.08 no errors

using 6.6.36 as it was advised in the messages file

Thoughts anyone?

I'm having this exact same problem -- all WUs end in "Computational error" from the very first WU I downloaded.

Here is my first WU result:

<core_client_version>6.6.36</core_client_version>
<![CDATA[
<message>
Incorrect function. (0x1) - exit code 1 (0x1)
</message>
<stderr_txt>
# Using CUDA device 0
# Device 0: "GeForce GTX 260"
# Clock rate: 1350000 kilohertz
# Total amount of global memory: 939524096 bytes
# Number of multiprocessors: 27
# Number of cores: 216
# Amber: readparm : Reading parm file parameters
# PARM file in AMBER 7 format
# Encounter 10-12 H-bond term
WARNING: parameters.cu, line 568: Found zero 10-12 H-bond term.
WARNING: parameters.cu, line 568: Found zero 10-12 H-bond term.
MDIO ERROR: cannot open file "restart.coor"
Cuda error: Kernel [fft_data_swizzle_in] failed in file 'c:\cygwin\home\speechserver\gpumd2\src\pme\CPME_cufft.cu' in line 44 : unknown error.

I'm running a single EVGA GTX 260 Superclocked Edition (190.xx driver) on a EVGA NForce 750i SLI FTW motherboard, Intel Q9300 cpu, 8Gb DDR2-800 memory. Everything is running at stock -- nothing is tweeked! OS: Vista64 Premium SP2

---

Let me think...

What thing I did different than you...?

Uhmmm.... I installed CUDA 2.2 Toolkit.


Try it.