[Rd] 64 bit R for Windows

[Martin Maechler]

>>>>> "PD" == Peter Dalgaard <p.dalgaard at biostat.ku.dk>
>>>>>     on 02 Sep 2005 18:48:24 +0200 writes:

    PD> "Milton Lopez" <mlopez at iattc.org> writes:

    >> I appreciate the update. We will consider using Linux,
    >> which leads me to one more question: what is the maximum
    >> RAM that R can use on each platform (Linux and Windows)?
    >> 
    >> Thanks again for your prompt responses.

    PD> On Win32, something like 3GB. Maybe a little more on
    PD> Linux32, but there's a physical limit at 4GB.

for a *single* object, yes.  However (and Peter knows this
probably better than me ..), R's workspace can be very much
larger which makes it realistically possible to start *using* R
functions on objects of around 4GB.
Someone (Venables & Ripley ?) have once stated the rule of thumb
that you need about 5--10 times the size of your "single" large
object for your "workspace", because of (intermediate) copies,
sometimes multiple ones are needed, or at least part of the
current implementations of many basic algorithms / functions.
In other words, if you got a 32 GB RAM, you could probably start
to work with objects of the size of (a little less than) 4GB
relatively comfortably.

Martin Maechler

    PD> On Linux 64, the motherboards set the limit in practice,
    PD> 32GB systems have been reported working and I think at
    PD> least 64GB should be possible. I seem to recall that the
    PD> maximum _virtual_ memory is not quite 2^64, but it will
    PD> be pretty huge (2^48, 256TB)?.