Forum > GPU crunching
Latest drivers (NVidia and ATI)
Urs Echternacht:
--- Quote from: Raistmer on 01 Nov 2011, 10:43:11 pm ---We need to locate place of failure. I've seen the same with windows Cat preview.
still unknown where exactly data start to differ between driver versions.
Please, post your findings in corresponding thread on AMD forum too.
http://forums.amd.com/devforum/messageview.cfm?catid=390&threadid=155591&enterthread=y
--- End quote ---
Added my findings over there.
Fredericx51:
I tried CAT 11.10, immediatly produced errors and driver crasht, so I strait went back with CAT 11.4.1332 ,
last ATI 5870S GPUs WU ;
<core_client_version>6.12.34</core_client_version>
<![CDATA[
<stderr_txt>
Number of period iterations for PulseFind setted to:1
Number of app instances per device setted to:2
Running on device number: 0
Priority of worker thread raised successfully
Priority of process adjusted successfully, below normal priority class used
OpenCL platform detected: Advanced Micro Devices, Inc.
BOINC assigns 0 device, slots 0 to 1 (including) will be checked
Used slot is 0; INFO: clCreateContext
OpenCL-kernels filename : MultiBeam_Kernels_r365.cl
Info : Building Program (clBuildProgram):main kernels: OK code 0
Windows optimized S@H Enhanced application by Alex Kan
Version info: SSE3x (AMD/Intel, Core 2-optimized v8-nographics) V5.13 by Alex Kan
SSE3x Win32 Build 365 , Ported by : Jason G, Raistmer, JDWhale
SETI7 update by Raistmer
Original GPU DCT by Jason G
OpenCL version by Raistmer, r365
Build features: SETI7 Non-graphics OpenCL USE_OPENCL_HD5xxx IPP AMD specific USE_SSE3 x86
CPUID: Intel(R) Core(TM) i7-2600 CPU @ 3.40GHz
Speed: 4 x 3413 MHz
Cache: L1=64K L2=256K
CPU features: FPU TSC PAE CMPXCHG8B APIC SYSENTER MTRR CMOV/CCMP MMX FXSAVE/FXRSTOR SSE SSE2 HT SSE3
CPU type 0x46
Number of OpenCL platforms: 1
OpenCL Platform Name: AMD Accelerated Parallel Processing
Number of devices: 2
Max compute units: 20
Max work group size: 256
Max clock frequency: 850Mhz
Max memory allocation: 134217728
Cache type: None
Cache line size: 0
Cache size: 0
Global memory size: 536870912
Constant buffer size: 65536
Max number of constant args: 8
Local memory type: Scratchpad
Local memory size: 32768
Queue properties:
Out-of-Order: No
Name: Cypress
Vendor: Advanced Micro Devices, Inc.
Driver version: CAL 1.4.1332
Version: OpenCL 1.1 AMD-APP-SDK-v2.4 (595.10)
Extensions: cl_amd_fp64 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_3d_image_writes cl_khr_byte_addressable_store cl_khr_gl_sharing cl_amd_device_attribute_query cl_amd_printf cl_amd_media_ops cl_amd_popcnt cl_khr_d3d10_sharing
Max compute units: 20
Max work group size: 256
Max clock frequency: 850Mhz
Max memory allocation: 134217728
Cache type: None
Cache line size: 0
Cache size: 0
Global memory size: 536870912
Constant buffer size: 65536
Max number of constant args: 8
Local memory type: Scratchpad
Local memory size: 32768
Queue properties:
Out-of-Order: No
Name: Cypress
Vendor: Advanced Micro Devices, Inc.
Driver version: CAL 1.4.1332
Version: OpenCL 1.1 AMD-APP-SDK-v2.4 (595.10)
Extensions: cl_amd_fp64 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_3d_image_writes cl_khr_byte_addressable_store cl_khr_gl_sharing cl_amd_device_attribute_query cl_amd_printf cl_amd_media_ops cl_amd_popcnt cl_khr_d3d10_sharing
Work Unit Info:
...............
Credit multiplier is : 2.85
WU true angle range is : 0.009569
New triplet to report:
score=10.53; power=10.53; freq_bin=-1.367e-121
New pulse to report: score=1.001, power=6.7813, fftlen=256, freq_bin=145, time_bin=2048
New pulse to report: score=1.037, power=4.5921, fftlen=1024, freq_bin=590, time_bin=512
New pulse to report: score=1.022, power=3.2453, fftlen=2048, freq_bin=1632, time_bin=256
New triplet to report:
score=11.2; power=11.2; freq_bin=-3.534e-047
New triplet to report:
score=10.54; power=10.54; freq_bin=-3.541e-194
New pulse to report: score=1.052, power=6.034, fftlen=512, freq_bin=71, time_bin=1024
New pulse to report: score=1.003, power=2.199, fftlen=2048, freq_bin=395, time_bin=256
New pulse to report: score=1.025, power=4.8706, fftlen=128, freq_bin=126, time_bin=4096
New pulse to report: score=1.01, power=2.6588, fftlen=1024, freq_bin=519, time_bin=512
New pulse to report: score=1.028, power=10.489, fftlen=512, freq_bin=221, time_bin=1024
New pulse to report: score=1.019, power=10.13, fftlen=1024, freq_bin=412, time_bin=512
Flopcounter: 17723986648026.246000
Spike count: 0
Autocorr count: 0
Pulse count: 9
Triplet count: 3
Gaussian count: 0
Wallclock time elapsed since last restart: 5072.1 seconds
class Gaussian_transfer_not_needed: total=0, N=0, <>=0, min=0 max=0
class Gaussian_transfer_needed: total=0, N=0, <>=0, min=0 max=0
class Gaussian_skip1_no_peak: total=0, N=0, <>=0, min=0 max=0
class Gaussian_skip2_bad_group_peak: total=0, N=0, <>=0, min=0 max=0
class Gaussian_skip3_too_weak_peak: total=0, N=0, <>=0, min=0 max=0
class Gaussian_skip4_too_big_ChiSq: total=0, N=0, <>=0, min=0 max=0
class Gaussian_skip6_low_power: total=0, N=0, <>=0, min=0 max=0
class Gaussian_new_best: total=0, N=0, <>=0, min=0 max=0
class Gaussian_report: total=0, N=0, <>=0, min=0 max=0
class Gaussian_miss: total=0, N=0, <>=0, min=0 max=0
class PC_triplet_find_hit: total=46928, N=46928, <>=1, min=1 max=1
class PC_triplet_find_miss: total=957, N=957, <>=1, min=1 max=1
class PC_pulse_find_hit: total=47871, N=47871, <>=1, min=1 max=1
class PC_pulse_find_miss: total=14, N=14, <>=1, min=1 max=1
class PC_pulse_find_2CPU: total=0, N=0, <>=0, min=0 max=0
class PoT_transfer_not_needed: total=46917, N=46917, <>=1, min=1 max=1
class PoT_transfer_needed: total=968, N=968, <>=1, min=1 max=1
19:30:15 (7060): called boinc_finish
</stderr_txt>
]]>
But this is older version and the problem is amount of CPU use when using GPUs and the latest driver, atleast above CAT 11.4?
perryjay:
Well, I went back to driver 267.59. I have been getting some downclocks with the newest driver 285.62. I'm also getting some invalids with high pulse counts both in the AP and MB work. I can't swear it's the driver as I also started using the rescheduler again to move things around so that might have been causing the problems. I've also been watching some videos. I'll see what happens with the older driver as it was working fine for me before updating the driver.
As always, I have my Intel E5400 dual with my GTS 450 slightly overclocked to 883/1766/1804 running one AP and one MB or two MBs at a time. (count set at .51/.49)
Oh, and I'm running 6.10.58 64bit and x39e Lunatics along with Raistmer's open CL for NVIDIA R521.
Mike:
I wouldn´t wonder perryjay.
I imagine nvidia is going thru the same issues AMD did month ago and still has.
Frizz:
I just wonder why both companies have similar issues (well, at least they *look* similar) at the same time. Maybe it's because they already add support for the next architectures (Kepler and GCN) in their drivers - and break their old code.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version