home *** CD-ROM | disk | FTP | other *** search
- Path: aplcomm.jhuapl.edu!dunlap
- From: dunlap@aplcomm.jhuapl.edu (Dunlap W. Brian F1D x8255 )
- Newsgroups: comp.sys.amiga.programmer
- Subject: Re: Amiga and Real Time Systems
- Date: 10 Apr 96 00:45:39 GMT
- Organization: Johns Hopkins University Applied Physics Lab, Laurel, MD, USA
- Message-ID: <dunlap.829097139@aplcomm.jhuapl.edu>
- References: <9604091927.AA000mo@harrison.demon.co.uk> <4keoj5$91e@swrinde.nde.swri.edu>
- NNTP-Posting-Host: aplcomm.jhuapl.edu
- X-Newsreader: NN version 6.5.0 CURRENT #2
-
- We are integrating a shiny new
- SGI Challenge multi-processor almost-top-of-the-line
- machine into our lab as the central real-time computer. The
- SGI real-time extensions are named React. Today I found
- out that React can't react as fast as a 1986 A1000 that it
- controls (the A1000 is an embedded processor for a transmitter
- rack). The interrupt response time (maximum latency) for the
- SGI is specified as 250 microseconds from the time the interrupt
- line on the VME expansion cage is asserted.
-
- The 7 MHz A1000 has been proven to have a maximum interrupt response
- time of 110 microseconds from an interrupt line on its Zorro I
- expansion cage.
-
- Go figure. :-/
-
- -Brian
-
-