[BioC] related question about computers and bioC

Phguardiol at aol.com Phguardiol at aol.com
Wed Jan 14 09:48:24 MET 2004


Hi all,we are planning to buy a workstation for our microarray platform which will analyse Affymetrix chips, Agilent chips and potentially other chip types. I ve been using gcrma under XP Pro with a DELL Xeon 2.4 and 3.62 Go RAM and it was difficult or impossible until now to run 24 HU133A chips because of memory size limitation (probably because we were limited with XP). We are thinking about moving to Linux OS. I dont have a large experience with workstations and large dataset (more than 100 chips). I was wondering how much RAM memory we should think of for this system if we are planning to run let say 200 chips with either RMA or probably with GCRMA or expresso with RMA2. Is there a threshold above which it is not necessary to increase the RAM size ? I have seen a WS with potentially 24 Go and I was wondering if it would make a difference if it was with "only" 12 Go ? Could you confirm that it would be better to have a 64 bit system ? last question which is in part related to the previous one: should we consider a WS with a Pentium a Xeon or an Itanium processor ? By the way, I ve seen people also working on Sun system servers, what would be the main difference if compared to a DELL or HP Workstation a part the fact that you can share data maybe more easily ? thanks for sharing your experience with mePhilippe



More information about the Bioconductor mailing list