Xref: utzoo comp.unix.sysv386:1409 comp.os.minix:12950 Path: utzoo!attcan!uunet!aplcen!wb3ffv!ka3ovk!raysnec!shwake From: shwake@raysnec.UUCP (Ray Shwake) Newsgroups: comp.unix.sysv386,comp.os.minix Subject: Re: 286 -> 386sx Upgrades Message-ID: <106@raysnec.UUCP> Date: 19 Oct 90 14:05:55 GMT References: <40019@cci632.UUCP> <1990Sep21.144108.1098@pmafire.UUCP> <1990Oct17.175700.24704@ico.isc.com> Followup-To: comp.unix.sysv386 Distribution: na Organization: IRS/CI - Technical Solutions Branch Lines: 21 rcd@ico.isc.com (Dick Dunn) writes: >I tried several times to upgrade an old NEC APC-IV. That's their basic AT >clone from some years ago. The first attempts at upgrading were add-in 386 >boards, either plugging directly into the CPU socket or plugging into the >bus with a cable to the CPU socket. Either they didn't fit (given space >constraints on the board) or didn't work (somebody playing BIOS games, I >suspect; that always gets UNIX). None of this surprises me based on my experiences with NEC's 386/20. I'd considered upgrading mine given the absence of memory cache and limited (and EXPENSIVE) memory expansion potential. Given my own suspicions regarding form factor, screw bracings and connectors - and considering that floppy, serial and parallel controllers run off the motherboard - I decided that it was not worth it. Hey, it's not *that* bad a box, and it's certainly adequate to support my DOS and Xenix work, not to mention occasional ISC UNIX work. I can always buy a more powerful box for more serious work. Given the radical surgery obviously called for with the APC-IV (and with some other boxes) it may be appropriate just to keep it around for AT-class use.