From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mailserv2.iuinc.com (qmailr@mailserv2.iuinc.com [206.245.164.55]) by puffin.external.hp.com (8.8.7/8.8.7) with SMTP id PAA08854 for ; Sat, 27 Nov 1999 15:12:20 -0700 Received: from storm.ca (dial02p57.ottawa.storm.ca [209.87.225.121]) by mail.storm.ca (8.8.8+Sun/8.8.8) with ESMTP id RAA19059 for ; Sat, 27 Nov 1999 17:14:12 -0500 (EST) Message-ID: <38401100.5FFDF373@storm.ca> Date: Sat, 27 Nov 1999 17:12:32 +0000 From: Sandy Harris MIME-Version: 1.0 To: parisc-linux@thepuffingroup.com References: <3835D808.18274C86@thepuffingroup.com> Content-Type: text/plain; charset=us-ascii Subject: [parisc-linux] Parallel port hardware List-ID: Can anyone point me to info on the electrical properties of the parallel ports on 712 or 715 machines? The Papers project (Purdue Adapter for Parallel Execution and Rapid Synchronisation) http://garage.ecn.purdue.edu/~papers/Index.html have an interesting approach to building parallel machines. Oversimplifying their fairly complex analysis (see page above for details), they argue that the hard issue is latency for synchronisation, not bandwidth, so you can do a lot of fairly interesting things with very simple synchronisation logic attached to a parallel port and software that exploits it from user space, without the overhead of kernel calls, packetisation, device drivers, .... Their most complex device has a few dozen TTL logic chips. Their simplest one handles up to about 8 machines using just a cable and their software on Linux PCs: http://garage.ecn.purdue.edu/~papers/WAPERS/ I'm thinking that 8 715s might make an interesting parallel engine, but the cable device only works if the parallel port's control lines are implemented as open collector TTL. Where can I discover whether they are?