Tasks getting "Timed out - no response" errors

Message boards : Number crunching : Tasks getting "Timed out - no response" errors
Message board moderation

To post messages, you must log in.

Previous · 1 · 2

AuthorMessage
rob smith Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer moderator
Volunteer tester

Send message
Joined: 7 Mar 03
Posts: 22160
Credit: 416,307,556
RAC: 380
United Kingdom
Message 1364846 - Posted: 5 May 2013, 17:23:45 UTC

My error rate on the gtx460 was certainly higher with earlier versions of the driver than with 314.22, but since I updated to r1761 it is hard to say which made the improvement.
Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?
ID: 1364846 · Report as offensive
Profile Fred E.
Volunteer tester

Send message
Joined: 22 Jul 99
Posts: 768
Credit: 24,140,697
RAC: 0
United States
Message 1364857 - Posted: 5 May 2013, 17:41:32 UTC

Let's see if the driver update fixes it. It's hard to say what's going on with the AP tasks with just one to look at. But, you're also getting a higher than normal inconclusive rate on the stock 6.10 gpu application as well as the optimized version that you stopped running, so you need to watch that to see if there is improvement.
Another Fred
Support SETI@home when you search the Web with GoodSearch or shop online with GoodShop.
ID: 1364857 · Report as offensive
ACP

Send message
Joined: 28 Mar 13
Posts: 31
Credit: 148,954
RAC: 0
United States
Message 1364877 - Posted: 5 May 2013, 18:08:20 UTC - in response to Message 1364845.  

That is where i download from.. My GPU software is linked to the website. All i did was clicked on "check for update's"
ID: 1364877 · Report as offensive
TBar
Volunteer tester

Send message
Joined: 22 May 99
Posts: 5204
Credit: 840,779,836
RAC: 2,768
United States
Message 1364884 - Posted: 5 May 2013, 18:26:00 UTC - in response to Message 1364841.  
Last modified: 5 May 2013, 18:42:39 UTC

I was not aware there was an update version for my GPU I just updated it. And about the Lunatic's program. I know some people don't have issues with it where as others may have, ever since i installed that program i had problems with BOINC Manager one after another. As it is right now i am doing just fine without the program, better even.

Are you talking about the two that are shown in progress? If so, do they run to completion?


It was ALL the tasks that was downloaded to do.. It was right after reinstalling Boinc Manager after the reinstall of Windows 7

AstroPulse issue is i get this every time it finish's the task. Was it due to the outdated GPU driver?

Error while computing 7,051.12 848.97 --- AstroPulse v6 v6.04 (opencl_nvidia_100)

See this thread; OpenCL AstroPulse crash after processing completion
You can probably avoid most of those Access Violation (0xc0000005) Errors by adding parameters to your ap_cmdline_6.04_windows_intelx86__opencl_nvidia.txt file.
ID: 1364884 · Report as offensive
ACP

Send message
Joined: 28 Mar 13
Posts: 31
Credit: 148,954
RAC: 0
United States
Message 1364997 - Posted: 5 May 2013, 23:51:07 UTC - in response to Message 1364884.  

Those manual installs is not my thing. I don't trust my self to do them.. not a programer even for copy & paste. I'll stay with the irritating issue i have unless it's an install type app.
ID: 1364997 · Report as offensive
TBar
Volunteer tester

Send message
Joined: 22 May 99
Posts: 5204
Credit: 840,779,836
RAC: 2,768
United States
Message 1364999 - Posted: 6 May 2013, 0:48:10 UTC - in response to Message 1364997.  

Those manual installs is not my thing. I don't trust my self to do them.. not a programer even for copy & paste. I'll stay with the irritating issue i have unless it's an install type app.

There isn't any type of Install. All you are doing is adding a line of text to a text file that has already been installed. The text file was installed, by default, for the sole purpose of adding such lines of text. They can't add the correct line of text during the install because it is different for every class of GPU. YOU have to add the correct line of text for YOUR GPU. Sorry, there isn't any other way. You add the correct line to the file or, it doesn't happen. The instructions are in every new App package released, the most recent is here, Current OpenCL apps are available for download on Lunatics again
    Command Line Parameters.
    High end cards (more than 12 compute units)
    -unroll 12 -ffa_block 8192 -ffa_block_fetch 4096 -hp

    Mid range cards (less than 12 compute units)
    -unroll 10 -ffa_block 6144 -ffa_block_fetch 1536 -hp

    Entry level GPU (less than 6 compute units)
    -unroll 4 -ffa_block 2048 -ffa_block_fetch 1024 -hp


All You have to do is add the correct line to your already present ap_cmdline_6.04_windows_intelx86__opencl_nvidia.txt File, or not.

ID: 1364999 · Report as offensive
bill

Send message
Joined: 16 Jun 99
Posts: 861
Credit: 29,352,955
RAC: 0
United States
Message 1365006 - Posted: 6 May 2013, 2:23:25 UTC

So what is a 'compute unit'?
ID: 1365006 · Report as offensive
Profile Mike Special Project $75 donor
Volunteer tester
Avatar

Send message
Joined: 17 Feb 01
Posts: 34253
Credit: 79,922,639
RAC: 80
Germany
Message 1365057 - Posted: 6 May 2013, 7:15:37 UTC - in response to Message 1365006.  

So what is a 'compute unit'?



Each GPU device (AMD, NVidea, or any other) is split into several Compute Units (MultiProcessors), each of which has a fixed number of cores (VertexShaders/StreamProcessors). So, one has (Compute Units) x (VertexShaders/compute unit) simultaneous processors to compute with, but there is only a small fixed amount of __local memory (usually 16KB or 32KB) available per MultiProcessor. Hence, the exact number of these multiprocessors matters.



With each crime and every kindness we birth our future.
ID: 1365057 · Report as offensive
bill

Send message
Joined: 16 Jun 99
Posts: 861
Credit: 29,352,955
RAC: 0
United States
Message 1365064 - Posted: 6 May 2013, 8:17:09 UTC - in response to Message 1365057.  

And the number of compute units for a specific card
are found where? GPZ? Documentation? Nothing I'm coming
up with says "X card has Y compute units."
ID: 1365064 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 5 Jul 99
Posts: 4654
Credit: 47,537,079
RAC: 4
United Kingdom
Message 1365065 - Posted: 6 May 2013, 8:26:45 UTC - in response to Message 1365064.  

And the number of compute units for a specific card
are found where? GPZ? Documentation? Nothing I'm coming
up with says "X card has Y compute units."

From your stderr_txt from one of your completed Wu's.

Claggy
ID: 1365065 · Report as offensive
Profile Mike Special Project $75 donor
Volunteer tester
Avatar

Send message
Joined: 17 Feb 01
Posts: 34253
Credit: 79,922,639
RAC: 80
Germany
Message 1365066 - Posted: 6 May 2013, 8:55:20 UTC - in response to Message 1365064.  

And the number of compute units for a specific card
are found where? GPZ? Documentation? Nothing I'm coming
up with says "X card has Y compute units."


Your card for example has 8 compute units.



With each crime and every kindness we birth our future.
ID: 1365066 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 5 Jul 99
Posts: 4654
Credit: 47,537,079
RAC: 4
United Kingdom
Message 1365072 - Posted: 6 May 2013, 9:21:05 UTC - in response to Message 1365066.  

For Example:

For Cuda:

setiathome_CUDA: Found 2 CUDA device(s):
Device 1: GeForce GTX 560 Ti, 1024 MiB, regsPerBlock 32768
computeCap 2.1, multiProcs 8
pciBusID = 1, pciSlotID = 0
clockRate = 1900 MHz
Device 2: GeForce GTX 560 Ti, 1024 MiB, regsPerBlock 32768
computeCap 2.1, multiProcs 8
pciBusID = 2, pciSlotID = 0
clockRate = 1800 MHz
In cudaAcc_initializeDevice(): Boinc passed DevPref 1
setiathome_CUDA: CUDA Device 1 specified, checking...
Device 1: GeForce GTX 560 Ti is okay
SETI@home using CUDA accelerated device GeForce GTX 560 Ti
pulsefind: blocks per SM 4 (Fermi or newer default)
pulsefind: periods per launch 100 (default)
Priority of process set to BELOW_NORMAL (default) successfully
Priority of worker thread set successfully

setiathome enhanced x41zc, Cuda 4.20


For OpenCL:

Running on device number: 1
Priority of worker thread raised successfully
Priority of process adjusted successfully, below normal priority class used
OpenCL platform detected: NVIDIA Corporation
BOINC assigns device 1
Info: BOINC provided device ID used
Used GPU device parameters are:
Number of compute units: 8
Single buffer allocation size: 256MB
max WG size: 1024
FERMI path used: yes

Build features: Non-graphics OpenCL USE_OPENCL_NV OCL_ZERO_COPY COMBINED_DECHIRP_KERNEL FFTW USE_INCREASED_PRECISION USE_SSE2 x86
CPUID: Intel(R) Core(TM)2 Quad CPU Q9550 @ 2.83GHz

Cache: L1=64K L2=6144K

CPU features: FPU TSC PAE CMPXCHG8B APIC SYSENTER MTRR CMOV/CCMP MMX FXSAVE/FXRSTOR SSE SSE2 HT SSE3
AstroPulse v.6
Non-graphics FFTW USE_CONVERSION_OPT
Windows x86 rev 1761, V6 match, by Raistmer with support of Lunatics.kwsn.net team. SSE2

OpenCL version by Raistmer

ffa threshold mods by Joe Segur
SSE3 dechirping by JDWhale
Combined dechirp kernel by Frizz
Number of OpenCL platforms: 1


OpenCL Platform Name: NVIDIA CUDA
Number of devices: 2
Max compute units: 8
Max work group size: 1024
Max clock frequency: 1900Mhz
Max memory allocation: 268435456
Cache type: Read/Write
Cache line size: 128
Cache size: 131072
Global memory size: 1073741824
Constant buffer size: 65536
Max number of constant args: 9
Local memory type: Scratchpad
Local memory size: 49152
Queue properties:
Out-of-Order: Yes
Name: GeForce GTX 560 Ti
Vendor: NVIDIA Corporation
Driver version: 310.70
Version: OpenCL 1.1 CUDA
Extensions: cl_khr_byte_addressable_store cl_khr_icd cl_khr_gl_sharing cl_nv_d3d9_sharing cl_nv_d3d10_sharing cl_khr_d3d10_sharing cl_nv_d3d11_sharing cl_nv_compiler_options cl_nv_device_attribute_query cl_nv_pragma_unroll cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_fp64
Max compute units: 8
Max work group size: 1024
Max clock frequency: 1800Mhz
Max memory allocation: 268435456
Cache type: Read/Write
Cache line size: 128
Cache size: 131072
Global memory size: 1073741824
Constant buffer size: 65536
Max number of constant args: 9
Local memory type: Scratchpad
Local memory size: 49152
Queue properties:
Out-of-Order: Yes

Name: GeForce GTX 560 Ti
Vendor: NVIDIA Corporation
Driver version: 310.70
Version: OpenCL 1.1 CUDA
Extensions: cl_khr_byte_addressable_store cl_khr_icd cl_khr_gl_sharing cl_nv_d3d9_sharing cl_nv_d3d10_sharing cl_khr_d3d10_sharing cl_nv_d3d11_sharing cl_nv_compiler_options cl_nv_device_attribute_query cl_nv_pragma_unroll cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_fp64


Legacy setiathome_enhanced V6 mode.


Claggy
ID: 1365072 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14649
Credit: 200,643,578
RAC: 874
United Kingdom
Message 1365080 - Posted: 6 May 2013, 9:54:22 UTC

There are a couple of useful Wiki pages:

Comparison of Nvidia graphics processing units
Comparison of AMD graphics processing units

The Nvidia (cuda) page lists "SM Count - Number of streaming multiprocessors" for each card, but whoever maintains the AMD page doesn't seem to bother with a compute unit count.
ID: 1365080 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14649
Credit: 200,643,578
RAC: 874
United Kingdom
Message 1365103 - Posted: 6 May 2013, 11:47:47 UTC - in response to Message 1365080.  

Too late to edit, but I've had a PM to say that the compute unit count can be found on the AMD wiki page:

The column "Config core" has three numbers separated by colons; Unified Shaders : Texture mapping unit : Render Output Unit (ROP).

It is the third number - the ROP - that you're looking for.
ID: 1365103 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 5 Jul 99
Posts: 4654
Credit: 47,537,079
RAC: 4
United Kingdom
Message 1365106 - Posted: 6 May 2013, 11:59:27 UTC - in response to Message 1365103.  

Too late to edit, but I've had a PM to say that the compute unit count can be found on the AMD wiki page:

The column "Config core" has three numbers separated by colons; Unified Shaders : Texture mapping unit : Render Output Unit (ROP).

It is the third number - the ROP - that you're looking for.

The AMD wiki and GPU-Z might say 16 ROPs for my HD7770, but OpenCL says 10 Compute Units, so they might not match.

Claggy

ID: 1365106 · Report as offensive
TBar
Volunteer tester

Send message
Joined: 22 May 99
Posts: 5204
Credit: 840,779,836
RAC: 2,768
United States
Message 1365164 - Posted: 6 May 2013, 16:25:07 UTC

There are a few exceptions, especially with the older GPUs. OpenCL says my NV8800 has 14 Compute units. On those occasions when I run an AP on that card, anything over an unroll of 6 causes the entire computer to become unresponsive. Even though I'm not using the monitor attached to the 8800. Using an unroll of 6, the GPU load is maxed out anyway, so, there isn't any advantage to using a higher setting. The other older card, the ATI 4670, is listed at 8 Compute units. The 4670 is maxed out using an unroll of 4, and I think I even get warning notices with a much higher setting. That's covered by the line 'Your mileage might vary' in the Readme. That's another reason those Command Lines are not set by the Installer.
ID: 1365164 · Report as offensive
bill

Send message
Joined: 16 Jun 99
Posts: 861
Credit: 29,352,955
RAC: 0
United States
Message 1365279 - Posted: 6 May 2013, 21:46:06 UTC

Thank you everyone.

I'm sure these answers will be as helpful
to others as they have been to me.

Again thanks.
ID: 1365279 · Report as offensive
Previous · 1 · 2

Message boards : Number crunching : Tasks getting "Timed out - no response" errors


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.