Network end users to pull down 2 gigabytes a day, continuously?

Sat Jan 13 10:45:50 UTC 2007

Of course, this below is for inter-domain. There is no shortage of  
multicast walled garden
deployments.

Regards
Marshall

On Jan 12, 2007, at 7:44 PM, Marshall Eubanks wrote:

>
>
> On Jan 12, 2007, at 10:05 AM, Frank Bulk wrote:
>
>>
>> If we're becoming a VOD world, does multicast play any practical  
>> role in
>> video distribution?
>
> Not to end users.
>
> I think multicast is used a fair amount for precaching; presumably  
> that would increase in this scenario.
>
> Regards
> Marshall
>
> P.S. Of course, I do not agree we are moving to a pure VOD world. I  
> agree with Michal Krsek in this regard.
>
>>
>> Frank
>>
>> -----Original Message-----
>> From: owner-nanog at merit.edu [mailto:owner-nanog at merit.edu] On  
>> Behalf Of
>> Michal Krsek
>> Sent: Wednesday, January 10, 2007 2:28 AM
>> To: Marshall Eubanks
>> Cc: nanog at merit.edu
>> Subject: Re: Network end users to pull down 2 gigabytes a day,  
>> continuously?
>>
>>
>> Hi Marshall,
>>
>>> - the largest channel has 1.8% of the audience
>>> - 50% of the audience is in the largest 2700 channels
>>> - the least watched channel has ~ 10 simultaneous viewers
>>> - the multicast bandwidth usage would be 3% of the unicast.
>>
>> I'm a bit skeptic for future of channels. For making money from  
>> the long
>> tail, you have to have to adapt your distribution to user's needs.  
>> It is not
>>
>> only format, codec ... but also time frame. You can organise your  
>> programs
>> in channels, but they will not run simultaneously for all the  
>> users. I want
>> to control my TV, I don't want to my TV jockey my life.
>>
>> For the distribution, you as content owner have to help the ISP  
>> find the
>> right way to distribute your content. In example: having  
>> distribution center
>>
>> in Tier1 ISP network will make money from Tier2 ISP connected  
>> directly to
>> Tier1. Probably, having CDN (your own or pay for service) will be  
>> the only
>> one way for large scale non synchronous programing.
>>
>>             Regards
>>                     Michal
>>
>>
>