Path: utzoo!attcan!uunet!image.soe.clarkson.edu!news From: nelson@sun.soe.clarkson.edu (Russ Nelson) Newsgroups: comp.sys.ibm.pc Subject: Re: Steve Gibson's 386sx column Message-ID: Date: 13 Feb 90 14:57:31 GMT References: <29140@amdcad.AMD.COM> Sender: news@sun.soe.clarkson.edu Reply-To: nelson@clutx.clarkson.edu Organization: Clarkson University, Potsdam NY Lines: 30 In-reply-to: phil@pepsi.amd.com's message of 12 Feb 90 20:22:51 GMT In article <29140@amdcad.AMD.COM> phil@pepsi.amd.com writes: 4) The Gateway Ethernet interface has no high memory footprint? Anyone care to comment on the Gateway? I've never seen it for sale. It's probably I/O based and doesn't have a boot prom installed. There are quite a few of these boards, among them the 3c501 and ni5010 (I wouldn't put either of these in a 386), Novell NE[12]000 and clones. I don't know how much performance you lose when you use an I/O based Ethernet card. There is also the ni6510, which is a bus mastering Ethernet card. I've got two of these (donated to Clarkson by Racal/Interlan so that I could write a packet driver for it), and they're pretty slick. It will write its packets anywhere in the first 16 Mbytes, so you could put your packet buffers up in extended memory, or, if you have a 386 control program that does expanded memory mapping, you could put your buffers in expanded memory. Their prototype cards wouldn't work in my prototype Zenith Z-386 -- does that surprise anyone? I have no idea whether the real cards will work in real 386ses. If you're thinking about buying one of these, contact Racal/Interlan to see if the board will work in your machine. They have a list of machines that the board is known to work in. For example, it works in a Z-248, but not in a Z-286-LP12 (that is, the prototype board doesn't work). -- --russ (nelson@clutx [.bitnet | .clarkson.edu]) Russ.Nelson@$315.268.6667 Violence never solves problems, it just changes them into more subtle problems