Re: [Rd] Whine in dual core Windows PC if R uses full CPU capacity

From: Duncan Murdoch <murdoch_at_stats.uwo.ca>
Date: Mon 15 Jan 2007 - 18:07:10 GMT

On 1/15/2007 12:02 PM, Ulrike Grömping wrote:
> Dear R-developers,
>
> the following may seem a weird question for R-devel – I try anyway, because it
> does seem to be related to usage of R and R's way of interacting with the
> hardware.
>
> When running R (2.4.0) on my new dual core Windows XP system (Intel Core 2 Duo
> E6600, Mainboard ASUS P5B, BeQuiet Power Unit), I observe various types of
> sounds (not related to fans) when running at full usage of at least one of the
> cores.
>
> The most annoying sound – a high-pitched noise like a whine – I have so far
> observed running simulations in R only and couldn’t reproduce with any other
> software (sometimes it’s also more a rustle than a whine, that one is not
> disturbing). I can clearly track it down to particular instances of R – as
> soon as I left-click on a blue window bar in the affected R-Gui, the noise
> stops (and there are sometimes other instances of R running in parallel which
> are completely unrelated to the whine). I’ve also stopped R when the computer
> just whined and started SAS, fully using the CPU as well – no whine. Shut down
> SAS, restart R à whine again. Apparently, it can also be related to what
> exactly R is doing at the particular moment. For example, I just discovered
> that simulations of one type were finished, because the computer – running
> another simulation under full load on R in a quiet way – started whining at
> the moment it switched to the new type of simulation.
>
> The hardware technician I’ve asked about this issue does not have any idea and
> suspects R to do something peculiar that causes the noise. It’s not the
> mainboards fault (that has already been changed for another reason, and the
> behavior remains). I can’t definitely locate the source of the sound. I
> suspect that it’s the Power Unit, but there are so many parts close together
> that I am not really sure.
>
> I do not know enough about interactions of R with hardware to know whether it
> is even possible for anyone deeper into the matter to develop an idea what
> could be behind this behavior. For the moment, I can get rid of the whine by
> using the mainboard tool „AI Gear“ to reduce the processor’s speed from 2.4GHz
> to 1.7GHz when I want to concentrate – then the whine is gone (but the
> simulation is slower).

I haven't heard this from R, but on occasion I've had noisy CPUs. I suspect it's a case of some repetitive calculation leading to repetitive electric fields which happen to cause physical resonance of some part of your system. That agrees with the observation that changing the speed changed the noisiness. Another way to get rid of it would be to change the contents of the loop: make it run faster or slower and the noise should go away.

Duncan Murdoch

>
> Does anyone have any idea – or hints regarding what else I could think about ?
>
> Regards, Ulrike
>
> ******************************
> Prof. Dr. Ulrike Grömping
> Fachbereich II
> TFH Berlin
> Luxemburger Str. 10
> 13353 Berlin
> mail: groemping@tfh-berlin.de
> www: www.tfh-berlin.de/~groemp/
> ******************************
>
> ------- End of Forwarded Message -------
>
> ******************************
> Prof. Dr. Ulrike Grömping
> Fachbereich II
> TFH Berlin
> Luxemburger Str. 10
> 13353 Berlin
> mail: groemping@tfh-berlin.de
> www: www.tfh-berlin.de/~groemp/
> ******************************
>
>
>
> [[alternative HTML version deleted]]
>
>
>
> ------------------------------------------------------------------------
>
> ______________________________________________
> R-devel@r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel



R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel Received on Tue Jan 16 05:24:09 2007

Archive maintained by Robert King, hosted by the discipline of statistics at the University of Newcastle, Australia.
Archive generated by hypermail 2.1.8, at Mon 15 Jan 2007 - 20:31:20 GMT.

Mailing list information is available at https://stat.ethz.ch/mailman/listinfo/r-devel. Please read the posting guide before posting to the list.