home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!pacific.mps.ohio-state.edu!linac!att!rutgers!apple!apple!netcomsv!mork!fisk
- From: fisk@netcom.com (Benjamin Fisk)
- Newsgroups: comp.dcom.cell-relay
- Subject: Re: Question
- Summary: Its later than you think
- Message-ID: <r1hmm#q.fisk@netcom.com>
- Date: 24 Jul 92 17:24:10 GMT
- References: <1992Jul17.144218.29461@walter.bellcore.com>
- Organization: Netcom - Online Communication Services (408 241-9760 guest)
- Lines: 25
-
- In article <1992Jul17.144218.29461@walter.bellcore.com>, Shikhar Bajaj <bajaj@thumper.bellcore.com> writes:
- >
- >
- > However, I believe that efficiency is not the major issue which needs
- > to be addressed. With fiber optic bandwidths in the Terabit range, at
- > least in the short to mid-term, we will have bandwidth to burn. Most
- > commercial applications and operating systems implementations cannot
- > even "fill the pipe" with data (at least for workstation class machines
- > and below) so what is the point about worrying about efficiency. This
- > may seem shortsighted and appear as if I am simply deferring the problem
- > to a later date, but I do not believe that IF ATM becomes widespread
- > throughout the network, current traffic patterns will persist. For
- > example, multi-media traffic (i.e. real-time video and voice) is so
- > unlike data traffic patterns that I do think that major pattern shifts
- > would occur. But, we don't know what these potential patterns may be.
- >
- The infinite pipe argument has always tagged along with fiber. But for those
- who have worked with developing fiber optic networks know that capacity
- runs out just as fast on fiber as any other media. The bottom line is
- transmission is expensive to install and it is always in great demand.
-
- However, I do agree that applications are tending twoards longer cell
- sizes and that ATM cells will prove out fine in the long run.
- --
- Ben Fisk : fisk@netcom.COM
-