re: ADSL is hot (further thoughts)
Brooks,
You raise some good points.
DMT integration is clearly behind CAP; but it is catching up fast.
CAP is simpler, earlier technology - it had a head start.
But DMT integration is coming along very, very quickly.
Besides, as we all know, integration is only a means to an end,
not an end in itself (the end = low cost!!), and for ICs cost
is driven above all by volume.
(and thanks for the nice mention of Aware/ADI)
You are dead right; there is **much** more to a real product than just
a technology or a chipset.
The silicon will become a commodity; the cleverness is in the
way that becomes a product, a system.
This is the intriguing stuff...
Who here knows what chipset is inside their modem ?
(odds are it is Rockwell, but you can't buy a Rockwell modem
- you buy US Robotics, or whoever.
(And -incidentally- Rockwell didn't invent V.34, or hold patents on it;
they just delivered on it and make lots of $$ on it!!)
Finally, "FLEXIBLE" rates.
CAP cannot deliver this. There are flexible only in a very crude way:
if you do not get 6, drop to 1.5.
DMT allows for essentially Any rate between min (a few hundred k
downstream) and max (many Mbps, certainly >8).
To try a graph
rate
! D DMT goes faster @ short loops
! D
!CCCCCD Every loop gets the maximum possible (no waste)
! C D
! C D
! CCCCD
! C D DMT supports low rates for long reach
! C D
__________________________
reach |
