[Bioclusters] fortran 9x compilers for linux clusters

andy law (RI) bioclusters@bioinformatics.org
Thu, 3 Oct 2002 09:29:40 +0100


Does anyone have any experience they'd like to share on the topic of Fortran 9x compilers for linux systems?

Specifically, I have a 30 CPU intel PIII cluster - each of the 15 dual-processor nodes is running RedHat. One of my 'customers' has some code developed to run and previously succesfully tested on a Cray system in his native Sweden. We are interested in compiling and benchmarking it locally. There are several other statistical packages developed by the quantitative genetics group and written in fortran that we would also like to try out on the system once we get more proficient. The latter are written in f77 and so we could just use g77. However, the former (and some of the newer stuff from the Quantitative Genetics groups around the world) is written in F90, so I guess we will have to go commercial.

We don't have anything fancy on the system in terms of parallel virtual machines yet, but will want to look into that. In fact, we have MPICH and a crude form of task-farming system and that's about it. I notice that some of the compilers' blurb mentions OpenMP support - only one mentions MPI specifically. I have no idea about the implications of choosing one MP architecture over another and if two can be run in parallel (irony not intended!).

For extra information. One project that we are committed to trialling is one for the parent organisation (a UK Research Council) which involves a GRID-type system, utilising spare processor cycles to run sequence similarity searches (blast, basically). I suspect we will be using Condor plus some extra bits and bobs to implement that.

Given all that background information, what options do I have and what should I take into account before making a purchase?

Thanks for any help you can provide.


Dr. Andy Law
Head of Bioinformatics - Roslin Institute

Unfortunately, legal niceties require me to add the following to this message...

The information contained in this e-mail (including any attachments) is confidential and is intended for the use of the addressee only.   The opinions expressed within this e-mail (including any attachments) are the opinions of the sender and do not necessarily constitute those of Roslin Institute (Edinburgh) ("the Institute") unless specifically stated by a sender who is duly authorised to do so on behalf of the Institute.