I posted a writeup on Faroese Telecom’s NGN project couple of weeks back. Apparently the telco has deployed VoIP equipment 2 years ago. And although it has found ways to utilize the deployed equipment, the utilization remains minimal. Enterprise VoIP for which the service provider had deployed Broadsoft and Audiocodes products is yet to be launched. Reason? Unresolved technical issues related to Audiocodes media gateways.

To make enterprise VoIP services a reality, the service provider needs to interconnect with the existing PSTN for which it uses the Audiocodes product. Unfortunately the interconnections have not worked as expected. The other problem has been the lack of technical support. The service provider was told that Audiocodes has stopped providing support to the product deployed by Faroese Telecom.

276 million registered users at the end of 4Q07. That represents 12% increase quarter-over-quarter and 61% increase year-over-year

Skype generated $115 million in revenue during 4Q07, representing 76% year-over-year growth, and 18% growth quarter-over-quarter

83% of the revenues from non-US customers

Posted fourth consecutive quarter of profitability

Skype announced a partnership with MySpace and launched the 3 Skypephone in eight countries including the United Kingdom, Italy and Australia

3Q07 update can be found here

Net revenues for the quarter were $540 million up by 8.6% yoy. Sequentially it was up by 3.3%. Communications contributed around $102.6 million. Other segments that contributed include Computing 31%, Consumer 29% and industrial 21%.

The net realizable bookings across segments were at $476.4 million down by 9% due to seasonality, fire at a major customers’ factory and the macro economic concerns of certain customers.

Maxim announced a cut in the R&D budget resulting in $15 million annual benefit towards operating expenses. It reduced investments in RF Wireless and Telecom. In the RF Wireless it announced discontinuation of investing into low-margin handset RF transceivers and in the area of Telecom it has decided to defer additional R&D investments till it sees market acceptance of recently introduced products.

Revenues in Q4 were up by a modest 2% qoq which stood at $323 million. New category products saw revenues grow 9% sequentially in this quarter. In terms of product families, CPLDs declined by a percentile which are used in a number of communication devices. FPGAs however grew by 4%. Among the products Max II CPLDs registered record revenue in this quarter which was up by 8%.

Operating expenses were around 44% of the revenues. Of these $71 million were spent on R&D. Although operating expenses were cut by 3% the R&D spending went up by 6% reflecting better expense management on selling, management and other costs.

Communications segment was the strongest segment in the quarter. It grew by 4% while other segments were pretty flat. Telecom business was down but wireless and networking grew by double digits.

VoIP lines were up 50k during the quarter to 436k (out of 1.5 million Internet subscribers)

The company added over 170k VoIP customers in F2007

VoIP rolled out to over 80% of the company’s footprint

Exploring business VoIP opportunities

Now able to offer the triple play of voice, video and data to approximately 85% of homes passed

Stay tuned during the month of February. SeaMeWe permitting, we will post for you a brand new CEO interview each working day for the remainder of this month!!! That is right. We take you direct to the horse each day. Drink deep from the fountain of Wisdom 2.0

That is what I have been doing myself. Over the last year or so, I have enjoyed interacting with the top bosses at some of the most innovative outfits. Think they are a bunch of accountants managing company books? Think again. While they may need to be calculated and careful about what they say, I have usually found them approachable, open, and intelligent. I hope you find the content from these CEO interviews stimulating.

Most of my interviews take place on phone. Some however will come in form of video clips.

Your TSB applications platform comes with a few web mashups which seems like a progressive move. What traction are you getting for this product?

As of now, we are seeing interest both from Italian and EMEA tier 2 operators and large multinational companies. We are basically positioning our TSB product to both the telco operators and enterprises that want an easy integration of web 2.0 services in the existing fixed mobile and enterprise networks.

What role do you see the social networks playing and could they become your future customers?

Social networking is a fundamental aspect of web 2.0 evolution and is changing the way people interact with the media and communicate. Due to the rise of social networking, media companies have to spend more on technology to handle such interaction. Cisco recently announced it is designing a new kind of operating system called EOS (entertainment operating system) dedicated to on-line communities. Italtel is evolving its next generation networks strategy with robust web 2.0 applications in order to support service providers’ transition to new media. For example, we enable live calls on Second Life and set-top-box interaction for IPTV open environment.

Does your TSB platform support advertising based telephony?

Advertising on free phone calls is effective on the Internet and it is becoming a viable business model also on cellular phones. We are ready to support service providers that decide to adopt this type of models. TSB can mix existing and new services with advertising in a fast and easy way to fit in with the time-to-market requirements of service providers. As an example, through TSB, it is possible to make a selective and personalized introduction of messaging with personalized advertising, either on mobile phones or on TV platforms.

Note: This interview was done a few days before CounterPath acquisition of FirstHand was announced.

What is the update on your commercial deployments?

Through the second half of 2007 we had thousands of seats out there through beta tests and trials. And those tests and trials across multiple OEMs have gone well and there have been related announcements related to general availability of PBX products integrated with our solution towards the end of 2007. We have pipelines built up with all our OEMs to supply our products for large enterprises as well as SMEs as we go into 2008.

There was a lot of interest in FMC at the end of 2006 but during 2007 there has been a lull although there was relatively more action on enterprise FMC side.

On the carrier side there has been a lull due to the fact that wireless carriers are asserting their control. And the way they are doing that is by exploring other FMC options like femtocell. However we did see some offers being launched in 2007 such as T-Mobile UMA based FMC service. On the enterprise side we have seen FMC continue to march forward. The delay there has been due to delays from PBX vendors in integrating this and bringing it to market. Nortel’s CS1000 which is their main PBX platform took six months till the end of 2007 to integrate our product. There are 3000 test cases just from the mobility aspect to validate a product for CS1000 platform. So it is a very strenuous process. We have also seen similar delays with Cisco and Avaya. But all major PBX vendors have announced their products now and hopefully that lull will disappear during 2008.

What percentage of IP PBX extensions do you expect to add FMC capability to, going forward?

Three years out we are looking at estimates of anywhere between 8% to 24% of enterprise PBX extensions being installed as mobile or FMC enabled. For 2008 I expect adoption of under 1% range. And that is a large figure. Enterprise PBX market is a huge market and we are looking at something in the region of 100 million lines per year being shipped worldwide.

Quarterly revenue for IDT was $201.2 million which reflects a decline of 5% qoq. Communications related revenues were approximately 32% of the total revenues which has declined from 34% last quarter. Other segments’ revenue shares were Consumer 19%, and Computing 43%. The rest came from audio segment.

The company continued their engagements with base station manufacturers on the wireless infrastructure side who are working on IDT’s Pre-processing Switch for next generation wireless rollouts.

I am not sure how familiar you are with the name CounterPath, but the mouse just swallowed another little cub from the FMC zoo. Bridgeport Networks is the second FMC vendor acquisition of the week for the relatively lesser known CounterPath that licenses out its softphones to vendors and service providers. The first acquisition of the week was Firsthand Technologies that also falls under the FMC category.

So where would CounterPath be taking its FMC assets? Neither of the two acquired companies offer UMA flavour. So there are not going to be any consumer FMC deployments just yet. The only successful consumer FMC offerings that can be seen out there are UMA based which is a more mature FMC option. The company will have to perhaps wait one more year to see VCC gaining traction. It now has both an enterprise FMC product (Fristhand) and consumer FMC product (Bridgeport) under its belt. I would guess that the company will perhaps target MVNOs that are the ideal VCC target customers.

On the enterprise FMC side it is likely to continue working with OEMs like Nortel, a channel that Firsthand had secured prior to their acquisition. A better strategy would be to target the SOHO service provider breed such as Vocalocity, Phone.com, and Toktumi. GIPS, another softphone vendor like CounterPath that powered the early versions of Skype, is also changing direction somewhat and offering feature server capabilities geared toward the SOHO service provider segment. GIPS’ first customer is Toktumi.

Your offering very much emphasizes connecting IP PBX to digital proprietary phones. What is the logic here? Utilize the existing phone wiring?

Exactly. The existing phone wiring as well as the existing phones as well. We make an adaptor that connects over 24 proprietary digital phones manufactured previously by the likes of Nortel and Avaya. So with our offering you take benefits of VoIP without getting rid of your digital phones and wiring.

Is the adaptor like a multi port ATA ?

It goes a step beyond that. It supports not just analog phones but also digital phones that the PBX manufacturers have made historically. You cannot take a Nortel business phone and plug it into an RJ11. Your proprietary phone at your desktop at work speaks a totally different language. It only knows how to talk to the PBX made by the same manufacturer. Our intellectual property is about how each of those proprietary digital handsets signal when you press ‘Conference’ or ‘Hold’. In our case it sends that signal to our box and our box then relays that signal onward to an IP PBX or a hosted PBX making the features transparently available to proprietary phones.

When we bought our PBX, we bought for about $1000 and the desktop phones cost really peanuts, something around $20 kind of range. I thought that it was counter intuitive to be concerned about handset costs rather than the PBX? But since you tell me that you preserve investment in digital phones, the approach makes sense.

Those phones used to cost upwards of 500 to 1000 dollars. They are really sophisticated phones with 20 buttons on them. And the old ones have better audio technology. They have better quality proprietary chips rather than cheap chips in the present day IP Phones.

Normally you would see IMS being pitched at the local incumbents. Belgacom International Carrier Services (BICS), however, chose IMS blueprint over the softswitch environment even when the cost of IMS implementation was higher. According to BICS, developing applications with softswitch environment would have cost them much more.

The service provider has already realized break even on the investment made in deploying IMS. It reached a break even point in just 10 months after deployment. On the CAPEX it saved 30% and the OPEX was remarkably lessened by 70%. And BICS says IMS stuff is green. Compared to the legacy switches and other infrastructure they saved around 20% in energy consumption which included power consumption of the actual equipment, ventilation and air conditioning of the place where the equipment is located.

Gareth talks about SIP trunking integrated with Microsoft OCS, possible hosted model for OCS, and Arena. Arena is a service offered by Interoute that includes things like switch partitioning and VoIP peering. This is a much shorter version of the video conversation. A full transcript of the interview will follow later this month.

Providing VoIP service since: August 2004. GoIP started with its own developed VoIP system for retail service. Subsequently GoIP International was formed to offer service as an Application Service Provider (August 2005) for consumer VoBB and Mobile VoIP. Retail business sold off to debitel.dk (www.debitel-ip.dk)

Most memorable challenges so far: Providing system deliveries faster to service providers, and avoiding custom development.

Number of service provider customers: 18

You offer services to SMBs mainly, which is now a crowded space? What is your unique selling point there?

Apart from maintaining high call quality, we also deal with organizations that require diversity on their network as well as flexibility in their growth. We are probably the only provider for contingencies in the network to ensure service availability.

You have been through the market downturn having managed Nx Networks. What advice would you have for start up companies in VoIP?

Everyone thinks that VoIP is an easy business to replicate and support. It is probably a lot more difficult operating in an unregulated market. A lot of new companies get into this market and panic and start selling on price alone forgetting to provide quality. So if they are going to enter this market, they better know what they are doing. Otherwise not only will they suffer but their customers will suffer fairly quickly.

You have been involved in acquiring VoIP companies in the past. We see some of the startups merging in the VoIP market. Does merger of startups usually work?

Everybody wants a rollup strategy right now especially since VoIP is a crowded space. Problem is that everyone thinks that they are worth a billion dollars even if they are losing money. They believe they are selling services like MySpace and Google is buying. They don’t realize they will be selling dial tone which has being commoditized for over 100 years now.

Revenues for the 4Q 2007 were $35.3 million, up 5% sequentially. Revenue from multiservice access VoIP processors increased by 13% forming 28% of the quarter’s revenues. High-performance analog products showed an increase of 4% totaling contribution of 30% to the revenues while WAN communications remained flat by contributing 42% towards the overall revenues.

Geographically, Asia-Pacific contributed half of the revenues while Americas earned 39% of the revenues. Europe contributed 11%. Cisco was the only 10% end customer.

You have recently articulated three main areas of focus: your traditional hosted PBX offering, mobility and Voice 2.0. How do these three areas gel together?

We are focussed on re-inventing voice for carriers. Although the contribution of voice in the overall telecom business is diminishing by the day but voice business generated over $1.2 trillion dollars for carriers last year. Re-inventing voice is a huge business opportunity. It can mean several things. It can mean integration of voice with PC based and mobile based applications. It can also mean making voice services available to the user anytime and anywhere he is.

From operator perspective they want certain common capabilities in their network. Take BT21CN project which is a reference blueprint in the industry. There is a push to minimize complexity of managing tasks. There are some common capabilities across networks and across services. That is what voice will have to adapt to. Sylantro’s multiplay application server will become a key application building block of tomorrow’s network architecture. It will serve as that common capability platform for different services. So in short our strategy has evolved to serve two areas: common capability multiplay platform and re-inventing voice.

You mentioned multiplay platform. There is one such deployment at Swisscom where the same feature server is being used for fixed as well as mobile voice. Right?

Yes. Swisscom deployed Sylantro platform when Siemens was their systems integrator. We deployed fixed line voice for consumer voice. At the beginning of this year Swisscom decided to merge and evolve their network to an IMS architecture to combine and converge two network architectures. And Ericsson took the lead and worked directly with us as supplier for the IMS architecture. Swisscom asked Ericsson to use the same feature server that was used in fixed voice. Today Sylantro platform serves the common voice capability in Swisscom network.

Any other operators out there that have deployed your platform for both fixed as well as mobile voice services?

Yes. Wateen in Pakistan is the largest IMS based VoIP over Wimax network. The service enables mobile Wimax at the access side serving both consumers as well as businesses. They launched the services in November and it has been a big success. Similarly we have a project going on with Korea Telecom. That is also a Wimax project where business customers are integrating VoIP with their Microsoft OCS platform. There are two large trials – one of them in the US – that also use our platform as a multiplay platform. These deployments really challenge the traditional FMC set up. We see a much more creative convergence than the traditional definition of FMC.

• 4.102 million VoBB subscribers in France as of end 4Q07. Added 617k subscribers during 4Q07.
• VoBB subscriber base represents about 36% ADSL customer accounts.
• 677k VoBB customers outside France (mainly in the UK). Poland: 132k VoIP subscribers, other European countries 544k.
• IP TV Subscribers: France 1,149k, Poland 40k, other European countries 54k.
• Orange Unik FMC subscribers: 573k users as of October 2007, up from 468k users at end 3Q07.

If you are a developer who is particular about the underlying chip for your application, your demands will by and large boil down to Processing power the chip has, the Space it occupies, and the Power it consumes.

You need the processing power and speed so that you are able to add on multiple applications which can perform various functions. But adding more applications result in two things. It demands more power consumption and since the base band will be capable of giving output in various forms and for many applications, there will be more and more peripherals added to it. So demanding more processing power effectively results in increase of power consumption and the size.

What sort of VoD content do you see out there in the IPTV market? Is it the same VoD content that you find with other forms of delivery?

The most popular content for VoD is recorded TV. You record the TV channels and you make that available on demand. The second most popular VoD content is movies. Local content is also becoming a big factor especially in IPTV services. These are the three major VoD content forms that are popular with IPTV users but none of these three are IPTV specific only.

What in your experience are VoD demand trends in emerging versus established markets?

In the US, the initial IPTV deployments do not necessarily have a VOD component. They rather focus on linear TV. You see more openness in deploying VoD in emerging markets versus the US where the traditional telcos have been more cautious about the introduction of new services. They typically go for linear TV first and then add some VoD and only later the real promise of IPTV services which is the new services over TV.

Are you involved in any hybrid offerings such as Verizon that is offering the VoD component over IP and the rest over legacy network?

We have a few customers in Israel, Far East and Europe, where the channels are being delivered over the air – either over satellite or over digital terrestrial – and the interactive on-demand content is delivered over IP connection through DSL or FTTH. We definitely see the hybrid approach as a trend.

When you evaluate central office of a telco for possible IPTV deployment, what would you ideally want to see there? What are the ingredients of a suitable telephone plant for an IPTV deployment?

Revenues for the 4Q 2007 touched $2.96 billion up by around $260 million compared to 3Q 2007. Total shipments for the period were 2.36 million units (all 8” wafers now). Shipments for the quarter were up by around 5.9%. In 3Q 2007, the revenues had increased by 17% and the shipments by 20% which in this quarter has considerably gone down to 5.5% and 5.9% respectively. 68% of the quarter’s production was consumed by fabless/system customers while IDMs consumed the remaining 32%.

Geographically North America accounted for majority of the sales: 79%. Asia followed with 11% of total. Europe accounted for 8% of the sales. Japan contributed 2% of the sales.

Communications segment continues to be major sales contributor segment for TSMC. In 4Q 2007, communications chip contribution remained flat at 42% of the sales. Other segments contributing to the quarterly revenues were Computer 35% (up 3%), Consumer 15% (down 2%), Memory 3% (down 2%) and others 5% (up 1%).

In terms of chip technology, 90nm chips were the maximum revenue contributors for the quarter by fetching 29% of the revenues. In the quarter TSMC also announced shipping of one-millionth 12” 90nm wafer in less than 5 years. That was followed closely by 0.15/0.18um chips contributing 27%. The 0.11/0.13um, 0.25/0.35um, 65nm and 0.50um+ contributed 20%, 10%, 10% and 4% in that order. By chip technology 90nm chips saw some increase while others had a dip or remained flatter.

Alastair discusses factors determining FMC growth and reasons for slow FMC uptake so far. Tango Networks claims that its hybrid FMC architecture is a win-win solution for both the PBX vendors and carriers thus avoiding possible cannibalization of revenues of these two important players in the FMC game. You can hear more from Tango Networks during our upcoming webinar on FMC security.

• Added over 2.5 million VoIP subscribers during the Year 2007 - an increase of 61%.

• Comcast ended 2007 with a total of 4.4 million VoIP subscribers.

• VoIP service now marketed to 42 million homes representing 86% of Comcast's footprint

• 4.4 million VoIP base represents 11% penetration of the total Comcast addressable market

• Phone revenue increased 85% to $1.8 billion due to significant growth in VoIP subscribers, offset by a $229 million, or 50% decline in circuit-switched phone revenues as Comcast transitions to marketing only VoIP in most areas.

• Entering 2008, Comcast has fewer than 200,000 circuit-switched customers, with the winding down of that business expected to be completed by the end of 2008.

• Comcast ended 2007 with 13.2 million high-speed Internet subscribers, or 27% penetration of homes passed.

Atheros posted a revenue of $114.3 million in 4Q’07 which was up by 8% reported in the last quarter. Income for the period was $21.4 million. The networking chipsets fetched revenues of 52% while PC OEM got in 43% and the rest of 5% was brought in through consumer chipsets.

Major growth was seen in core wireless LAN segment, particularly the 11g solutions that contributed primarily to the overall growth in this quarter. Wireless LAN chipsets contribution by types was 11a/g (16%), 11g (67%) and 11n (17%) of the total wireless LAN revenues.

Hon Hai Precision Industry was the only 10% customer for the period.

By the end of the quarter, Atheros had shipped over 20 million Ethernet ports to around 20 different customers on a cumulative basis. In 2006 they had just 1 customer for this product line.

Why provide a service rather than license the platform to service providers?

Our background is actually building platforms for carriers. Our team built platforms, and enhanced applications on top of those platforms, for AT&T, Sprint, Pacbell, GTE and several wireline and wireless carriers. What we experienced is that innovation always came from carriers and at their own initiative. Your fantastic features and APIs don’t mean much to them. At the end of the day, what we found was that carriers did not have the infrastructure or the business model to allow their networks to be opened for third party developers. With Ribbit, we have switched gears a little. We decided instead to open the platform to developers. And we are not going to wait for two years to bring new features to consumers that could be offered yesterday.

There are service bureaus out there like Tellme that offer something similar to enterprises. Would you regard them as similar to Ribbit?

We have seen service bureaus out there that serve large enterprises providing APIs that are telco oriented. These companies use SIP APIs but there are very few developers out there that really understand SIP stacks. We are trying to make those APIs one level easier. We provide Flash APIs so that the developer does not have to necessarily know the bits and bytes of telephony protocols.

Is time-to-market the main reason why you go direct to the consumer or is it the lack of compatible carrier infrastructure?

The issue is the time it takes carriers to incorporate things into their network. The issue I think is much less how they open up their APIs to the core elements of their network. You see some of the progressive carriers attempting to do this. BT has their web21c program. But even so it is coming on slow. And again if you look at the interfaces that are being provided it is being provided with the concept of deriving a telephony application, adding telephony to an existing workflow type of application for example.

This model can easily be replicated by social networking sites like Facebook. Are we looking at some sort of competition from those guys in the future? They could aggregate these Voice 2.0 applications and offer a services bundle.

It is possible. But telephony is a different game than providing various types of web services. You have to have telephony experience. There is also the billing issue. In communications we have applications that we need to bill for and we are going to see that for a long time to come. Social networking sites generate revenues based on advertising. It is very hard to subsidise phone calls through advertisements today. Social networking sites have a different mentality. We help developers monetize the application. Facebook type sites are not concerned about monetization aspect of the applications.

• Added 56,000 net subscriber lines during the quarter to end the year at nearly 2.6 million lines.
• Lost 230k customers during the quarter. Average monthly customer churn during the quarter stood at 3%
• Marketing cost of customer acquisition: $223 per customer. Total marketing costs during the quarter: $63 million.
• IP litigation costs during the quarter were $1.3m compared to $133m in the previous quarter
• 4Q07 revenues $216 million, 2% increase q-o-q and 19% increase y-o-y. Net loss $9 million
• ARPU: $28.19 down from $28.24 in 3Q07.
• Current cash and marketable securities and restricted cash at quarter end was $190 million.

Providing IPTV service since: August 2007 (in Norway)

List of challenges: Integration between ERP systems and IPTV provisioning systems; STB problems; VOD software bugs; Content owners that do not allow the company to sell channels one at the time (only bundled in packages); difficulties raising enough capital; content distributors that own distribution infrastructure.

Number of IPTV subscribers: not disclosed

IPTV equipment sourced from:
Set-top box: Motorola 1510 and 1910.
Middleware: Dreampark
Encoder: Latens
VOD Server: BitBand

I am not yet clear if stress leads to a creative outlet or creative exhaustion. For me the months of Jan and Feb are always stressful in terms of workload. And I have noticed that I am not able to generate any micky mouse ideas these days that I could feel excited about. So although I have only one data point here to make an inference, I would say that extra workload is not good for creativity. There is another thing that is not good for creativity. I am officially (and regretfully) a smoker but my average is like 5 cigarettes per month. When I have a smoking streak (like in Jan and Feb), I lose it all. I mean creativity :) The fags just do not work for me. They make me dumb. Not sure how Freud and Einstein managed!

Anyway, I will blog about a few ideas that my smoke infested grey matter came up with. One is a mobile VoIP aggregator. Since you are going to see dozens of mobile VoIP offerings during 2008, why not develop a Least Cost Routing softclient that interfaces with various mobile VoIP offerings out there. Making cheap calls has to be one of the main motivations for using Mobile VoIP. And a single provider can never ensure best rates for all long distance routes. The LCR softclient can also similarly aggregate the mobile callback offerings out there if the APIs are available from companies like mig33 and Jajah. And while you are at it, you can make this a flash phone client so that we do not have to download anything.