RE: BIOS versions?

["B. P. Lame'" <blame@prolog.com>]

On Monday, 02 February, 1998 01:31, Peter Kalbus[SMTP:peteka@pep-kaufbeuren.de] wrote:
> we want to bring up a new system slot cpu
> on base of socket 7 processors. These
> cpu will mostly have standard coponents
> and some special enhansements.
> 
> For these cpu we will integrate our own bios
> on base of a standard bios (e.g.: AMI, Award
> or Phoenix).
> 
> My question is, which bios is the best
> soloution for this project. Main aspects
> of may questions are costs for a
> sources licence, code quality of the
> source code for adding features, time
> to market for new standard components
> and update policy for the source code.
> 
> Thank you for your answer
> 
> Peter Kalbus (peteka@pep-kaufbeuren.de)
> Company PEP Modular Computers Germany

System slot CPU - CompactPCI, or PCI/ISA?  Well, as far as BIOS vendors and my limited experience go, the almost infallible rule is that you get what you pay for.  This rule also seems to apply to the support you receive after your initial purchase.  Actual pricing is on a per customer, per project basis, with source code licenses being very expensive in my opinion.    Never buy sight-unseen if you can avoid it.  If yours is an industrial application, and compatibility is a concern, keep in mind that many off-the-shelf BIOS cores are written for desk-top PCs which typically do not experience the many unique and limit-pushing configurations of industrial systems.  Ergo, the programmer(s) may have taken certain shortcuts that would be undesireable in your application.  BIOS's are no longer trivial to write.  There are many industry specifications that need to be closely followed, and many subtleties/updates that should be inferred.  And, speaking as one who's had to verify some BIOS functionality, just keeping up to date on these specs is not an easy task.  I think shopping for a BIOS vendor is like shopping for a long-term business partner.  Good luck.
-- 
Brooks Lame'
Engineer
Motorola Monterey Design Center
brooks_lame@mcd.mot.com
blame@prolog.com