In his new post as vice president and chief scientist at Broadcom Corp., cable veteran Rich Prodan is busy working on the next generation of cable products, from silicon up.
Broadband Week senior editor Karen Brown recently talked to Prodan about his recent move from Terayon Communication Systems Inc. to the world's biggest silicon provider and his views on what the future of cable technology holds. An edited transcript follows.
BW: Your move from Terayon to Broadcom took some people by surprise. Can you talk a little about that decision? What attracted you to Broadcom, aside from its leading position as a chip supplier?
I've always been in [research and development] for my entire career, and basically, when I was CTO of [Cable Television Laboratories Inc.], interacted with a lot of different vendors and brought them into various projects within CableLabs.
After the establishment of DOCSIS [Data Over Cable Service Interface Specification], then the appeal of S-CDMA [synchronous code-division multiple access] — after it was actually built, implemented and millions of these things were operating worldwide — in fact, Terayon became the major proprietary modem supplier behind Motorola Inc. It sort of attracted me to say, 'Well, I think the timing is right now for this technology to be introduced into the industry,' and after talking with the Rakib brothers [Terayon CEO and co-founder Zaki Rakib and president and CTO Schlomo Rakib] about this — really saying that they should be looking at a standards-based product that is really going to be accepted in the industry, which initially wasn't their desire.
But they said, 'Well, OK, we agree at this point. We'd like you to come on board and help to do this.' So that was my entry into the vendor side of the business joining Terayon.
And really, the push for getting DOCSIS 2.0 evaluated, and seriously looked at, and finally adopted, which was really my major contribution while I was at Terayon. And with the acceptance of DOCSIS 2.0, it really is now at a point where it is execution by various companies to get that technology out into the marketplace.
So looking around there, my job is sort of done, and I have always been kind of a very broad-based guy — I came from the [consumer-electronics] industry, 12 years working for Philips [Consumer Electronics Co.] basically with a [high-definition television] and video signal-processing background. So Broadcom has been always a favorite company from a technology point of view being so broad-based, and it's sort of a natural match to my background and talents in video, in data, in storage.
And so after my job at Terayon was more or less complete I looked at the opportunity there. They have a huge base of talent — over 300 Ph. D.'s, and now 301 with me joining. It's a huge opportunity to get involved in a very broad-based business, and all of the new and exciting opportunities that are just on the horizon for the cable industry, that in the next couple of years will be changing the face of services as cable operators know it today. I really wanted to be sort of in the thick of that, more of these new and advanced services — the new millennium for the cable operator.
So this is the natural fit for me to move at this time, I thought. And that's where I wound up.
BW: Much of the development buzz surrounding cable modems these days seems to be focused on DOCSIS 2.0, and there are now several products — including those from Broadcom — that are capable of supporting the standard. When do you think the cable industry will be ready to switch on the 2.0 capability? What are the MSOs telling you?
Well, the actual data of turning on 2.0 is probably a little bit fuzzy at this point, but I think what I have been hearing in a lot of my discussions with the MSOs is that if they have silicon in products that they will be buying beginning early next year that are 1.1-certified in terms of cable modems, but are 2.0-based, that a lot of them will require their modem purchases to be 2.0 based, in anticipation of when they will swap out [cable-modem termination systems] to support the 2.0 functionality.
Because really, all of the 1.1 [quality of service] is part of DOCSIS 2.0. So it is fairly easy for them, because they are in control of the headend when they want to switch. It really is an upgrade — a swap-over of the CMTS — which is a sort of central decision they have to make, rather than the individual subscribers in their homes when they have all of these modems available, and modems they are buying at retail.
And that is a much harder thing to control, so they're concentrating on getting the DOCSIS 2.0 silicon in 1.1-certified modems added to the market, and then deciding when to enable the 2.0 features as a centralized headend upgrade in the CMTS as a natural progression.
BW: Do you have any guesses as to when that will happen?
You know, it depends on how wildly successful they are in deploying new services that are more than best effort. If they are going to deploy a lot of tiered services — where they will have to give service-level agreements for the gold/silver/bronze sort of strategy for different data tiers and charge accordingly — to be able to control the amount of bandwidth you give to people who want to run businesses with servers over your network, the business rather than the residential customer, these will be using a lot more upstream bandwidth.
Because with cable operators, in my history with the industry, it has always been a question of need. They don't spend any money until that investment has a very near-term return for them. When the business starts to pick up, in terms of the new services that they are offering, and if they are successful, then I think they will have this capacity issue and I think it will be a natural for them to turn on 2.0 without requiring upgrades, which is a very happy story that they can tell to Wall Street.
BW: As DOCSIS 2.0 nears its first wave of certification testing, what's the next step in cable-modem technology? Where is silicon development headed?
I won't get into where the next generation of development in silicon is headed. I think that is a whole other discussion that most people aren't ready for right now. But I think that with the establishment of 2.0 and the beginning of embedding DOCSIS into different devices to control power-meter security systems, lighting, as well as entertainment products and whatever in this embedded DOCSIS or eDOCSIS program that they are starting.
They have residential gateways now. People are running small [local-area networks] in their homes and their offices, so the CableHome project is taking DOCSIS as a base and then bringing it out into home networking. So I think just using DOCSIS as a single computer connection to a best-effort service is going to branch out into these LAN-based QoS and embedded applications in different appliances in the home — embedding DOCSIS in set-tops to overcome some of the difficulties with scaling interactive services and bringing voice, video and data together in an integrated fashion, rather than independent silos, as they are today.
That's really where I think the next step is for where DOCSIS is going. And once we have 2.0 cable-modem technology certified — and the CMTSs qualified — then I think these natural extensions will be shortly following for a whole lot of exciting products for the industry.
BW: Now that you are at Broadcom, you also get to play with digital set-top box silicon. Where is chip development headed on that front? All of the new services that could potentially be pumped into set-top boxes — i.e. video-on-demand, personal video recording, HD and one day, interactive services — what are they doing to silicon and memory requirements?
As you can guess, a lot of the services now that people want to deliver do have to be integrated in a package that is plug-and-play and easy to use from a subscriber perspective. So what the silicon is going to do — and a company like Broadcom, which is uniquely positioned to do this — is there are a lot of groups that have core technology, that we can build complete systems on silicon for a set-top box, where you can support a lot of the services you are mentioning here within a single-chip solution. It's a big chip. It's a complete system.
But there are very few companies that have all of the technology in-house that can put it together and do this. Embedding a DOCSIS 2.0 modem with the out-of-band to support legacy at the same time; multiple tuner; silicon-based tuner technology to support to parallel data and video transmission, as well as voice service to make actual gateways with home networking technology — which is also a major focus in Broadcom. It can all be brought in to a systems approach to bring down the cost of offering these new services.
BW: What are your biggest priorities these days in terms of cable-related products?
I guess it is the global adoption of DOCSIS 2.0 technology. I have been working with my colleagues in both the European community and with the Japanese government and industry to basically come up with an international standard within the [International Telecommunications Union] for DOCSIS 2.0 that will address all regions in the world with any regional differences required for their markets. We've successfully completed a consented draft that covers all regions of the world.
In terms of video, it's been what you touched on earlier — it's the system-on-silicon approach and the integration of HD and digital TV services for the OpenCable technology that is just starting to be worked out and introduced, hopefully some time next year, with the agreements reached between the [Federal Communications Commission], the cable operators, the technology providers and the content owners. These four people are playing all the different roles in establishing what this open digital-TV environment is going to be: Open in the sense that it can be available at retail — you can carry signals from broadcasters, as well as cable providers and satellite, in a ubiquitous way — but you still maintain control that the content owner is not afraid to release their intellectual property in a digital format for fear of copying and loss of management of their rights. So all of that is just coming together.
BW: Is there anything that you want to add, anything that you think is particularly important for the cable industry going forward?
The major thing is to train their people, their staff in how to deploy this technology successfully, flawlessly. It's something that no matter how cool a technology is, people buy services — they don't buy technology. So we have to help the industry to really roll out a plug-and-play set of services without hiccups. And if we can do that, I think nobody can touch cable as the ultimate broadband end provider of services to both the home and to businesses.
Nobody else has the two-way bandwidth. Nobody else has the breadth of content. Nobody else has the understanding of the customer relationship. This is something that, if we integrate this and provide these packages, I think nobody can touch cable as being the ultimate broadband provider.