Einstein hard on the CPU
19 Jan 2005 16:08:24 UTC
Topic 187262
(moderation:
Hi,
I've just starting crunching for Einstein@home today and noticed that it is very hard on the cpu - temperature rises approximately 2-3 degrees C higher than when crunching on any other Boinc project, even CPDN which is quite cpu intensive. Has anyone else noticed this? I didn't see anything on these message boards about it.
Unfortunately, now the cpu temp is close to the critical level where the pc gets unstable - I guess more cooling is required once again :-(
Einstein hard on the CPU
)
Einstein@Home is mainly CPU-bound, thus it's not unusual that it takes the cooling to the limit. We have had some requests for a CPU-throtteling feature, but we have things with much higher priority keeping us busy. You may take a look at the thread http://einsteinathome.org/node/187169 for alternatives.
BM
BM
Well, it's not a complaint,
)
Well, it's not a complaint, rather an observation - It just gives me some challenges making the system run stable by providing even more cooling ;-)
Well I have been running a
)
Well I have been running a day now, and watching the temperatures. Have seen no drastic changes in my temperature ratings. Maybe it is due to the huge cooling fan on my power supply ? It is 120CM in diameter. And before putting it in I was having cooling problems. Plus lack of extra power.. But so far the program is flying with colors...
One question.... Where can we find info on what we are seeing on the display ?
How do I use the photo below
)
How do I use the photo below to show up ?
> One question.... Where can
)
> One question.... Where can we find info on what we are seeing on the display ?
Actually the answer is in the oldest thread of this forum: take a look at
http://einsteinathome.org/node/187050
BM
BM
Thanks...Found it and
)
Thanks...Found it and somewhat understand it !
The green and blue arms represent the three LIGO instruments (at just two sites) whose data we are currently analyzing. Blue is the two instruments (2km, 4km) in Hanford, Washington. (If you look closely at the blue arms you'll see that they are actually made up of a pair of arms: half-length and full-length.) Green is the 4km instrument in Livingston, Louisiana.
The red arms represent the GEO instrument (0.6km) in Hannover, Germany.
These detector arms are correctly oriented but are NOT to scale!
If you have set your system clock to the correct time, the instruments are shown in the correct relationship to the celestial sphere (constellations). If you watch over the period of a day you'll see that they move around the celestial sphere once in the space of 24 hours.
The purple dots represent the known pulsars, which have been detected electromagnetically. Notice that these are clustered in the plane of the Galaxy (the Milky Way), predominantly towards the center of the Galaxy.
The red dots represent the know supernova remnants. These too are clustered towards the center of the Galaxy. Note that these are of particular interest because some of these supernovae may have left behind a pulsar or spinning neutron star.
Finally, the red bulls-eye circle and cross-hairs, and the right ascension and declination shown in the bottom right, indicate the current sky position being searched.
We'll add some more information to the graphics in the near future, but for the moment we're concentrating on getting the internals of the search working well!
By the way, if you go to the WORK tab and right click on a running job, you'll find a 'SHOW GRAPHICS' option. If you select this, you'll get a graphics window. Now, you can use the right and left mouse buttons (and scroll) to change the size and orientation of the celestial sphere.
> Hi, > > I've just
)
> Hi,
>
> I've just starting crunching for Einstein@home today and noticed that it is
> very hard on the cpu - temperature rises approximately 2-3 degrees C higher
> than when crunching on any other Boinc project, even CPDN which is quite cpu
> intensive. Has anyone else noticed this? I didn't see anything on these
> message boards about it.
> Unfortunately, now the cpu temp is close to the critical level where the pc
> gets unstable - I guess more cooling is required once again :-(
>
here the Same
my P4 3.2 is normally running under Full load about 65 Degrees, with a silent EKL Cooler, under SAH and under Prime also, but with Einstein@home
[/url]
Ulli, that's pretty hot! I
)
Ulli, that's pretty hot! I would definitely think about a new cooler. I have a P4 3.2 HT machine at the office, and it normally sits at around 55-60, running two BOINC projects. I'm just using the stock Intel cooler.
-Gene
Ya I had the voltage up on my
)
Ya I had the voltage up on my athlon 64 to over clock it. When running Einstein it was up around 64-64C. I dropped the voltage back down and it's fine now, but I was having some heat related issues.
I have a watercooled machine,
)
I have a watercooled machine, and the hottest I've seen it is 38C. I don't see any tempurature difference between E@H and any of the other projects.