Just so you'll know, my utility reports that I downclocked the card. GPU-Z reports that I downclocked the card. All of the results, Pending and Valid, show the card is still at the factory speed.

Yes, it is at factory speed.

Did you try MSI Afterburner ?
Its the easiest way to do.

You can save a few speedsets and decide which to use afterwards.

---

With each crime and every kindness we birth our future.

Just so you'll know, my utility reports that I downclocked the card. GPU-Z reports that I downclocked the card. All of the results, Pending and Valid, show the card is still at the factory speed.

um - did you restart BOINC? Maybe it saves whatever the settings were at the start and that part of the stderr output isn't freshly generated with each task?

I wouldn't know, just a thought, if output disagrees.

Host
3324167 is making a lot of faults, -1 exit.

WU ID
840393334.

Host
5256028 is making -9 (informational overflow)

Both running CUDA 2.3.

WU ID
843782391.

Host
3324167. Found triplets twice, Invalid (241) | Fault (63)

CUDA & CUDA 2.3

My host,
having a bit much restarts, probably Collatz C. throwing WUs with a 1 hour
deadline, running High Priority.

---

No I didn't use MSI Afterburner, I used what came with the card.

Are we saying that GPU-Z may be telling lies?

I'll try it in a few minutes.... sheeesh.

No I didn't use MSI Afterburner, I used what came with the card.

Are we saying that GPU-Z may be telling lies?

I'll try it in a few minutes.... sheeesh.

No, It's just that MSI Afterburner is able to do much more then just read the cards...

---

The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's

Host
5256028 too much false -9 overflows.

Host
3324167 , also has too much false -9, >300 and
Cuda error 'find_pulse_kernel2<3, false>' in file 'c:/sw/gpgpu/seti/seti_boinc/client/cuda/cudaAcc_pulsefind.cu' in line 1201 : unknown error.
Cuda error 'find_pulse_kernel2<4, true>' in file 'c:/sw/gpgpu/seti/seti_boinc/client/cuda/cudaAcc_pulsefind.cu' in line 1229 : unknown error.

Card getting too hot? Dust?
Both running CUDA and CUDA 2.3 and are/were my wingmen, while checking my ATI host.
One of my host with a GTS250, has 280.26 and reBOOTs, for now apparent reason.
I noticed a newer driver is available, gonna try it. It hasn't made mistakes, yet!

---

No I didn't use MSI Afterburner, I used what came with the card.

Are we saying that GPU-Z may be telling lies?

I'll try it in a few minutes.... sheeesh.

No, like Vic said its perfect for just read the card.
With Afterburner you can set all core speeds manually and save them to profiles.

So i would adjust 1 profile with stock speeds and 2 - 3 with different overclocked speeds.
Choose a profile and click adjust overclocking at startup.
One card i readjusted this way is running now with near the factory settings.

---

With each crime and every kindness we birth our future.

I've all of my cards running stock speed, but 1 9800GTX and 1 8500GT are damaged
in some way and were taken out, so the GTS 250, is going to be downclocked, or switched for a FERMI.
If lowering Core-Speed, doesn't have enough effect. (And if I've the $$, for this.)

Afterburner, can come in handy, since this is the only way to decrease
clock-speed.
Has to be driver related, since it crashed several times, only with newer driver.
About GPUz, on 2, with Cross-Fire Bridge, HD5870, 1st GPU is always on 100%
and 2nd gives more real data. Sometimes, after restarting BOINC, it'll give proper
readings.

---

Well... crud.

I rechecked GPU-Z. It repeated what I already knew. The card was downclocked to nVidia specs.

Ok, I restarted the computer (again), used Gigabyte's overclocking tool to downclock (again) and updated and then ran GPU-Z (again) and it shows the slower speeds (the nVidia defaults, not the factory overclock).

I'm going out of town and don't want it to sit there and do what it is doing. I've shut it down with 1100 WUs Pending, so this ought to be fun.

I checked some of the Pendings and they are full of -9s.

I'm frustrated.

New motherboard, CPU, and RAM will be waiting for me when I return. If something's making noise it may not be the card, and the BIOS of this old board isn't giving me a lot of options. I'm going to do everything I can think of to rule out everything else, then I'm going to yell at Gigabyte.

The motherboard and RAM are old. The Phenom x 4 was a "new" CPU. The power supply is a PC P&C 750, less than a full year old.

We'll see what we see soon enough. In the meantime, at least it isn't making any more errors.

If it starts kicking-out errors after all this upgrading everything else I may see if Gigabyte will take the card in trade for a GTX 460.

Based on my first Gigabyte overclocked 560ti (solid from day one), I thought everyone having trouble with the 560tis were abusing them in some way or maybe this was only an EVGA problem. I don't think that anymore.

This is how a 'normal' overflow looks like with ATI rev177 and on Darwin (APPLE), on CPU, both are valid.

<stderr_txt>
Number of period iterations for PulseFind setted to:1
Number of app instances per device setted to:2
Running on device number: 1
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 1 device, slots 2 to 3 (including) will be checked
Used slot is 2; Info : Building Program (clBuildProgram):main kernels: OK code 0

ar=0.778301 NumCfft=150263 NumGauss= 607305872 NumPulse= 65352439296 NumTriplet= 9094714425344
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 177 , Ported by : Jason G, Raistmer, JDWhale

OpenCL version by Raistmer, rev177

Build features: 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 3386 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 : 0.778301
SETI@Home Informational message -9 result_overflow
NOTE: The number of results detected exceeds the storage space allocated.

Flopcounter: 5527315516.363931

Spike count: 30
Pulse count: 0
Triplet count: 0
Gaussian count: 0
Wallclock time elapsed since last restart: 24.0 seconds

class Gaussian_transfer_not_needed: total=8, N=8, <>=1, min=1 max=1
class Gaussian_transfer_needed: total=4, N=4, <>=1, min=1 max=1


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=88, <>=0, min=0 max=0
class Gaussian_skip6_low_power: total=68, N=88, <>=0.77272725, min=0 max=1


class Gaussian_new_best: total=15, N=15, <>=1, min=1 max=1
class Gaussian_report: total=0, N=0, <>=0, min=0 max=0
class Gaussian_miss: total=73, N=73, <>=1, min=1 max=1


class PC_triplet_find_hit: total=7, N=7, <>=1, min=1 max=1
class PC_triplet_find_miss: total=2, N=2, <>=1, min=1 max=1


class PC_pulse_find_hit: total=9, N=9, <>=1, min=1 max=1
class PC_pulse_find_miss: total=0, N=0, <>=0, min=0 max=0
class PC_pulse_find_2CPU: total=1, N=1, <>=1, min=1 max=1


class PoT_transfer_not_needed: total=7, N=7, <>=1, min=1 max=1
class PoT_transfer_needed: total=3, N=3, <>=1, min=1 max=1

class T_FFT<8>: total=1, N=1, <>=1, min=1 max=1
class T_FFT<16>: total=1, N=1, <>=1, min=1 max=1
class T_FFT<32>: total=1, N=1, <>=1, min=1 max=1
class T_FFT<64>: total=1, N=1, <>=1, min=1 max=1
class T_FFT<128>: total=1, N=1, <>=1, min=1 max=1
class T_FFT<256>: total=1, N=1, <>=1, min=1 max=1
class T_FFT<512>: total=1, N=1, <>=1, min=1 max=1
class T_FFT<1024>: total=1, N=1, <>=1, min=1 max=1
class T_FFT<2048>: total=1, N=1, <>=1, min=1 max=1
class T_FFT<4096>: total=1, N=1, <>=1, min=1 max=1
class T_FFT<8192>: total=1, N=1, <>=1, min=1 max=1
class T_FFT<16384>: total=1, N=1, <>=1, min=1 max=1
class T_FFT<32768>: total=0, N=0, <>=0, min=0 max=0
class T_FFT<65536>: total=0, N=0, <>=0, min=0 max=0
class T_FFT<131072>: total=0, N=0, <>=0, min=0 max=0
class T_FFT<0>: total=0, N=0, <>=0, min=0 max=0
class T_FFT_strip<8>: total=0, N=0, <>=0, min=0 max=0
class T_FFT_strip<16>: total=0, N=0, <>=0, min=0 max=0
class T_FFT_strip<32>: total=0, N=0, <>=0, min=0 max=0
class T_FFT_strip<64>: total=0, N=0, <>=0, min=0 max=0
class T_FFT_strip<128>: total=0, N=0, <>=0, min=0 max=0
class T_FFT_strip<256>: total=0, N=0, <>=0, min=0 max=0
class T_FFT_strip<512>: total=0, N=0, <>=0, min=0 max=0
class T_FFT_strip<1024>: total=0, N=0, <>=0, min=0 max=0
class T_FFT_strip<2048>: total=0, N=0, <>=0, min=0 max=0
class T_FFT_strip<4096>: total=0, N=0, <>=0, min=0 max=0
class T_FFT_strip<8192>: total=0, N=0, <>=0, min=0 max=0
class T_FFT_strip<16384>: total=0, N=0, <>=0, min=0 max=0
class T_FFT_strip<32768>: total=1, N=1, <>=1, min=1 max=1
class T_FFT_strip<65536>: total=3, N=3, <>=1, min=1 max=1
class T_FFT_strip<131072>: total=11, N=11, <>=1, min=1 max=1
class T_FFT_strip<0>: total=0, N=0, <>=0, min=0 max=0
13:40:36 (4392): called boinc_finish

</stderr_txt>
]]>

WU ID
844500581, has found 30 spikes, 1 by GPU (ATI 5870) and 1 CPU,
[b]
Intel(R) Core(TM) i7 CPU 860 @ 2.80GHz [x86 Family 6 Model 30 Stepping 5], running DARWIN (Apple)
<stderr_txt>
setiathome_enhanced 6.03a XCode GCC 4.0.1 (Apple Inc. build 5494) i386

libboinc: 6.12.35
libboinc: 6.12.35

Work Unit Info:
...............
WU true angle range is : 0.778301
Optimal function choices:
-----------------------------------------------------
name
-----------------------------------------------------
v_BaseLineSmooth (no other)
v_vGetPowerSpectrumUnrolled 0.00011 0.00000
sse3_ChirpData_ak 0.01190 0.00000
v_vTranspose4x16ntw 0.00375 0.00000
BH SSE folding 0.00097 0.00000
SETI@Home Informational message -9 result_overflow
NOTE: The number of results detected exceeds the storage space allocated.

Flopcounter: 16938960949.655954

Spike count: 30
Pulse count: 0
Triplet count: 0
Gaussian count: 0
22:39:59 (82986): called boinc_finish

</stderr_txt>
]]>

WU ID
844332516, -9 error, 1 CUDA 2.3 and 1 ATI rev.177, the ATI rev.177 app.
hasn't made mistakes/errors on MB work.
However, most CUDA/OpenCL tasks, are checked against a CPU host, this time
24 spikes and 6 pulses were present in both WUs.

Another issue, quite some difference, 5.5GFLOPS on ATI and ~16GFLOPS on the CPU,
maybe something todo, ATI app. runs 2 instances_per_device, CPU 1 ?

---

Please go download, install MSI Afterburner 220 Beta 8, once a profile is saved and set, You can forget It, As a saved profile will stay in effect until overwritten and the MSI Afterburner is better than the Gigabyte program, Here's a description of MSI Afterburner. I use It 24/7, I've seen no problems, It can also adjust and set the gpu fan speed as well as down clock Your card.

---

The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's

MSI Afterburner and EVGA Precision are both written by the same author.
And he is the Russian programmer known as Unwinder
You can trust him.

He develops RivaTuner for more than 10 years
(and the above 2 programs are based on the RivaTuner's core, only the GUI/interface is different)

Interview: Alex 'Unwinder' / RivaTuner
http://www.techpowerup.com/reviews/Interviews/Alex_Unwinder/

"Relations with hardware companies
Despite modifying ATI and NVIDIA software and making their job more difficult, Alex says he has a good relationship with both vendors and their product partners."


If you don't believe me that MSI Afterburner and EVGA Precision are almost the same and based on RivaTuner
look at this:

---

 


- ALF - "Find out what you don't do well ..... then don't do it!" :)
 

No I didn't use MSI Afterburner, I used what came with the card.

Are we saying that GPU-Z may be telling lies?

I'll try it in a few minutes.... sheeesh.

No, It's just that MSI Afterburner is able to do much more then just read the cards...

Yeah. I was using GPU-Z to verify what I had done with Turbo Boost which also let's you reduce the clock speed and adjust the fan speed.

That may be, But the MSI version is better as It's more filled out and is newer as it's still being updated, Me I create a shortcut and copy that to the startup folder and I set the shortcut to start minimized and the MSI program starts when My PC does and at the settings that I set It at, no more, no less.

---

The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's

No I didn't use MSI Afterburner, I used what came with the card.

Are we saying that GPU-Z may be telling lies?

I'll try it in a few minutes.... sheeesh.

No, It's just that MSI Afterburner is able to do much more then just read the cards...

Yeah. I was using GPU-Z to verify what I had done with Turbo Boost which also let's you reduce the clock speed and adjust the fan speed.

Question is do You have to set turbo boost every time You start or restart Yer PC? If so I'd change to Afterburner instead, much less of a hassle and the program is Free to download and use. And I've never heard of TB, How about a link as I can't find It?

---

The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's

@ Bilbg, MSI-Afterburner, has 3 month's older dll's, executables, program looks
completely similar to EVGA-Precision. Probably UPdated and upgraded. And sound
added.
(I use Total Commander also, free version, looks alot like "the old Norton Commander")
It's widely used and works very good.
NVidia ships/delivers SmartDoctor with their GTX480, 580's which works OK for these
monsters. Nothing more is needed.

@ Crun-Chi, I find the factory speed, 1.8GHz., high compaired to other FERMIs, when I raised clock-speed on a GPUgrid WU, it errored out, inmediatly above 1.75GHz.on a 480.

GTS250, doesn't come which such a DVD, full of toys ;-)
So, I install EVGA-Precision on this VISTA 32bit rig, along with the right drivers.
Thanks.

---

Please go download, install MSI Afterburner 220 Beta 8, once a profile is saved and set, You can forget It, As a saved profile will stay in effect until overwritten and the MSI Afterburner is better than the Gigabyte program, Here's a description of MSI Afterburner. I use It 24/7, I've seen no problems, It can also adjust and set the gpu fan speed as well as down clock Your card.

[ img ]http://www.guru3d.com/imageview.php?image=25573[ /img ]

Another tool I like and have been using recently is NVIDIA Inspector

---

SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[

Please go download, install MSI Afterburner 220 Beta 8, once a profile is saved and set, You can forget It, As a saved profile will stay in effect until overwritten and the MSI Afterburner is better than the Gigabyte program, Here's a description of MSI Afterburner. I use It 24/7, I've seen no problems, It can also adjust and set the gpu fan speed as well as down clock Your card.

[ img ]http://www.guru3d.com/imageview.php?image=25573[ /img ]

Another tool I like and have been using recently is NVIDIA Inspector

I like Nvidia Inpector also. I do my GPU overclocking with it, and use MSI Afterburner to do my monitoring. Both are excellent tools.

Steve

---

Warning, addicted to SETI crunching!
Crunching as a member of GPU Users Group.
GPUUG Website

@ Crun-Chi, I find the factory speed, 1.8GHz., high compaired to other FERMIs, when I raised clock-speed on a GPUgrid WU, it errored out, inmediatly above 1.75GHz.on a 480.

I think that 560TI has good OC potential, but when you ( or factory) raise frequency , you must also raise voltage. Step by step is best approach to problem. Some GPU don't need raise of voltage, some GPU need less or more, but nothing until experiment cannot tell you what is good choice.
My 560 TI works well on games but on SETI give me one error in 20 WU. Then I start OCCT test, find errors after few minutes, makes few versions of BIOS with Nibitor, and flash it. When I find optimal values I never get any WU error again.

---

I am cruncher :)
I LOVE SETI BOINC :)

@ Bilbg, MSI-Afterburner, has 3 month's older dll's, executables, program looks
completely similar to EVGA-Precision. Probably UPdated and upgraded. And sound
added.

Don't mind the dates at all - I don't have the latest versions
(I have integrated GPU NVIDIA GeForce 6150SE nForce 430 so installed MSI Afterburner and EVGA Precision only of curiosity)

---

 


- ALF - "Find out what you don't do well ..... then don't do it!" :)