Path: utzoo!attcan!utgpu!watserv1!watmath!att!pacbell!pacbell.com!ucsd!swrinde!mips!decwrl!sgi!shinobu!odin!anchor!olson From: olson@anchor.esd.sgi.com (Dave Olson) Newsgroups: comp.sys.sgi Subject: Re: Imaginary PIs [ clarification; desired i/o features ] Message-ID: <1990Sep10.023340.19858@odin.corp.sgi.com> Date: 10 Sep 90 02:33:40 GMT References: <1411@contex.UUCP> <3330@dftsrv.gsfc.nasa.gov> Sender: news@odin.corp.sgi.com (Net News) Organization: Silicon Graphics, Inc. Mountain View, CA Lines: 42 In <3330@dftsrv.gsfc.nasa.gov> buck@drax.gsfc.nasa.gov (Loren (Buck) Buchanan) writes: | In article <1411@contex.UUCP> frank@contex.UUCP (Frank Perdicaro) writes: | >A Modest Proposal | ... | >Proposed Configuration 3. | > Very minor sheet metal modification make it possible to run a full height | >drive in the top slot of a 4D 25. Make these modifications and place a drive | >there. Cut the existing SCSI cable in half, and have the top-slot drive | >serve the secondary cpu. This still provides for SCSI expansion in one cpu, | >but provides two independent cpu in one box. | I apparently missed this when originally posted. The top slot in the current PI can be used for either 2 half/high devices, or one full high, so no sheet metal (or other mods) are necessary. Of course, this doesn't do anything for the proposed 2 cpu's in a single chassis! On a related note, there has been some discussion of possibly providing an optional second SCSI bus on some future ESD (low-end) machine, similar to what was recently added on the IO3 board for the high end machines. Is this something that some of you might find useful/desirable? What other i/o features are folks looking for? (Don't bother complaining about pricing, I'm only interested in engineering related features!) Please reply via E-mail to olson@sgi.com, unless you think it is of general interest. I'm only soliciting opinions, not making any promises. I speak only for myself, etc., etc. -- Dave Olson Life would be so much easier if we could just look at the source code.