Pipeline Profile: Gregg Grigaitis

SCTE Member Since 2003 Title:Vice President, Customer, Broadbus Broadband Background:Grigaitis previously served as general manager for the near Video on Demand (NVOD) and GuestServe systems at SeaChange International. Prior to that, he was director of IT for The Cable Advertising Network, and held several technology and business positions with SBC. What’s working on the VOD front? Open standards, for one. Clearly, Time Warner was onto something with its Integrated Services Architecture (ISA), as is Comcast with its Next Generation on Demand (NGOD) architecture, and the broader Next Generation Network Architecture (NGNA) of several operators, with open standards and best-in-class technology. Second, purpose-built technology. Operators are steadily improving VOD performance and have derived operational ease of use and simplicity in systems deployments. How does Broadbus fit into this landscape? Broadbus has arrived at a time when, in order to rollout new services, operators are demanding a level of reliability that they are simply not achieving with their existing disk-based VOD vendors. Open standards allow operators the flexibility to easily deploy a highly reliable architecture such as Broadbus. We’re not only easing the current VOD deployments, but are allowing the foundation that will reduce the time to market for new products, such as Television on Demand, which equates to more revenue sooner for the operators. 2005 will be a watershed year in the deployment of open standards VOD pumps. Is content no longer a sticking point? There has been a clear and measurable up-tick in VOD usage as a function of content availability. The last two years have seen a tremendous increase in both quantity and quality; usage statistics are tracking very favorably. What about the question of centralization vs. decentralization? I’d suggest that centralization/decentralization is not really a debate. It’s a configuration and economics issue, weighing transport costs, VOD hardware costs and configuration preferences. Different operators have different approaches, and VOD vendors must provide a solution that can be precisely configured to fully leverage the operators’ preferences. So what would a new architecture look like? The cutting-edge deployment of choice that Broadbus is seeing across multiple Tier 1 MSOs is a completely redundant bi-directional ring, with 2 super-headends—each headend containing a VOD pump that equally provides reachability to every edge device/set-top in the network. Given that Broadbus is able to deliver 38,000 MPEG2 3.75 Mbps SD streams from a single rack, we’re very well suited to this architecture. Additionally, new deployments are completely breaking apart the "storage and streaming" linkage of the disk-based pumps. Broadbus has deployments from under 1,000 streams to over 12,000 streams, with storage ranging from 1,800 hours up to 5,000 hours, and several variations in between. So we’re equally well suited for the smaller deployments. What’s not working? That would be the opposite of my first answer: disk-based closed systems. The disadvantages of the old VOD deployments, where the server vendor "handled everything," with a box or cluster of boxes repurposed for video delivery, are becoming more and more apparent to the operators, as evidenced by their migration to open standards and best-in-class technology. Operators are replacing systems that are a mere 18 to 24 months old as they have already run out of gas. What else, more specifically? The operators have 2 major challenges: functionality and reliability. The tightly coupled disk-based VOD pump just isn’t working. When you use a hard disk as a streaming device in order to grow streams, you have to increase hard disks to provide the requisite bandwidth, often requiring several hundred hard disks. Broadbus routinely deploys with only 14 or 28 hard disks, dramatically improving reliability. Additionally, with the clear direction toward real-time content being delivered with VOD servers and ingest demands rapidly increasing, conventional pumps simply do not have the horsepower and I/O bandwidth to keep up. As DRAM pumps are deployed, expect to see a marked increase in new functionality and services, primarily around non-linear television, such as nDVR and Television on Demand. Seems like this sector is heating up? Is that so? The sector is clearly heating up. Improved content is driving usage rates; usage rates are driving the need for more streams. What’s with Broadbus and the music thing? Do you play an instrument? Broadbus is "rocking" the cable industry in more ways than one…. To get the job, I said I played the drums. They’ve called my bluff, so I better take some lessons quickly. Stayed tuned for more from Broadbus at NCTA. [Editor’s Note: For more on the topic of "VOD: What’s Working, What’s Not," see the current issue of Communications Technology, or check out the CT archives at http://www.ct-magazine.com.]