• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Riddle me this...

dringdahl

Senior member
OK been testing some dual xeon cluster computers. Here is an interesting fact.

Running Seti (2 copies)

CPU1 Fan: 5400 RPM (min = 2848 RPM, div = 2)
CPU2 Fan: 5443 RPM (min = 2848 RPM, div = 2)
Sys Fan: 0 RPM (min = 1424 RPM, div = 4)
Sys Temp: +32°C (high = +50°C, hyst = +45°C) sensor = thermistor
CPU2 Temp: +43.0°C (high = +85°C, hyst = +80°C) sensor = thermistor
CPU1 Temp: +41.0°C (high = +85°C, hyst = +80°C) sensor = thermistor

Running FaD (2 Copies)

CPU1 Fan: 5487 RPM (min = 2848 RPM, div = 2)
CPU2 Fan: 5443 RPM (min = 2848 RPM, div = 2)
Sys Fan: 0 RPM (min = 1424 RPM, div = 4)
Sys Temp: +31C (high = +50C, hyst = +45C) sensor = thermistor
CPU2 Temp: +46.5C (high = +85C, hyst = +80C) sensor = thermistor
CPU1 Temp: +43.5C (high = +85C, hyst = +80C) sensor = thermistor


So does this mean that FaD is doing more computation??? Less memory access?? If you run top they are both showing 100 % utilization.

TIA

BTW This is a 2.8ghz setup.
 
I think it depends on which parts of the CPU are being utilized more. SETI seems to be more memory bandwidth limited, so perhaps it's not able to use as much CPU as it could otherwise because of the bandwidth (I really have no idea, please enlighten me someone who knows). Folding@Home with the Gromacs core is said to be in some cases even more stressful than Prime95, which is considered one of the most stressful non-real-work-related things a CPU could do. Of course, the advantage of Prime95 is that it is checking the results of its calculations against known good results, so you'll know if something is screwed up. Of course, Folding@Home has a way to tell if WUs are getting corrupted due to an overclocking situation, but not in the same way that Prime95 can, probably since the latter deals with exact previously-known numbers.
 
Hmmm, while were dealing with interesting facts, my P4 w/ HT running a single process is 3-4C cooler than while running two F@H consoles (with diff machine Id's).

I'm guessing that FAD is using a part of the CPU that is closer to the thermistor, thats another possibility.
 
FaD makes all my cpu temps go up 3-4 C as compared to running SETI I. all non-hyprethreading boxes but the temp is definately a constant rise topping out at 3-4 C higher.
 
Back
Top