Thursday, September 4, 2008
Interview with David Wong, D2 Technologies
Story by Benjamin F. Kuo
Our interview this morning is with David Wong, Chairman and CEO of Santa Barbara-based D2 Technologies (www.d2tech.com), a firm which has a very interesting history of developing Voice-over-IP technology. Wong has a long and noted career in the area -- in the 1970's he was on the ARPA team that completed the first phone call over ARPAnet, the predecessor to today's Internet--and he also co-founded Spectron Microsystems, which developed the first DSP (digital signal processor) and was acquired by Texas Instruments. He later started and sold D2 Technologies in its initial incarnation for a very healthy sum to Virata (which became GlobesanVirata, and later Conexant). We sat down and spoke with David about his company, his views on the future of VOIP and telecom, industry efforts like Google's Android, and more last week.
You've got an interesting history--D2 was acquired by GlobeSpanVirata, but a few years ago you revived D2 as an independent firm. Can you telll us the story?
David Wong: Back in 1993, when we started, DSP was a relatively new technology. I had actually worked very closely with TI on the development of the first TI DSP in the 1970's, as a graduate student working on voice compression. They had developed DSP for secure military communications, and in the 1980's decided to introduce DSP in the commercial arena. The area they had a lot of traction with was voice modems, and then in the 1980's voice messaging and PBX type applications. I started the company in 1993--back then, Voice over IP wasn't practical, the Internet was not built out, nor was broadband. So we were primarily working on DSP technology, which was ultimately embedded into every major system in the world. We created DSP that went into conferencing systems at Cisco, created DSP switches for Lucent, Ericcson, etc. etc.
In the late 90's, the Internet boom started, and broadband began coming on line from people like GlobespanVirata. We saw lots of interest in technology for voice over DSL, voice over cable, voice over Frame Relay -- it wasn't even called voice over IP in those days--it was called voice over packet. We created the first voice over packet, voice over Frame Relay, and voice over ATM, and consequently voice over DSL solutions, which is the reason GlobespanVirata acquired us. There was a real rush for the DSL and cable modem and chip companies to acquire voice technology. TI acquired Telegy, to secure the voice over cable business, Broadcom acquired another company to compete in the cable/DSL area, and we were kind of the last standalone of the top three voice technology companies. We had a very nice exit in late 1999/early 2000.
After joining GlobesanVirata. they turned D2 into their corporate technology center for voice-over-DSL, and ultimately Voice-over-IP. We acquired lots of networking technology, which really created the foundation of our product, with a complete VoIP engine, DSP, codec, echo cancellation, including protocols as well as application software.
So how did you end up restarting D2 as an independent firm--was it a spinout?
David Wong: It was kind of a spinout. After I sold the company, I stayed, as the firm changed from working on DSL to Voice-over-IP, to complete my stock vesting. After I left, having sold the company, I had a non-compete, and dabbled in other investments and also tried my luck in the film industry. In the Internet nuclear winter in 2002/2003, as you know, all of the revenues for companies went off a cliff. Profitability became the key to the survival of all technology companies--including GlobespanVirata. They were in relatively good shape, having just raised something like $400 to $500M before the crash, but they decided that VoIP was dead in the water. The broadband industry--the CLEC industry--pretty much imploded. All of those broadband businesses went back into the control and hands of the incument LECs, such as Verizon and ATT, who were not in a hurry to go to VOIP. VOIP was hot for the competitive LECs to compete, and it was a very compelling story--and they would have wanted it, if the financial markets hadn't turned against them. So, since they all imploded, GlobespanVirata decided that VOIP wouldn't see daylight for a few years. There was no VOIP revenue, and incumbent LECs didn't want to see VOIP. They decided to mothball the technology, and relocate the team from Santa Barbara to their headquarters in New Jersey. As you can imagine, that went over really well. Those in Santa Barbara said they didn't want to go, so the firm decided there was no point of having the technology without the team, because it would be useless in two years.
However, they had some marquee customers--Sonus, Cisco, Lucent, and Siemens--customers who were outraged on hearing about the shutdown. They said--when you first acquired D2, you said you would maintain that software, and how can you do that without a team? GlobespanVirata got in some serious talks with the OEMS, who told them--we'll never buy a DSL chip from you if you do this to VOIP. So they did a deal. They told us we'll give you back the technology and the customers, if you will basically agree to sustain the code for a couple of years. The deal was struck, and we spun out that way. The group actually came back to me--I had left already--and told me--we're just a bunch of engineers, would you mind coming back to run the show? I cam back at first a day a week, but before I knew it it was full time. Then, when VoIP and Vonage was formed in 2003, broadband really started to take off. Now, in 2007/2008, VoIP there are countries like Japan where VoIP will be 100 percent of their telephony technology by the end of next year. They are actually shutting down their old phone network. British Telecom said it will be shutting down its old phone systems, because it's so much more expensive to run--especially when people are not building new equipment, everyone is using twenty year old designs, and twenty-year old semiconductors. It's not a sustainable technology anymore. I think in the next three to five years, you will see most telcos announcing they have, or plan to quickly shut down their old legacy telephony systems, and go all to VoIP.
Explain what your company does now, how your software fits into today's market?
David Wong: Our software fits in a nice way. We started developing DSP technology, which was already used in the old legacy phone systems for some new features, such as interactive voice response, conferencing, and voice messaging. It was still not packet based, but that technology was needed for voice over packet. DSP companies then turned that into voice over packet. We then focused on the IP side, dropping Frame Relay and DSL, and that technology has been deployed and proven over the last decade or more. You need all of that technology as people transition to voice over IP. For example, in my house, I have a cable modem, and VOIP, where I plug a phone into it. That turns your cable modem into a small central office. It does everything the central office did in that big phone building, on a Class 5 switch, where our software used to be. That's now shrunk from a 5000 point switch to a 1 point version, on the end point device. That's where we are. What we've added is the SIP protocol, SRTP, firewall traversal, and lots of other new technologies which have been necessary to integrate legacy voice technology into an IP broadband network. We basically took a very solid base of voice software, added networking technology as part of a router, into the product. We're now taking it one more step.
We're still in a transition, where we--as a user group--have been trained since we were one year old to the Ma Bell concept of a phone. But if you look at a cell pone, that's the next concept--it's got a screen, widgets, a GUI, and all of these other features. IM, text, that's the new paradigm. If you look at the Ma Bell pone, which you now plug into a cable modem or DSL to get voice technology, I think you'll see it's an oddity like a buggy and wagon. It's like seeing a buggy and wagon on the freeway. People now text more than they use the cell phone, but you can't text with that Ma Bell phone. What we see, is the entire user experience, and user features and functions, are going to transition from voice and telephony, to people-to-people communications. That includes messaging, IM, text, presence, voice, conferenceing, etc. etc. We've now gone beyond just providing the protocol and voice stack, to include UI software, in our new product, mCue, geared at the mobile user experience.
So who is the customer for your products now?
David Wong: The mCUE product is really targeted at the enterprise space. Companies spend millions of dollars on their PBX, and for those phones sitting on everyone's desk. In the last five years, the phone on the desk has had a drop of 90 percent plus. When I'm in the office, I hardly ever make a call or pick up my phone. I'm on my mobile, or on my PC. My PC, which is next to the phone, has a soft phone, and Yahoo Messenger and Google Talk. Why do I need a PBX? On the other hand, once you walk away from your desk or walk away from your laptop, you lose all those capabilities. With Verizon or other providers, they have a different phone, and a different UI. What we're trying to provide is software on those phones to deploy by enterprises, allowing them to add or embed the application on all of their employee phones. Now, your mobile phone will become an office communications tool, whether you are in the office or on the road.
With the iPhone, Winmobile, and Android, the phone is now going to be an open platform, and you will be able to embed new applications like you do on a PC. Some PCs are used for CAD, some as an IDE for software development, other for accounting--and most now have instant messaging and voice on them as well. Communication is going to be a very fundamental application on all mobile devices, whether that is a GPS device, a game device, a PMP--once you have a network, people will want to have communications on it--and it will be people-to-people communications, rather than people-to-website.
The other way we get to market is through the service providers, who also have a problem. Their problem is that right now, with enterprises, if you are an employee of company X and make a phone call, they can show you who called, who called you, all in that PBX. Now, they've lost touch. You might have a sales person with a mobile phone, and when they walk out the door they take their entire account list and history with them--they have no clue for those last 5 years what has happened with those customers. Companies are going to the mobile guys, saying--we need a product from you. With their mobile users, they don't know what they're doing, and can't control it. They could be making terrorist calls, and they have no idea. They could be calling their competitors, and they don't know. They can't manage them, they can't do cost controls--and what they'd really like to have is a mobile service than an enterprise can manage, which can keep records, and control. As we move into 3G and 3 1/2G phones, telecommunications providers are starting to be relegated to the pipe. That's a commodity, where you price goes to hell, and your margins go to zero. They want to add value, and the most willing payers are not those people paying $19.95 a month, the consumers--but those paying $150-$200 a month per user, the enterprise. They want to capture more value, and provide more services to enterprises. So it could be to services or directly to users, the business model isn't yet clear.
You mention you're involved with Google's Android, what's your assessment of that effort and other similar efforts for open phone standards and operating systems, and how are they are going?
David Wong: Actual market adoption hasn't started. There are lots of products still in development and being worked out. There has been some early adoption, such as IBM, with fairly good adoption of their Sametime product. However, that's been on desktop, not mobile. Talking to Tier 1 carriers, they all have projects to take enterprises and IP mobile. They haven't announced anything yet, but we're seeing early instances of where they are going. Mobile carriers will be offering wireline service--you'll be able to buy a thin card from TMobile or others, and for $9.95 you can plug that into a phone box. There's also the femto stuff. Those are all incremental steps towards fixed mobile convergence. It's pretty clear--PBX are all going towards IP, home cable or ponei s going IP, and people with an IP PBX and presence can be anywhere in the world, as if they were at an extension in the office. You see fixed line wanting to move into mobile phones. This is where it is going to happen, from both sides, depending on which carrier or OEM you are talking about. It will be a very exciting next 5-10 years.
You've had quiet a bit of experience from the telecom side of things, what's your view on how the telecoms versus the Internet companies will handle this convergence?
David Wong: The telecom's entree to VoIP is to hide it. The cable company also hides it, and doesn't tell customers it's offering VoIP. Instead, they are offering better packages, and features, such as being able to be reached remotely--all of that because it's VoIP. They are offering these packages but don't market it as VoIP. They're selling it much more as an extension and enhancement of their telco products--lower prices, and features. The Internet guys think that this should be free. Email is free, why not voice? But, because they don't have the consumer experience, if you use Google Talk, or Yahoo, you'll find they have nice features like presence and a GUI, but in terms of quality and service they haven't quite gotten it. Ultimately, I do think voice will be practically free, it's just a matter of the telco's figuring out business models and how to make money on other things--whether that's messaging, maybe call management services, or other kinds of enhanced services.