Posts by ohiomike

1) Message boards : Number crunching : BoincView web site gone? (Message 848888)
Posted 3 Jan 2009 by Profile ohiomike
Post:
Anyone tried BV with BOINC 6.4.5? Are they compatible?

V1.5.8 Beta works with Win XP X64 & Boinc 6.4.5/6.5 with minor quirks.
"Accelerators" does not show GPUs correctly, etc.

2) Questions and Answers : GPU applications : JACT (Jet Another Cuda Thread): GPU undeperforms when all CPU cores are processing WUs (Message 847387)
Posted 31 Dec 2008 by Profile ohiomike
Post:
Configuration:

- AMD Phenom (4 cores)
- 8600GT on a PCIE slot
- 8300GS on board
- Windows XP SP2
- Boinc 6.4.5, Nvidia drivers 178.24

Now that CUDA seem to work as expected on my system, I have a couple of reports about scheduling and one important question. Here is what I noticed:

- Sometimes only one CUDA thread is started, other times two of them (as I have got two CUDA-capable devices). If, when both GPUs are crunching, I manually suspend one CUDA WU, no new CUDA WU is started - i.e. only one CUDA thread is left working.

- Sometimes three CPU-based threads are started, sometimes four of them. Note that I did not rise the CPU count artificially by edithing the XML file.

- And finally the most important issue: when four CPU-based crunching threads are running, my CUDA devices seem to underperform: nVidia monitor says that GPU load is about 25%, and GPU temperature is considerably lower as well. If I manually stop one of the CPU tasks, GPU usage goes to 100% and temperature rises about 10 degrees C.

Can anybody confirm the latter fact?
Maybe it has all already been reported and/or addressed, but if this is not the case, owner of four cores might as well decide to reserve one of them to GPU crunchers until the issue is resolved: what they lose in CPU processing power should be more than balanced by GPU power increase.


Just to add a note to this, when I had my Q6600 setup with 2 8800 GTS cards as a test, it looked like the Cuda tasks my have gotten mixed up about which GPU to use. I could not confirm this without more debug info from the tasks however. I like how Folding@Home does it, you lock a GPU to a task slot.
3) Message boards : Number crunching : Win2K, AMD CPU, 8800 GT, and Cuda?? (Message 847217)
Posted 31 Dec 2008 by Profile ohiomike
Post:
Same result with DirectX 9c:

<core_client_version>6.5.0</core_client_version>
<![CDATA[
<message>
There are no child processes to wait for. (0x80) - exit code 128 (0x80)
</message>
<stderr_txt>
setiathome_CUDA: Found 1 CUDA device(s):
Device 1 : GeForce 8500 GT
totalGlobalMem = 536543232
sharedMemPerBlock = 16384
regsPerBlock = 8192
warpSize = 32
memPitch = 262144
maxThreadsPerBlock = 512
clockRate = 918000
totalConstMem = 65536
major = 1
minor = 1
textureAlignment = 256
deviceOverlap = 1
multiProcessorCount = 2
setiathome_CUDA: CUDA Device 1 specified, checking...
Device 1: GeForce 8500 GT is okay
SETI@home using CUDA accelerated device GeForce 8500 GT
Rise priority modification by Raistmer based on rev380 of SETI@home sources
Priority of worker thread rised successfully
setiathome_enhanced 6.02 Visual Studio/Microsoft C++
libboinc: 6.3.22

Work Unit Info:
...............
WU true angle range is : 0.556810
Optimal function choices:
-----------------------------------------------------
name
-----------------------------------------------------
v_BaseLineSmooth (no other)
v_GetPowerSpectrum 0.00048 0.00000
v_ChirpData 0.01866 0.00000
v_Transpose4 0.01402 0.00000
FPU opt folding 0.00409 0.00000
Cuda error 'cudaAcc_GetPowerSpectrum_kernel' in file 'd:/BTR/seticuda/Berkeley_rep/client/cuda/cudaAcc_PowerSpectrum.cu' in line 56 : unknown error.

</stderr_txt>
]]>

Validate state Invalid
4) Message boards : Number crunching : Win2K, AMD CPU, 8800 GT, and Cuda?? (Message 847189)
Posted 31 Dec 2008 by Profile ohiomike
Post:
There are no child processes to wait for. (0x80) - exit code 128 (0x80)

Update your DirectX version to be at least one of the newer DirectX 9.0c versions. The error message you get is usually due to you trying to run with an outdated DirectX version. If you never updated Dx, Windows 2000 uses by default Dx 7, if I am not mistaken.

OK- I'll give that a try. Thanks.
5) Message boards : Number crunching : Win2K, AMD CPU, 8800 GT, and Cuda?? (Message 847175)
Posted 31 Dec 2008 by Profile ohiomike
Post:
Has anyone gotten any successful results with Windows 2000 and a 8500 GT?
I have tried the stock app, all 3 versions of the opt app, and the standard and Beta NVidia drivers.
I have also tried Boinc 6.45 and Boinc 6.5.
Note this GPU did return several valid results when installed in a different system
(System was Intel Q6600 under Windows XP X64).

Current System= Windows 2000, AMD 5600 + X2, 1 GB Ram, ASUS 8500 GT

With Boinc 6.5, latest opt app, NVidia 180.84:

<core_client_version>6.5.0</core_client_version>
<![CDATA[
<message>
There are no child processes to wait for. (0x80) - exit code 128 (0x80)
</message>
<stderr_txt>
setiathome_CUDA: Found 1 CUDA device(s):
Device 1 : GeForce 8500 GT
totalGlobalMem = 536543232
sharedMemPerBlock = 16384
regsPerBlock = 8192
warpSize = 32
memPitch = 262144
maxThreadsPerBlock = 512
clockRate = 918000
totalConstMem = 65536
major = 1
minor = 1
textureAlignment = 256
deviceOverlap = 1
multiProcessorCount = 2
setiathome_CUDA: CUDA Device 1 specified, checking...
Device 1: GeForce 8500 GT is okay
SETI@home using CUDA accelerated device GeForce 8500 GT
Rise priority modification by Raistmer based on rev380 of SETI@home sources
Priority of worker thread rised successfully
setiathome_enhanced 6.02 Visual Studio/Microsoft C++
libboinc: 6.3.22

Work Unit Info:
...............
WU true angle range is : 0.433189
Optimal function choices:
-----------------------------------------------------
name
-----------------------------------------------------
v_BaseLineSmooth (no other)
v_GetPowerSpectrum 0.00048 0.00000
v_ChirpData 0.01880 0.00000
v_Transpose4 0.01406 0.00000
FPU opt folding 0.00487 0.00000
Cuda error 'cudaAcc_GetPowerSpectrum_kernel' in file 'd:/BTR/seticuda/Berkeley_rep/client/cuda/cudaAcc_PowerSpectrum.cu' in line 56 : unknown error.

</stderr_txt>
]]>

Validate state Invalid


- OR With Boinc 6.5, latest opt app, NVidia 178.48:

<core_client_version>6.5.0</core_client_version>
<![CDATA[
<stderr_txt>
setiathome_CUDA: Found 1 CUDA device(s):
Device 1 : GeForce 8500 GT
totalGlobalMem = 536543232
sharedMemPerBlock = 16384
regsPerBlock = 8192
warpSize = 32
memPitch = 262144
maxThreadsPerBlock = 512
clockRate = 918000
totalConstMem = 65536
major = 1
minor = 1
textureAlignment = 256
deviceOverlap = 1
multiProcessorCount = 2
setiathome_CUDA: CUDA Device 1 specified, checking...
Device 1: GeForce 8500 GT is okay
SETI@home using CUDA accelerated device GeForce 8500 GT
Rise priority modification by Raistmer based on rev380 of SETI@home sources
Priority of worker thread rised successfully
setiathome_enhanced 6.02 Visual Studio/Microsoft C++
libboinc: 6.3.22

Work Unit Info:
...............
WU true angle range is : 0.432071
Optimal function choices:
-----------------------------------------------------
name
-----------------------------------------------------
v_BaseLineSmooth (no other)
v_GetPowerSpectrum 0.00048 0.00000
v_ChirpData 0.01867 0.00000
v_Transpose4 0.01391 0.00000
FPU opt folding 0.00490 0.00000
Cuda error 'cudaMemcpyToSymbol(cudaAcc_GaussFit_settings, (void*) &settings, sizeof(settings))' in file 'd:/BTR/seticuda/Berkeley_rep/client/cuda/cudaAcc_gaussfit.cu' in line 454 : invalid device symbol.
Cuda error 'cudaMemcpyToSymbol(cudaAcc_PulseFind_settings, (void*) &PulseFind_settings, sizeof(PulseFind_settings))' in file 'd:/BTR/seticuda/Berkeley_rep/client/cuda/cudaAcc_pulsefind.cu' in line 669 : invalid device symbol.
Cuda error 'cudaAcc_CalcChirpData_kernel2' in file 'd:/BTR/seticuda/Berkeley_rep/client/cuda/cudaAcc_CalcChirpData.cu' in line 106 : invalid device function .
Cuda error 'cudaAcc_GetPowerSpectrum_kernel' in file 'd:/BTR/seticuda/Berkeley_rep/client/cuda/cudaAcc_PowerSpectrum.cu' in line 56 : invalid device function .
Cuda error 'cudaAcc_summax_kernel<8>' in file 'd:/BTR/seticuda/Berkeley_rep/client/cuda/cudaAcc_summax.cu' in line 156 : invalid device function .
SETI@Home Informational message -9 result_overflow
NOTE: The number of results detected exceeds the storage space allocated.

Flopcounter: 331089936.000000

Spike count: 30
Pulse count: 0
Triplet count: 0
Gaussian count: 0
called boinc_finish

</stderr_txt>
6) Message boards : Number crunching : if you think cuda is fast (Message 846234)
Posted 29 Dec 2008 by Profile ohiomike
Post:
as far as to which is faster, they will switch back and forth as the new versions are released. ATI will release a card that is faster than the Nvidia latest. Then Nvidia will release another in a couple months that is faster than that. Same with Intel and AMD...

And there is going to be the ATI is faster on project A while NVidia is faster on project B to come...

Kinda moot really right now as ATI is not capable of doing BOINC ... yet ...

The thing that may help ATI in the long run is the fact that their GPUs will do double precision FP (NVidia is single precision only).
7) Message boards : Number crunching : Someone please explain this (Message 845653)
Posted 27 Dec 2008 by Profile ohiomike
Post:
A set and forget user might have a passively cooled CUDA card (the only one I have is) and not notice until it's too late that the CUDA app has over-heated it...

-Dave

If the system is cooled correctly, a user shouldn't have to notice the card is overheating. All GPUs come with a VBIOS that checks the GPU's temperature. The drivers in combination with the VBIOS know what is the best temperature for this sort of GPU to run on, knows its minimum and maximum. If the maximum is reached, the GPU will change the fan speed (on active cooling) or clock down its clocks/voltage (on passive cooling) fully automatically.

Not everyone has adequate cooling and venting in the rest of the computer. That's something you really need when using passively cooled cards.

Now when comparing Folding@Home with Seti's use of the GPU, FAH does calculations in bursts. It'll use the GPU for a couple of minutes, then back off and use the CPU. SAH will use the GPU continuously at full load. The work does come in bursts, but the time between the bursts is very short, not enough to cool the GPU down.


The thing that is easy to overlook here is where the heat goes. I have 2 8800 GTS cards running Folding@Home and it keeps one card @ 81C and the other (with the partially blocked air intake) @ 89C. The problem is all the heat is not being blown out the back, everything inside the case heated up until I cooked 2 sticks of ram. (This unit is now running again with new ram and the case open and on it's side so the heat goes up into the air instead of up into the ram).

Moral of the story- Keep an eye on the case temp along with the GPU temps.

That being said, I agree that the GPU option should default to "OFF". The system needs to be checked and monitored after turning it on.


8) Message boards : Number crunching : Modified SETI MB CUDA + opt AP package for full GPU utilization (Message 845374)
Posted 26 Dec 2008 by Profile ohiomike
Post:
I have had two tasks with too many results (waiting on wingmen to verify or not), and this one:

setiathome_CUDA: Found 1 CUDA device(s):
Device 1 : GeForce 8500 GT
totalGlobalMem = 536543232
sharedMemPerBlock = 16384
regsPerBlock = 8192
warpSize = 32
memPitch = 262144
maxThreadsPerBlock = 512
clockRate = 918000
totalConstMem = 65536
major = 1
minor = 1
textureAlignment = 256
deviceOverlap = 1
multiProcessorCount = 2
setiathome_CUDA: CUDA Device 1 specified, checking...
Device 1: GeForce 8500 GT is okay
SETI@home using CUDA accelerated device GeForce 8500 GT
Rise priority modification by Raistmer based on rev380 of SETI@home sources
Priority of worker thread rised successfully
Cuda error 'cufftPlan1d(&fft_analysis_plans[FftNum], FftLen, CUFFT_C2C, NumDataPoints / FftLen)' in file 'd:/BTR/seticuda/Berkeley_rep/client/cuda/cudaAcc_fft.cu' in line 49 : out of memory.
setiathome_CUDA: CUDA runtime ERROR in plan FFT. Falling back to HOST CPU processing...
setiathome_enhanced 6.02 Visual Studio/Microsoft C++
libboinc: 6.3.22

Work Unit Info:
...............
WU true angle range is : 2.715856
Optimal function choices:
-----------------------------------------------------
name
-----------------------------------------------------
v_BaseLineSmooth (no other)
v_GetPowerSpectrum 0.00021 0.00000
v_ChirpData 0.01462 0.00000
v_Transpose4 0.00563 0.00000
FPU opt folding 0.00172 0.00000

Flopcounter: 5215429468138.598600

Spike count: 3
Pulse count: 0
Triplet count: 1
Gaussian count: 0
called boinc_finish

9) Message boards : Number crunching : Some testing tool for GPU needed (Message 844821)
Posted 25 Dec 2008 by Profile ohiomike
Post:
Do you know any testing utility that checks validness of GPU computations?
Something like CPU burn

I haven't found anything. I have noticed that my system ram is right over a 70 drg C GPU however. It cooked the ram. More fans, and under-clocking the GPU right now. Makes me wonder if the errors we are seeing are ram failures caused by hot ram. If you find a way to load up a GPU and monitor it, I would keep an eye on the ram temp if you can.
10) Message boards : Number crunching : Minor Cuda app Error (Message 844819)
Posted 25 Dec 2008 by Profile ohiomike
Post:
In looking at some Cuda WUs I've run, I noticed that the message always says that it is using GPU #2 (I have 2 8800 GTS cards in the system). The message never shows #1 as being used (both are loaded and running).
Msg=
cudaAcc_initializeDevice: Found 2 CUDA device(s):
Device 1 : GeForce 8800 GTS
Device 2 : GeForce 8800 GTS
cudaAcc_initializeDevice is determiming what CUDA device to use...
user specified SETI to use CUDA device 2: GeForce 8800 GTS


11) Message boards : Number crunching : Thinking of Platform change (Message 753836)
Posted 16 May 2008 by Profile ohiomike
Post:
Hans Dorn is running BOINC under Wine. Seems less dramatic.

Speaking of Wine, I've been waiting forever for this to work ---> Darwine

I think Apple is gonna beat 'em to the punch with their next version of OSX... ;)

(I did have minesweeper running at one point, but that was about it.)

If they were smart they would get into the PC market and put out an operating system for all and watch Microsoft sqirm...It can't be that hard at this point...

It's actually a little dangerous.

Apple is, and has been since the mid-80's, a software company who funds their software development through hardware sales.

When Apple started allowing authorized clones, other companies made the money and OS development basically died. Fortunately, they quit before it was too late.

If they released a PC compatible version of Darwin, they'd have to price it low enough to get a significant number of converts, and still hold the bottom line while their hardware sales drop dramatically.


I have asked the same question a hundred times and this was the first answer that actually seems to make sense. Too bad it would be nice to have out of the box competition with the support they would be able to gain...Hardware companies would be climbing all over themselves to make drivers and such...It hasn't hurt Microsoft any...

Not quite legal but: http://wiki.osx86project.org/wiki/index.php/Main_Page shows how to run OSX on almost anything.
12) Message boards : Number crunching : Set Your Credit Target on Nez (Message 723661)
Posted 9 Mar 2008 by Profile ohiomike
Post:
Target will be reached in 13214.7 days on May 11 2044

Based on his credits as of 5 minutes ago....351,474,066...and counting.....

Hmmm....I'd be 87 years old.........

You are one of the few people that stand a chance of being alive when they reach that point! :)
13) Message boards : Politics : Moderation Change suggestion(s) - CLOSED (Message 723660)
Posted 9 Mar 2008 by Profile ohiomike
Post:
Jeremy et al

Without all the boring detail, every moderator knows what has been done to a particular message and why it was removed. In some cases a message was put back because "as a group" it was felt that message was within the posting guidelines (mostly new moderators learning). In some cases when a user asked "SetiMod" why a message was removed an explaination of what was not with the guidelines helped a user understand, or in some cases what was questionable was understood. Yes, we understand that English is not everyones first language. We are all human.

To answer the question how does Admin know when a moderator is having problems. In most cases it is the moderators policing the group that handles that "everyone learns about moderation on the job." The current group of moderators are from diverse backgrounds, education and nationality. While they may not all agree on "everything" the moderators have worked together long enough that they can reach a consensus.

The moderators email addresses are in the "accept list" in Eric's Spam Filter. Should someone have a question anyone can pose an issue to Eric personally.

For the larger part, anything that a moderator can do another moderator can undo.

If all else fails, there are several moderators who have phone numbers that can contact Eric immediately.

For the last part, Eric stated that he would like to have more moderators from different parts of the world to insure that all time zones are covered. If anyone would like to "volunteer or recommend a person" You are welcome to email me, I will insure that is forwarded to Eric.
For those that may not recall my public email address it is al.setiboinc at gmail.com (Yes, replace "at" and the spaces with the @ symbol). Yes, there is an unpublished selection process that is involved.

C: How does admin know or not know when a moderator is abusing his/her powers?


There is at least one SETI Administrator that receives all Modlist emails (if not more than one). Assuming this(these) person(people) have the time to read their emails, they know about every single action and discussion taking place. They can see who is "abusing" and who is simply trying to perform their duty.

Ok, fair enough. But expanding on this a little, we all know Matt doesn't care about the forums, Eric doesn't have the time to read the forums, and really pushed to have more moderators here to handle the work to allow him and the rest of the crew to focus on the important items, such as the project itself.

So with that, How does admin know or not know when a mod is being abusive?

Very good answer, thank you.
14) Message boards : Politics : Moderation Change suggestion(s) - CLOSED (Message 723556)
Posted 9 Mar 2008 by Profile ohiomike
Post:
Looking at
http://setiathome.berkeley.edu/forum_thread.php?id=45813
I begin to question the moderation system. May I suggest:
1) That moderators be elected and also subject to a recall vote?
2) That the moderators and their last x number of actions be currently displayed somewhere so that moderation does not appear to be a "behind the scenes" click/conspiracy.
15) Message boards : Number crunching : Problems switching to 64 bit client. (Message 720619)
Posted 2 Mar 2008 by Profile ohiomike
Post:
Just to put in my 2 cents worth- I always install Boinc in it's own partition:
1)It writes data to it's program file area- very bad practice (Proper design is for separate program and data areas)! I avoid putting anything in the Program Files directory that uses that area for data. (If you upgrade to Vista it is not allowed- finally).
2) With Boinc in it's own partition, my backups on my C drive are much smaller (data is not changing constantly like Boinc apps do).
16) Message boards : Number crunching : six cores on die... (Message 719106)
Posted 27 Feb 2008 by Profile ohiomike
Post:
<snip>
IBM chose EBCDIC.

Electronics engineer for 35+ years.

Wash your mouth out with bits :)
17) Message boards : Number crunching : GPU crunching on G8x (Message 718924)
Posted 27 Feb 2008 by Profile ohiomike
Post:
Why not spend $650 for a Tesla C870 GPU Computing Processo (128 cores).
http://www.nvidia.com/object/tesla_computing_solutions.html
18) Message boards : Number crunching : And we think Berkeley has computer problems (Message 718923)
Posted 27 Feb 2008 by Profile ohiomike
Post:
The problem is- it is most likely a legacy system (Cobol or such) which would require a re-write from page 0. I have a friend that still does a lot of Cobol work on IBM 360's and TI990/12's for hospitals that still used them for records and billings.
19) Message boards : Number crunching : Has the O.S. used, MUCH inpact on calculating speed? (Message 713939)
Posted 16 Feb 2008 by Profile ohiomike
Post:
<snip>
And if you go for multi-booting, install Windows FIRST. Windows will wipe out anything other than Windows on your disks. The later Linux install will give you a boot option for whatever is already there.

Good luck,

Happy crunchin',
Martin


You can install in the reverse order if you have disk imaging SW.
1) Make a disk image of the Boot disk.
2) Install Windoze to a spare partion (must be active partition).
3) Restore MBR & track 0 from the image.
4) Add Windows boot to grub.conf (or menu.lst)
(I found this the hard way after Windows XP ate it's self and I had to re-install).
20) Message boards : Number crunching : Need advice for Antivirus - Antispyware for XP Pro. 64 bit. (Message 711630)
Posted 12 Feb 2008 by Profile ohiomike
Post:
Avast Free edition works nicely, has a very low CPU over-head (unlike McAfee and Norton).


Next 20


 
©2025 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.