Re: [R] Running R on a 64 bit processor

From: Prof Brian Ripley <ripley_at_stats.ox.ac.uk>
Date: Sat 29 Jul 2006 - 05:37:03 EST

Be careful not to use clock speed as a measure of computer performance. Pentium 4s (and the comparable Xeons) were intended to be run very fast, but never managed it. So a 2.4GHz P4 proved to be slower than a 1GHz PIII. Unless you are running Windows 64, the chip having some 64-bit instructions is irrelevant, and even if you were it is irrelevant to binary builds of R for Windows (which will actually access less memory than is possible under 32-bit Windows XP).

R runs large tasks much better under Linux, and there having a 64-bit CPU and 64-bit OS pay off once you have 2Gb or more of RAM.

On Fri, 28 Jul 2006, Lewis G Coggins wrote:

> Greetings,
>
> We recently obtained a new computer in our lab with a Pentium 4 3.86 GHz
> processor and 4 gb of ram running windows xp with service pack 2. After
> installing R on this machine, I ran a bit of code and found that the
> execution time was actually significantly slower than a machine running
> windows xp with an older Pentium chip 1.73 GHz and 1 gb of ram. After
> speaking with the manufacturer of the new machine, I am told that the
> processor in the new machine is 64 bit whereas I believe the processor in
> the old machine is 32 bit. I have tried to sort through the
> documentation on the CRAN page relative to performance of R under the 32
> vs 64 bit sub architecture, however, I am no computer genius and find some
> of this stuff extremely confusing.

Well, it is under the Unix/Linux section, and you are running Windows so it does not apply to you.

> In a CRAN document entitled
> "Installation and Administration" , there is reference to sub
> architecture... it reads:
>
> 8.1 Windows
> Currently the Windows build of R is a 32-bit executable. This runs happily
> on Windows 64 on AMD64 and EM64T, but is limited to (we are told) a 2GB
> address space. It will not be possible to build a native version for
> Windows 64 until suitable compilers are available, and currently
> (mid-2006) that is not in prospect.
> So my question is: are there any options to allow R to take advantage of
> the faster chip, (with 64 bit architecture), and more ram. I see in the
> documentation that a linux version of R may be able to take advantage of
> this chip... is that true what would be involved in making that work? Are
> there other options? As we are beginning to use R more and more around
> here, we may send this computer back and get a celeron 3.2 GHz chip that
> has 32 bit architecture... is this an intelligent choice?
> Thanks in advance for considering my question,
> Lew Coggins
> [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help@stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>

-- 
Brian D. Ripley,                  ripley@stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

______________________________________________
R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
Received on Sat Jul 29 05:43:01 2006

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 Sat 29 Jul 2006 - 08:17:20 EST.

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