From owner-freebsd-isp Sat Nov 16 08:00:16 1996 Return-Path: owner-isp Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id IAA04088 for isp-outgoing; Sat, 16 Nov 1996 08:00:16 -0800 (PST) Received: from etinc.com (et-gw-fr1.etinc.com [204.141.244.98]) by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id IAA04079 for ; Sat, 16 Nov 1996 08:00:12 -0800 (PST) Received: from ntws (ntws.etinc.com [204.141.95.142]) by etinc.com (8.6.12/8.6.9) with SMTP id LAA13615; Sat, 16 Nov 1996 11:06:07 -0500 Date: Sat, 16 Nov 1996 11:06:07 -0500 Message-Id: <199611161606.LAA13615@etinc.com> X-Sender: dennis@etinc.com X-Mailer: Windows Eudora Version 2.0.3 Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" To: John Hay From: dennis@etinc.com (dennis) Subject: Re: Decision in Router Purchase Cc: isp@freebsd.org Sender: owner-isp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk >> >> > > >> Is SDL forthcoming with the technical information required >> > > >> to write and maintain drivers? >> > > >> ... >> > > >Yes, I have written the driver with the information they have given me. >> > > >They have been very helpful. They even supplied and shipped 2 N2 boards >> > > >and a N2pci board from the USA to me here in South Africa. >> > > > >> >> Have you done any systematic tests of the relative performance of >> the N2 (which uses the ISA bus) and the N2pci (which uses the PCI bus) ? > >No, I used a 486 for the ISA development and a Pentium for the N2pci >development. We did testing (we had a dual-T1 PCI board on the table in March), and the results were not significant for 1 or 2 T1lines, which is pretty much why we scrapped it and decided to do something different. Given that the design of the processor allows for full bandwidth utilization, the overall gain is minimal. the vast majority of the processing time is in software. When you get to about 10Mbs aggragate bandwidth (which is about 5 T1 with average full-duplex usage of 2Mbs), you get a worthwhile gain. The ethernet ISA-to-PCI differences are much more dramatic, mainly because the ISA cards had processors and designs that were not capable of filling the bandwidth (ie, there was no auto-transmit, generally only 2 transmit buffers, etc)..plus you're talking about 10Mbs. Dennis