From owner-freebsd-hackers Fri May 16 13:48:56 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id NAA19009 for hackers-outgoing; Fri, 16 May 1997 13:48:56 -0700 (PDT) Received: from plains.nodak.edu (tinguely@plains.NoDak.edu [134.129.111.64]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id NAA19003 for ; Fri, 16 May 1997 13:48:52 -0700 (PDT) Received: (from tinguely@localhost) by plains.nodak.edu (8.8.5/8.8.5) id PAA19763; Fri, 16 May 1997 15:48:29 -0500 (CDT) Date: Fri, 16 May 1997 15:48:29 -0500 (CDT) From: Mark Tinguely Message-Id: <199705162048.PAA19763@plains.nodak.edu> To: black@zen.cypher.net, sthaug@nethelp.no Subject: Why use ATM, (was Cluster Computing in BSD) Cc: hackers@FreeBSD.ORG Sender: owner-hackers@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Steinar Haug says: > I haven't seen any sign of inexpensive 155 Mbps ATM NICs, why do you > think inexpensive 622 Mbps ones are coming soon? they are coming down in price, but not nearly low enough to be mainstream. the cheapest cards are the ones that use host memory for packet segmentation and re-assembly. I wish I had a performance study to see how much speed is lost using host memory for SAR operations. Ben Black asks: > and why would you want to use something as inefficient as ATM for a > cluster interconnect? in-efficiency is not a concern problem IF the end result is faster than anything else on the market. ATM, lost that window of opportunity and today is not the fastest thing on the market because everone wants IP. I think even Sonet PPP/frame relay will upstage ATM. yeah, this prob. belongs in -chat or -atm. --mark.