Silicon Investor (SI) -- The First Internet Community

STOCKTALK

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 : Next Generation Internet, Internet2 & other video networks -- Ignore unavailable to you. Want to Upgrade?

To: Scott C. Lemon who wrote (63)	9/10/1999 8:33:00 AM
From: Frank A. Coluccio	Read Replies (1) \| Respond to of 74

Hi Scott, your question has an interesting texture to it. It inadvertently points to the potentiality of the double edgedness of the approach: "...are there pages on their web site that you can recommend that provide some real insight into the uniqueness* of their solution...?"* ---- I'm witnessing a very interesting phenomenon these days. I'm seeing almost everything, soon possibly very close to everything, being converted from legacy approaches to IP, or becoming IP-enabled. This holds true for even the staunchest of hold outs, like IBM mainframes, where channel extension had previously been done of SNA lines, then combined with ATM and Frame Relay, but now going to channel and front end IP interfaces. Not only the external interfaces of the main frames, but even the mainframe's logical partitions (LPARs) and internal fabrics are receiving IP addresses. Soon, voice will take this approach. And still, for those applications which require isochronicity, even in the largest IP Clouds, these applications are most often still mapped to ATM virtual circuit flows. As the capacities of cores increase with DWDM enablement, and as MPLS matures and gains wider spread acceptance, however, we will see an offloading of many of those deterministic flows from ATM to label switching over IP. What this says is that there is still a window left for circuit switched, and shared ATM networks. The question one must ask, however, is, How long will this window stay open for end user oriented applications? Regards, Frank Coluccio

To: Scott C. Lemon who wrote (63)	9/10/1999 8:43:00 AM
From: TheSlowLane	Respond to of 74

Scott, Some good questions, I will attempt to address them. The company is offering to lease a complete, bundled solution that encompasses hardware, software, network gear, bandwidth, operations, maintenance and user support - all for a fixed monthly fee. Sort of like buying a cell phone, if you will. I think in many cases, the intended use will be within an organization. However, Sean Coughlin of Cinetech has already said that he wants to use this technology to communicate with his customers (film production companies like MGM, Warner Bros, Pixar, Sony) so that they can meet and review a films progress without having to be in the same place. Also, if a network is installed in a single location, no telecom bandwidth is required. The network will run on an ATM switch. I think that in most cases though, at least some long-distance or even short-haul telecom links will be used. The web site has recently been overhauled, I don't know when you visited it last, but I encourage you to look at the whole thing. The company is working on the marketing stuff and I believe that they are developing ROI materials as well. The uniqueness of the solution is hard to fully appreciate without actually seeing the technology. I have seen it demo'ed twice. The quality in terms of frame rate (no jerky movements, lip movements and audio are synchronized), color and usability are astonishing. Also, the company claims that the network can scale to include many thousands of endpoints with no degradation in performance. This is not multicasting, this is interactive and real-time. The largest number of endpoints I have seen any other solution claim is about 32 (for conferencing). I don't believe that the revenue opportunity is detailed on the web site. It's hard to say what that will be until they actually start signing customers. However, the early adopters with whom they are working are certainly capable of generating very large orders. Remains to be seen if that will happen though, but I think it bears watching. Hope that helps... p.s. TeraMedia will be interoperable with all existing standards. So you will be able to video-conference with a non-TeraMedia platform (H.323) but the quality will be limited to the lowest common denominator. p.p.s. You asked about whether this solution can run on existing PC's. The answer is no. The level of capabilities that they are interested in delivering require a processor that can execute a gigaflop. The AltiVec is such a processor. It is a supercomputer on a chip, that is NOT just Apple marketing hype. In order to compress and decompress multiple data streams in real-time on one processor, it has to be extremely fast. Don't be fooled by MHz comparisons. The AltiVec executes up to 20 instructions per clock-cycle. It is blazingly fast, consumes 4 watts of power and was designed for precisely this type of application. The next generations will be faster and use even less power.