home .. forth .. misc mail list archive ..

Re: Speculation on ARM licensed by DRAM companies


>>>>> "wmor1" == wmor1  <wmor1@student.monash.edu.au> writes:

    wmor1> Question? how hard (extra equipement) would it take to
    wmor1> drive ethernet off the F21 chip.  I had thought about
    wmor1> protocols that would avoid the collision aviodance problem
    wmor1> on ethernet that would get around the performance drain of
    wmor1> to many senders on the network.  Actually I rember
    wmor1> something about being able to implement rings and ethernet
    wmor1> over the same networks.  The reason why is that something
    wmor1> like it would make an interesting bus.

The collision problem on ethernet and the collision problem in hash
tables are both essentially the same thing, except that one is a
collision in address space and the other is a collision in time
space.  Schemes to avoid collision invariably involve some sort of
token passing protocol.  The overhead for hash tables implies
significant efficeincy degradation at over about 75 % utilization.
Ditton for ethernet collisions.  It should not seem strange to you
then that the overhead of token passing is about the same also.  The
trick then is to use a bus with 33% more bandwidth than you actually
need.  This will eliminate most collisions, or alternatively provide
the extra overhead for a robust token passing protocol.  I prefer the
ethernet approach, since it is amenable to RF transmission over open
air also.

-- 
--------  "And there came a writing to him from Elijah"  [2Ch 21:12]  --------
Robert Jay Brown III  rj@eli.wariat.org  http://eli.wariat.org  1 847 705-0424
Elijah Laboratories Inc.;  37 South Greenwood Avenue;  Palatine, IL 60067-6328
-----  M o d e l i n g   t h e   M e t h o d s   o f   t h e   M i n d  ------