Hi Pat,
>>Steve, this seems like a big issue yet I don't hear much talk about it.
Pat sorry for the late response, I've been very busy. DMT vs. CAP has been a big issue on this
thread for quite a while. Out in the market place, you've got a small company with limited
resources, zero PR (until recently), an ANSI Standard product, that is trialing all over the world.
CAP has been hype city. Low power/cost/real-estate and its here today! I don't know how you
can say its "here today" when you're in the
middle of trials yet redesigning chips with new never before heard of features like
RATE ADAPTATION (DMT has always been rate adaptive). The magic of marketing.
This is a great example of the importance of standards. Amati's DMT technology
is presently going through the cost reduction process via integration. If an RBOC
trails Amati's Ov8, then they can extrapolate the same level of performance
with CopperGold or any other ANSI Standard product (ADI, Alctel,Orkit).
The new ("here today") high performance CAP chip isn't out until Nov. Are
the RBOCs going deploy a "back to the drawing board" technology without
retrialing? Does this new ,untested (at Bellcore) CAP technology (lost at Bellcore) offer
such overwhelming advantages that standards should just be throw out
the door? If AT&T invented DMT and Amati CAP, I wouldn't
be writing this today and Amati wouldn't exist.
>>Are copper lines so good few will notice the difference?
What it all boils down to is how much bandwidth is required for the application? The "killer app" is the Internet. We're all frothing
at the mouth for anything greater than 28.8 or ISDN so 1.5Mb/s sounds dammed good. For how long? My hunch is
that it is like memory, we'll never have enough. Remember Billy Gates and his infamous "640k should be enough" statement. So maybe 1.5Mbs is good enough until VDSL
happens (I mean the part where the drop plow that runs fiber makes it to my neighborhood, not the technology which will happen in 97). I should live so long. Anyway, if you get Rate Adapted down
to 1Mbs from 1.5Mbs you will not notice it until ADSL apps grow. I think work at home apps will
feel the bandwidth crunch if all you have is a 64Kbs. Even .7Mbs looks small. Internet bandwith
is varible by nature so I think most users will take what they get (unless they have a choice). This is
America though. I heard on CNBC (I think Bob Metcaff) saying that the internet may be
broken up into a free/pay network to relieve congestion. ADSL modems may be marketed
the same way. It would be easy restrict bandwidth the way DMT modems are designed today (reconfigure).
Who knows if an RBOC is even thinking this way. When your trying to milk bandwidth from a
CAP modem, who would be thinking that?
Video is a bandwidth hog that an RBOC can't charge much for. 1.5Mbs is required for MPEG1 video.
Its digital so if the full bandwidth isn't available you'll see compression errors (blank squares on the screen, not snow). MPEG1 works well for movies
but won't cut it for sports. If the RBOCs are going to get into video (which I doubt any time
soon for $$ reasons) 1.5Mbs isn't going to cut it with MPEG1 compression technology.
>> I've been told a noise as moderate as a
hairdryer will disturb a CAP connection
The Bellcore tests proved that DMT will outperform CAP with all types of impairments.
I don't remember a hairdryer test though. Maybe we messed up, but I doubt it. <grin>
Some of the impairments tested were related to noise that could couple into the ADSL copper pair
from adjacent pairs within the same bundle (T1noise for example). Radio stations, impulse noise
and Cris A. noise were also a big hit (couldn't resist Cris).
>> If you wanted to go from Munich to Vienna
would you take the autobahn or one of the secondary roads?
I'd fly a Glasair.
>> What am I not
seeing?
The politics.
Later
Steve |
