[BioC] large data set under 32 bits architecture

[Marcus Davy]

Hi,
As far as I am aware each process on 32-bit architecture can only
address theoretically a maximum of approximately  4 gigabytes of memory
(2^32 bits), even with the PAE extention allowing additional memory
available to the architecture. See
the R 1.9.0 windows FAQ for details on allowable memory.

http://www.stats.ox.ac.uk/pub/R/rw-FAQ.html#There%20seems%20to%20be%20a%20limit%20on%20the%20memory%20it%20uses!


A previous (disjointed) thread recently suggests you *may* run into
difficulties with 330 CEL files.

https://www.stat.math.ethz.ch/pipermail/bioconductor/2004-June/004915.html

People will be making algorithms more memory efficient, although to
future proof the 32-bit addressable memory limits, you could investigate
64-bit architecture options with a suitable operating system if you are
run into problems.


marcus



>>> <Phguardiol at aol.com> 22/06/2004 6:15:52 AM >>>
Hi all,
we are using a 32 system under Win2003 Server Enterprise with a quadri
Xeon and 24 Go RAM memory. I d like to know if this configuration can
allow us to work with affy RMA and 330 CEL files from U133a affy
chips.... next step being to use GCRMA.
or should we move to a Linux OS with this server to run all these in 1
round ? or any other strategy to propose ?
thanks
Phil

_______________________________________________
Bioconductor mailing list
Bioconductor at stat.math.ethz.ch 
https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor

______________________________________________________

The contents of this e-mail are privileged and/or confidenti...{{dropped}}