SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Technology Stocks : The *NEW* Frank Coluccio Technology Forum -- Ignore unavailable to you. Want to Upgrade?


To: MikeM54321 who wrote (2527)4/16/2001 3:15:59 PM
From: aladin  Read Replies (1) | Respond to of 46821
 
MikeM,

Absolutely not, but sorta :-). ftth is stating that for their system to work the local application program would have to cache enough of the stream to make it usable for these type of features.

My comments were generally meant to be non-specific to their server. Multicast is a transport methodolgy - not an application. However, the concept of local application caching could work with multicast as easily as unicast.

My one specific comment was that the largest backbone providers already support multicast. Why is this good? For example if I buy a UUNet connection and launch a DVD quality video service at 5.4mb then I need only buy enough bandwidth for the number of unique streams I require. So lets assume 8 streams - this translates into 43.2 mb which fits in a DS-3 which will cost me about $23,000 per month to operate.

If I were to use DF enabled servers on unicast I would need only 700kb per stream, but would need this on a per user basis! If my intended audience is more than 62 users I begin to lose money.

At 200 users I must upgrade to an OC-3 (but because I am now buying in bulk I am probably only doubling my cost - instead of tripling it).

Since they support multicast - their real advantage is the compression - or whatever they want to call the process - that permits a DVD-like image in only 700kb.

Thanks,

John



To: MikeM54321 who wrote (2527)4/16/2001 3:34:21 PM
From: ftth  Read Replies (1) | Respond to of 46821
 
Similar at a high level of network resource utilization, but I was referring to the VoD case, where multicast makes little sense.

In the original article Bernard posted a link to, they say:
"...IP multicast currently cannot accommodate video-on-demand."

VoD is by definition a user-unique stream so multicast makes no sense for this. It is unicast. IP multicast was never intended for such a use so to try and make multicast look deficient in this regard is spin.

If you combine digital fountain w/ push multicast, it adds a slight level of convenience to the multicast. It allows time-shifted start time with greater granularity that nVoD (and its 15 minute or half hour start time increments). So, in this sense, I could see how it would save an MSO some number of channels equal to the number of time-shifted nVoD streams the are constantly cycling thru. In other words, all those duplicate channels with the same movie--starting at 15 minute delay intervals--could be eliminated (except for one, of course). That seems like a fairly significant advantage and they don't even mention it.