home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.dcom.cell-relay
- Path: sparky!uunet!walter!att-out!pacbell.com!iggy.GW.Vitalink.COM!nocsun.NOC.Vitalink.COM!RODIN!dgr
- From: dgr@RODIN (Daniel Robinson)
- Subject: Re: Computers dont like ATM?
- Message-ID: <1992Nov18.162627.15214@NOC.Vitalink.COM>
- Sender: usenet@NOC.Vitalink.COM (Usenet News Admin)
- Nntp-Posting-Host: rodin.eng.vitalink.com
- Organization: Vitalink Communications, Fremont, California
- References: <1992Nov18.025754.14749@trl.oz.au>
- Date: Wed, 18 Nov 1992 16:26:27 GMT
- Lines: 26
-
- In article <1992Nov18.025754.14749@trl.oz.au> lampard@titan.trl.OZ.AU (Greg Lampard) writes:
- +His claim was that small
- +packets are good for delay-sensitive applications, but they make a lot
- +of unnecessary work for the computer that has to receive them, since it
- +has to process a header every 53 bytes,
-
- Much of that processing will be done in silicon, not by a stored
- instruction processor.
-
- +He also seemed to be suggesting that this storage and retrieval process
- +could become a bottleneck (at 100s of Mbit/s?) because of the limitations
- +of memory bandwidth.
-
- This is the right number for a stored instruction processor that is
- not a Cray.
-
- +He was advocating variable length packets (an idea
- +called PTM - packet transfer mode - that IBM has proposed), because it
- +puts the complexity in the switch rather than in the receiving
- +terminal.
-
- It is hard to see where a variable length packet will put the complexity
- in the network mesh node, not the terminal node. Remember that this
- the company that brought us source routing.
-
- Daniel Robinson
-