[c-nsp] 100 meg throughput

[Chris Cappuccio]

This is absurd.   There is no reason you can't fill a 100Mbit FE (Full duplex)
with the theoretical max 148810 packets per second in _each_direction_ if
they're all 64bytes.  If you plug one port on your smartbits directly into
the other, and leave things like the inter-packet gap standard, you will
achieve exactly 148810pps at the lowest packet sizes.

TCAM is clearly fast enough and cheap enough these days that even the cheap
$20 10/100 switches are able to achieve full line rate.  (Well, this doesn't
explain why 6500/7600 cards are so expensive :)

Joe Shen [sj_hznm at yahoo.com.cn] wrote:
> To our measurement with Smartbit, throughput of FE
> veries with packet size. With 64Byte packet, it could
> only reach 67Mbps or so; with 1500Byte packet, it
> could reach much close to 100Mbps. 
> 
> The same result with GE.
> 
> Joe 
> 
> 
> --- Paul Stewart <pstewart at nexicomgroup.net>:
> 
> > Hi everyone...
> > 
> > I realize this is a loaded question with many
> > variables but here
> > goes....
> > 
> > How much throughput should we expect to see on a
> > copper FastE
> > connection?  This connection is used to backhaul
> > Internet traffic and is
> > starting to hit the 80 meg mark on occasion....
> > 
> > Thanks in advance,
> > 
> > Paul Stewart
> > IP Routing/Switching
> > Nexicom Inc.
> > http://www.nexicom.net
> > 
> > _______________________________________________
> > cisco-nsp mailing list  cisco-nsp at puck.nether.net
> > https://puck.nether.net/mailman/listinfo/cisco-nsp
> > archive at
> > http://puck.nether.net/pipermail/cisco-nsp/
> > 
> 
> 
> __________________________________________________
> ?????????????????????????????
> http://cn.mail.yahoo.com
> _______________________________________________
> cisco-nsp mailing list  cisco-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
-- 
"You were about to change the channel when God healed you" -- Benny Hinn