[R] Maximum amount of memory

Peter Dalgaard p.dalgaard at biostat.ku.dk
Mon Mar 21 19:19:52 CET 2005


Tim Cutts <tjrc at sanger.ac.uk> writes:

> On 21 Mar 2005, at 4:42 pm, marvena at tin.it wrote:
> 
> > Hi,
> > I have a problem:I need to use the maximum amount of memory in order to
> > perform a very tough analysis. By purchasing the suitable computer,
> > what's
> > the maximum amount of memory obtainable in R?
> 
> Assuming that R is happy to use 64-bit memory pointers, the limit will
> be your wallet.  You could buy an SGI Altix and just keep buying more
> and more memory for it.  I don't know the limit - I know that SGI have
> sold one machine in Japan with 13 terabytes of memory.  We have two of
> them here with 192 GB of RAM each, but I haven't tried R on them yet -
> they're used for other things.

Actually, before the wallet-limit, you might bump into the limit on
the size of a single object, about 2G-items (i.e 16GB with 8-byte
doubles). We could fairly easily raise that limit, but beneath it is
another restriction, namely that Fortran subroutines tend to work with
integer sizes, so even if we had larger objects, it might be difficult
to do anything with them.
 
> Whether such a course of action is sensible is another matter.  Large
> memory machines rapidly become *extremely* expensive; once you have to
> use DIMMs larger than 1GB each, the price becomes prohibitive.
> Consider spending the same amount of money on employing several
> programmers and/or statisticians to break your problem down into
> smaller tasks than are tractable on smaller machines.
> 
> Our 192 GB machine cost quite a lot more than 192 desktop PCs with 1GB
> of RAM each.  In fact, the memory becomes so expensive the rest of the
> machine is virtually free, in comparison.  :-)
> 
> If you can get away with more modest amounts of memory, then a machine
> like the HP DL-585 might suit you - a quad processor Opteron, which
> can take up to 32GB or so of memory.  Fairly modest price.
> 
> Tim
> 
> -- 
> Dr Tim Cutts
> Informatics Systems Group, Wellcome Trust Sanger Institute
> GPG: 1024D/E3134233 FE3D 6C73 BBD6 726A A3F5  860B 3CDD 3F56 E313 4233
> 
> ______________________________________________
> R-help at 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
> 

-- 
   O__  ---- Peter Dalgaard             Blegdamsvej 3  
  c/ /'_ --- Dept. of Biostatistics     2200 Cph. N   
 (*) \(*) -- University of Copenhagen   Denmark      Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk)             FAX: (+45) 35327907




More information about the R-help mailing list