Dear Christine,
I can't comment on whether RAID will speed up your
analysis - though as a general rule, using SCSI disks
will provide some performance increase over IDE ones.
Available RAM will impact performance, though presumably
there'll be a plateau above which adding more RAM won't
affect things. Then performace of hard disks becomes an
issue, as you try to read data into memory as fast as you can.
I assumed that RAID just gave you disk redundancy, rather
than performance increases, but could be wrong - wouldn't
be the first time :)
As far as Linux and RAID arrays goes - I noticed in that
SuSE are selling their own RAID array systems:
http://www.suse.de/de/hardware
So you should be OK there.
Yours,
Jon.
_____________________________________________________
Jonathan Brooks Ph.D. (Research Fellow)
Magnetic Resonance and Image Analysis Research Centre
University of Liverpool, Pembroke Place, L69 3BX, UK
tel: +44 151 794 5629 fax: +44 151 794 5635
On Mon, 22 May 2000, Christine Preibisch wrote:
>Dear SPM'ers,
>
>A question concerning computer hardware once again.
>
>Reviewing previous emails regarding this issue I found a mail from Joe
>Devlin (14.02.00) where he recommended RAID arrays for further acceleration
>of analysis. Is it worthwhile using them if the RAM is big enough already
>(e.g. 1GB)? Can anybody comment on how much faster this systems really are,
>and which RAID level should be used?
>
>Does anybody have experience using a RAID array (of level 5) under (Suse)
>Linux? Are any RAID arrays on Intel-based PC's are properly supported by any
>Linux species?
>
>Any comment would be highly appreciated.
>
>Thank you,
>
>Christine Preibisch
>
>Universitdt Frankfurt
>ZRAD - Institut f|r Neuroradiologie
>Schleusenweg 2-16
>60 528 Frankfurt
>
>Tel: ++49 69 6301 4651
>Fax: ++49 69 6301 5989
>
>email: [log in to unmask]
>
>
>
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|