haas@cs.utah.edu (Walt Haas) (03/01/90)
This year should bring a number of FDDI protocol chip sets to market from various competing vendors. Looking at what it costs to install FDDI, I find that the "F" is the largest expense. In particular it costs me about ten times as much to install an ST connector as a thinnet BNC, and fiber is about ten times as expensive/foot. Therefore it would seem to me that the ideal way to network clusters of powerful workstations within a small area, such as the University of Utah Computer Science Department for a random :-) example, would be to build an interface that used the FDDI protocol but ran over thin Ethernet hardware rather than fiber. Presumably such a cluster would be tangential to a fiber FDDI ring to connect between buildings. Is anybody working on this? What's your reaction? Cheers -- Walt Haas haas@cs.utah.edu utah-cs!haas
cyc2@sword.bellcore.com (Chung Yu Chen) (03/02/90)
125 Mega baud over a pair of copper? I think you meant wave guide. Cheaper? guess again. Jason Chen
haas@cs.utah.edu (Walt Haas) (03/04/90)
In article <549@sword.bellcore.com> cyc2@sword.bellcore.com.UUCP (Jason Chen) writes: >125 Mega baud over a pair of copper? I think you meant wave guide. Cheaper? >guess again. Our broadband runs TV out to 450 MHz on RG6. That's a lot cheapr than fiber. -- Walt Haas
carroll@bcsaic.UUCP (Jeff Carroll) (03/05/90)
In article <1990Mar3.224014.26968@hellgate.utah.edu> haas@cs.utah.edu (Walt Haas) writes: >In article <549@sword.bellcore.com> cyc2@sword.bellcore.com.UUCP (Jason Chen) writes: >>125 Mega baud over a pair of copper? I think you meant wave guide. Cheaper? >>guess again. >Our broadband runs TV out to 450 MHz on RG6. That's a lot cheapr than fiber. Yup. Cutoff of the second mode (forget which one it is) in RG58 is something like 10 GHz. Twisted pair doesn't sound practical to me, though I haven't done the calculations. Anybody care to present numbers on single-mode BW and group delay. Related question: how does FDDI compare, cost-wise, to HPPI? Jeff Carroll carroll@atc.boeing.com
cyc2@sword.bellcore.com (Chung Yu Chen) (03/06/90)
In article <1990Mar3.224014.26968@hellgate.utah.edu> haas@cs.utah.edu (Walt Haas) writes: >In article <549@sword.bellcore.com> cyc2@sword.bellcore.com.UUCP (Jason Chen) writes: >>125 Mega baud over a pair of copper? I think you meant wave guide. Cheaper? >>guess again. > >Our broadband runs TV out to 450 MHz on RG6. That's a lot cheapr than fiber. ^^^^ ^^^^^^^^^^^ > >-- Walt Haas We are not talking on the same basis. Yours is an analog 450 MHz, simplex signal in the lab, while I was talking about 125 Mbaud, full duplex, digitally coded signals over a stretched distance. If you caculate the required analog bandwidth for FDDI digital signals with the necessary distance and error rate requirements, you may find that 450 MHZ is only a very small fraction. -- Jason Chen
shj@ultra.com (Steve Jay) (03/07/90)
cyc2@sword.bellcore.com (Chung Yu Chen) writes: >125 Mega baud over a pair of copper? I think you meant wave guide. Cheaper? >guess again. My company, Ultra Network Technologies, builds very high speed networking products. We can use either coax or fibre to connect our hubs together. Using good quality 75 ohm coax cable (RG-11), we can go 230 meters at 250 Mbits/sec, 330 meters at 125 Mbits/sec, or 460 meters at 62.5 Mbits/sec. Cut the distances in half for cheaper (RG-59) cable. We use two cables for full duplex operation. This is real stuff....we have machines scattered all over our building connected to our UltraNet via coax cable using our 125 Mbits/sec serial links. So, using relatively cheap copper for FDDI data rates over limited distances (~1000 feet) would seem quite feasible. Steve Jay shj@ultra.com ...ames!ultra!shj Ultra Network Technologies / 101 Dagget Drive / San Jose, CA 95134 / USA (408) 922-0100 x130 "Home of the 1 Gigabit/Second network" P.S. The above is not intended to be an advertisement (honest), just a confirmation that very fast serial data over copper is possible.