Posts by Helsionium

21) Message boards : Number crunching : CUDA problems (Lunatics) / zero status but no finished file (Message 1313292)
Posted 9 Dec 2012 by Profile Helsionium
Post:
Thank you all for your opinions.

I used an old version of MSI Afterburner that couldn't make use of all the functions. With the new one I can indeed adjust the voltage (but can still not turn the fan speed of my 660 above 74% which is strange). But if, as you say, this is already a rather high temperature I'd rather not fiddle with the voltage control unless necessary. However, according to several internet sources, the temperature is pretty much the expected temperature of my card - the Zotac GTX 660 Amp! Edition - under maximum load, maybe a few degrees higher, but that's expected too (multi-GPU system, very demanding CUDA load instead of the game-based benchmarks)

I never enabled SLI to begin with, and also switched to "single display performance mode", but that too didn't help. I've set the process priority to high via the mbcuda.cfg file.

I'm out of time now, but tomorrow I will try switching the cards around and find out if hyperthreading affects the problem or not. The good thing is that the error is relatively benign and that all of the affected tasks I followed eventually successfully validated.
22) Message boards : Number crunching : CUDA problems (Lunatics) / zero status but no finished file (Message 1313134)
Posted 9 Dec 2012 by Profile Helsionium
Post:
The problem persists even after I did this:

- reduce CPU speed from 4.3 GHz to 4.0 GHz
- reduce RAM frequency from 1600 to 1333
- disabled XMP
- installed x41zb
- tested GPU with OCCT, no errors, no artifacts

I can't change the GPU core voltage on the 660, the software doesn't allow that. I didn't find any options to adjust Vdd in the BIOS. The 660 isn't overclocked (past its factory-overclocking) and the software doesn't allow me to go lower than the default clock rates, so I can't do this. The error is probably not related to temperature, though. There was one instance of the error shortly after reboot, when temperatures were still much lower.

Just shortly before the installation of the new 660, I disabled hyperthreading because it was recommended by some in this forum. Could that be related?
23) Message boards : Number crunching : CUDA problems (Lunatics) / zero status but no finished file (Message 1313022)
Posted 9 Dec 2012 by Profile Helsionium
Post:
Alright, I checked:

- Hard drive integrity: OK (and no processes interfering with BOINC or SAH)
- DPC latency: OK, consistently below 500 microseconds, sometimes slightly above 500 if BOINC is active.
- Command rate: OK, set to T2.

About the temperature: is 70°C already too hot? Both GPUs and the CPU typically run at 68-74 degrees under full load. According to the information I had, this is well within their specifications.

About RAM Vdd: I can't really check this now, as I want to avoid restarting the machine unless absolutely necessary. Can this be a problem with RAM? Why would that affect only the tasks running on the 660 GPU and not the older, but almost equally fast 560 Ti?

The power supply should be OK as well. The system ran just fine with two overclocked 560 Tis which used more power.
24) Message boards : Number crunching : CUDA problems (Lunatics) / zero status but no finished file (Message 1312721)
Posted 8 Dec 2012 by Profile Helsionium
Post:
Hello there,

ever since I replaced one of my two 560Ti GPUs with a GTX 660, a relatively large amount of tasks (I guess 20-30%) restart at least once (exited with zero status but no 'finished' file) after encountering this error:

Error on call (cudaMemcpy(PowerSpectrumSumMax, dev_PowerSpectrumSumMax, (cudaAcc_NumDataPoints / fftlen) * sizeof(*dev_PowerSpectrumSumMax), cudaMemcpyDeviceToHost)), file c:/[Projects]/X_CudaMB/client/cuda/cudaAcc_summax.cu, line 239: unknown error

Here is a stderr output from one task that restarted five times:
http://files.helsionium.eu/stderr-2739587852.txt

It seems this problem only affects tasks that run on the new 660. What could be causing it? The same system ran fine with the two 560 Tis before. This problem isn't particularly grave - all of the affected workunits eventually complete and validate but I guess it's still having an impact on overall crunching performance, aside from being a major nuisance.

I'm using BOINC 7.0.28 on Windows 7 64-bit, the Lunatics x41g optimized app and the beta driver 310.70 (same problem with the latest WHQL driver, though).

Thank you in advance,
Helsionium
25) Message boards : Number crunching : So what's going on here? (Message 1308801)
Posted 22 Nov 2012 by Profile Helsionium
Post:
I had 5000+ ghost units as well. Since I decided that I'd rather crunch for other projects until the current issues are resolved, I purposely abandoned them rather then have them timeout over the next few weeks. This was at a time when I couldn't reach the scheduler or download server for days. Now that these issues are (partially?) resolved you might just as well sit back and relax, the ghosts will be resent eventually.
26) Message boards : Number crunching : Best RAC for your buck GPU? (Message 1307349)
Posted 18 Nov 2012 by Profile Helsionium
Post:
I've asked the same question some time ago and I believe the consensus was:
The 500 series are currently the best for SETI@home and they are cheaper than the 600 series. However, they use significantly more power, something to keep in mind.

However, I'm waiting for the next iteration of the optimized applications, because they might render the 600 series more useful.
27) Message boards : Number crunching : Noticed something strange... (Message 1298606)
Posted 25 Oct 2012 by Profile Helsionium
Post:
Thank you, I didn't know that.
Mods, please close this thread.
28) Message boards : Number crunching : Noticed something strange... (Message 1298595)
Posted 25 Oct 2012 by Profile Helsionium
Post:
Hello,
I just noticed something strange - nothing grave, but possibly a bug:

On the application details page for my main host
http://setiathome.berkeley.edu/host_app_versions.php?hostid=6141926
in the section Astropulse v505 (anonymous platform, CPU):

Number of tasks completed: 231
Consecutive valid tasks: 233

How can there be more valid tasks than completed tasks?

Kind regards,
Helsionium
29) Message boards : Number crunching : Can't download the lunatics installer 0.40 (Message 1291549)
Posted 5 Oct 2012 by Profile Helsionium
Post:
The important line is:

Installers have been temporarily removed pending resolution of some concerns related to licensing (20 Sept. 2012).


If I remember correctly, there were some licensing issues with Intel and Lunatics is working on resolving these issues. There's nothing that can be done about that, please just wait it out.
30) Message boards : Number crunching : MSI 760GM-E51 (FX) Mainboard no good for BOINC; can/will fry at 100% use (admitted by manufacturer) (Message 1291080)
Posted 4 Oct 2012 by Profile Helsionium
Post:
I had a MSI P67A-GD65 mainboard, which does not carry this warning, which failed after about half a year of heavy usage (SETI@home optimized with 8 CPU threads and 2 GPUs). It may be a one-time thing, though. I never had any stability issues with this board befor or after the replacement.

While I was still running a dual-GPU setup, I had three (!) MSI N560GTX-Ti Twin Frozr II/OC that failed (the fans broke all three times) in nine months.
So far, it has only been an annoyance because all defects were covered by warranty, but I don't think I'll ever again buy MSI products. I had very good experiences with their products before, don't know why quality seems to have gone down so much.
31) Message boards : Number crunching : Please help with cruncher upgrade (Message 1285566)
Posted 19 Sep 2012 by Profile Helsionium
Post:
OK thank you very much, I'm considering buying a GTX 660 Ti now.
32) Message boards : Number crunching : Please help with cruncher upgrade (Message 1285500)
Posted 19 Sep 2012 by Profile Helsionium
Post:
Oh I didn't know that the Tesla series is already that old. I also don't understand why a card that old still sells for around € 500 - I can get two GTX 660s for that price.
What is the difference between the Ti and the non-Ti versions? Sorry for asking, but this is not immediately obvious when comparing the numbers.
33) Message boards : Number crunching : Please help with cruncher upgrade (Message 1285410)
Posted 19 Sep 2012 by Profile Helsionium
Post:
Hello,

this computer was a dual-GPU setup until one of them broke a few months ago.
Now I'm going to replace the defective unit with something else, and I ask for your assistance in finding something suitable.

The main question is: Does anyone know how well a Tesla C870 performs vs another MSI N560GTX-Ti Twin Frozr II/OC? I tried comparing the numbers, but then I thought that this might not be a good idea, because of the different architectures.

If the C870 is not to be recommended, what is currently the GPU with the best price/performance ratio (less than $500), assuming I'm only using this replacement GPU for SETI@home?

Thanks in advance,
Helsionium
34) Message boards : Cafe SETI : Speed Limit: 85 (Message 1243364)
Posted 8 Jun 2012 by Profile Helsionium
Post:
85 mph is my typical driving speed on German and Austrian autobahns (Austria has a speed limit of 130 kph/81 mph), I consider that speed very safe. The vast majority of fatal autobahn accidents here are due to bad weather/visibility or driving under influence or at motorway junctions. The general speed limit is not a factor in any of those.

I believe the "German system" is perfect for long-distance motorways if they are appropriately built. I personally don't feel very comfortable driving faster than 160 kph/100 mph, but if others do so safely (keeping distance!) I have no problem with it.

In my opinion, the most dangerous situations arise from (relatively) slow drivers who drive in the left/middle lanes all the time, frequently enticing faster drivers to drive "more aggressively". While I condemn any aggresive style of driving, I must say such situations can be very frustrating or even aggravating. Once, while driving an emergency vehicle, it took two minutes for the driver in front of me to notice me, even though I had all lights, sirens and horns on. I guess that without those, he would have probably never noticed me at all.

So it's not just the faster drivers who must drive more responsibly, but everyone else, too. If this does not work well, the lower speed limits might remain necesseary.
35) Message boards : Cafe SETI : Earthquake hits N Italy (Message 1238767)
Posted 29 May 2012 by Profile Helsionium
Post:
And once again, it could be felt all the way up here, as powerful or even more powerful than any local earthquakes we've ever had in the past twenty years - and we have many. There will probably me more aftershocks. It is strongly advised to stay clear of any damaged buildings if possible.
EDIT: Seems that this earthquake was even more damaging. Ten fatalities reported so far.
36) Message boards : Cafe SETI : Earthquake hits N Italy (Message 1235005)
Posted 22 May 2012 by Profile Helsionium
Post:
The problem is that the predictions are either to vague (March to September? Can't evacuate an entire region for half a year) or on very short notice (animal behaviour). Such predictions are of very little use in evacuating or even just warning the population.
37) Message boards : Cafe SETI : Earthquake hits N Italy (Message 1234747)
Posted 21 May 2012 by Profile Helsionium
Post:
Today, a seismologist from our national earthquake service said that aftershocks may continue for up to two months, with about 200 of them powerful enough to be felt throughout the Emilia-Romagna region. He said aftershocks of magnitude 5 or greater are unlikely, though. The main earthquake was reportedly felt in some of the higher buildings of Vienna, 530 kilometers northeast of the epicenter.

A strange coincidence: A joint civil defense exercise (forces from Austria, Germany, Switzerland) is currently ongoing in my region. It was planned for over two years; the scenario is a 6.3 magnitude earthquake shortly followed by a 6.2 in a neighbouring region on May 21st (today) in the early morning. The scenario is remarkably similar to the actual earthquake that happened yesterday.

Now I just hope the weather in the Emilia-Romagna region gets better soon. Disaster relief is always difficult, but bad weather can severely impact these efforts.
38) Message boards : Cafe SETI : Earthquake hits N Italy (Message 1234131)
Posted 20 May 2012 by Profile Helsionium
Post:
The earthquake and the first aftershock both woke me up at Innsbruck, Austria, 270 kilometers north of the epicenter. The aftershock of 15:18 could also be felt here. Our national earthquake service said aftershocks may continue for weeks, with some of them potentially powerful enough to cause further damage.
My thoughts are with all the Italian people.
39) Message boards : SETI@home Staff Blog : Can Alec Baldwin Crash a Plane With a Cell Phone? (Message 1192962)
Posted 9 Feb 2012 by Profile Helsionium
Post:
While I completely agree with any measures in place to minimizes risks, I do not completely understand the reasoning behind the idea that cell phones might be dangerous to aircraft operation.

Cell phones are relatively low-powered radio transmitters, even at their maximum output power - I think it's 2 Watts for GSM handsets, and much less for more recent technologies. One would think that aircraft capable of flying through thunderclouds are supposed to be able to deal with electromagnetic fields of relatively minor magnitudes.

Moreover, it is very unlikely that a cell phone will be able to receive a signal except close to takeoff/landing. According to my knowledge, cell phones do not transmit at all while out of coverage.

The famous 'GSM buzz' that was already mentioned is also unlikely to be a problem, because it has a very short range and newer technologies don't cause similar interference at all.

In 2007, one pilot recounted an instance when the navigational equipment on his Boeing 737 had failed after takeoff. A flight attendant told a passenger to turn off a hand-held GPS device and the problem on the flight deck went away.


This simply is not possible. Reminds me of an instance where a patient of mine told me to turn off the navigation system of my ambulance car because he feared negative effects by electromagnetic radiation. However, GPS devices don't transmit. They only receive (very weak) signals from the GPS satellites and do not cause any radiation at all.
40) Message boards : Number crunching : # of tasks per GPU list (Message 1171912)
Posted 18 Nov 2011 by Profile Helsionium
Post:
I conducted a test on my machine with 2 OC'ed GTX 560 Ti cards (900 MHz Core / 1070 MHz Memory, 1 GB VRAM, Lunatics x38g).

The result is that there is a significant advantage of running 2 or 3 WUs at once over running only one at a time. However, the test does not clearly show any advantage or disadvantage of running more than two WUs at a time. The two GPUs have different memory usages because the first one displays to two 1080p screens. At least that's what I think.

1 WU at a time:
Memory 473 MB / 248 MB
GPU Load 65-80%

8 WUs completed. When the timer stopped, the active WUs were at 56.9 / 57.7 percent.
Thus the score is 9.15 WUs / 10 minutes.

2 WUs at once:
Memory 741 MB / 496 MB
GPU Load 95-99%

12 WUs completed. When the timer stopped, the active WUs were at 50.2 / 46.2 / 36.8 / 29.9 percent.
Thus the score is 13.63 WUs / 10 minutes.

3 WUs at once:
Memory 937 MB / 743 MB
GPU Load 98-99%

12 WUs completed. When the timer stopped, the active WUs were at 39.3 / 30.3 / 28.9 / 27.7 / 25.0 / 17.7 percent.
Thus the score is 13.69 WUs / 10 minutes.


Previous 20 · Next 20


 
©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.