RTC Magazine

RTC: The telecommunications and networking market/industry is undergoing some roiling changes at the present time. As a longtime participant in this industry, what is RadiSys’ view of where things are going in certain aspects of the industry? This is a big question, so maybe we can break it down into parts: One would be the issue of convergence—the mass of data, including voice and video—that service providers want to get to users. How do you see the industry addressing this? Will it be cable, fiber optic, satellite or some combination of these?

Grout: In both wireline and wireless environments the phenomenon is “triple play” or the packaging and delivery of voice, video and data services. These services require huge amounts of bandwidth and high-level security, both of which Ethernet supplies in a cost-effective, scalable manner. IP packet technology will play a key role in allowing various IP-based services such as VoIP, VOD and IPTV to be quickly and easily deployed.

Service providers, especially in the U.S. and Europe, are now starting to invest in infrastructure after an extended slow down in capital spending. Their strategies are sharply focused and investments are being made in products and technologies that allow them to reduce costs and increase revenues and profits.

In wireless, significant investments are now being made worldwide by service providers to upgrade their existing 2 and 2.5 systems to 3G. 3G systems allow the seamless integration of voice, video and data into a unified, high-bandwidth, cell-based infrastructure that accommodates mobile or fixed use. The shift to 3G means service providers can deliver next-generation integrated voice, wireless data and IP-based multimedia applications to more customers and ultimately realize more revenue at a lower cost. As the number of users, and services they access from laptops, phones, PDAs, instant messaging devices, etc. continues to increase, the 3G transition will likely accelerate. Other technologies such as WiMax will play a role in providing broadband access in fixed wireless applications and may end up taking market share from cable, twisted pair and WiFi (for local premise distribution).

Scott Grout, President and CEO, Radisys Corp.

Delivering integrated wireline services to private residences is currently handled via existing media such as twisted pair, coax and power line. Both carriers and cable operators are achieving remarkable penetration of high bandwidth services over the pre-existing cable and twisted pair venues. This media must be fully exploited before it will make economic sense to retrofit, or overlay, a new pathway to existing homes. If a viable third media is required, utility companies can increasingly enter the market with relatively low capital expenditures using existing power distribution infrastructure to offer consumers even more choice. Unfortunately, the most expensive network upgrade is retrofitting the “last mile”. Because of this high cost, the bar for increasing revenues and profits, and lowering costs, is set fairly high. Service providers have critical ROI (return on investment) targets to meet before approving the capital required for massive last-mile upgrades. Once the existing venues of twisted pair, coax and power line are exhausted, new fixed wireless broadcast technologies such as WiMax may be more economical to deploy than a mass physical upgrade of the existing last-mile infrastructure.

In business applications, there is already deep penetration of fiber to the enterprise and clearly this practice will continue. Depending on the economics, WiMax may also play an expanded role in delivering fixed broadcast broadband services to the enterprise.

For telecommunication equipment manufacturers (TEMS), who are our immediate customers, the delivery of voice, video and data and the shift to 3G are forcing them to look for more economical methods of developing next-generation products, faster. The emergence of industry standards, like AdvancedTCA, and the drive to better leverage internal resources have made outsourcing both at a platform and supply chain management level an attractive option. Incorporating AdvancedTCA-compliant products into their designs allows TEMs to spend critical resources on developing feature-rich applications and services rather than building the next compute or switch blade for a proprietary platform. To maximize budgets and meet time-to-market requirements, TEMs will need to rely on embedded vendors to handle the integration and validation of various components, from a variety of suppliers and trust that they will deliver proven, tested, application-ready platforms. RadiSys is at the forefront of leading the industry with this transition and addressing the TEM needs.

RTC: Then there is the wireless phenomenon. Wireless seems to be growing faster in some other countries than in the U.S. where it is not exactly standing still either. Here bandwidth is also an issue. Will the popularity of, say picture phones—not to mention video on cell phones—put a significant burden on the wireless infrastructure? How do you see the wireless aspect playing out in the broad context and where do you see the largest opportunities for participants in the embedded computer industry?

Grout: The proliferation of wireless devices has been significant in geographies where landline systems were yet to be deployed. In China and India for example, it can be more cost-effective to invest in wireless capability rather than install or update a wired infrastructure. Deploying base stations instead of laying expensive copper has enabled connectivity for hundreds of thousands of users in these regions in a very short interval. Expansion of wireless in more developed nations and those operating on 2 and 2.5G systems, will come from 3G infrastructure deployment, which is necessary to handle the growing number of users operating more powerful devices. Service providers are accelerating their plans for 3G infrastructure and services. Cingular, Korea Telecom, Nextel, Orange Communications, Verizon and Vodafone among others have all announced plans to upgrade their infrastructure and intend to roll out 3G services in test markets. The 3G services cater to higher bandwidth data services including devices such as picture and video phones, which are aimed at increasing the average revenue per unit.

In the embedded computer industry there are two categories of growth—basic service for developing geographies and enhanced features for developed economies. Both pose a significant opportunity for embedded systems suppliers. In developing economies, lowering costs (capital and operating expenditures) is a priority. New standards-based technologies like AdvancedTCA bring cost-effective enterprise technologies such as Ethernet, Linux, storage and processors to the once-proprietary world of telecom equipment. These technologies combined with the power of standards-based volumes enable a lower cost infrastructure than has been possible in the past.

In developed economies, users are demanding far more complex services with higher levels of security. Here, new architectures that provide wire-speed packet routing, processing and security are in high demand. Architectures like AdvancedTCA enable higher bandwidth and more sophisticated functionality than has been achievable in the past. Moving forward there will be a growing need to complement embedded computer solutions with more software, more integration and more sophisticated technologies like network processors to accommodate more users and smarter devices.

RTC: Perhaps the more immediate phenomenon is the gradual but steady move toward VoIP displacing the traditional POTS and/or circuit-switched networks. While this is not happening overnight, how do you see the dynamics playing out between the traditional telecom companies and the cable/Internet providers and, for that matter, free interlopers such as SKYPE?

Grout: With Vonage taking advantage of the broadband growth and addressing the residential voice market, the traditional PSTN service providers are looking to protect their customer base with their own VoIP migration. Traditional telecom companies use DSL services while broadband providers like Comcast are exploiting the cable infrastructure to deliver VoIP services. The continued growth in broadband will accelerate this transition and result in more customers, which means opportunity for new comers like Vonage and SKYPE. In either case, the infrastructure will be upgraded to replace the traditional POTS infrastructure. Due to these changes, RadiSys sees opportunities for next-generation solutions in B-RAS, media gateways and softswitches.

RTC: Today’s cable/Internet providers may have some advantage in that they use commodity-level blade servers and generate revenues whether the traffic is voice or data. Do you expect this advantage to continue as customers expect more and more sophisticated services? Does this present an opportunity for a standard like ATCA to move into both the traditional telecom as well as the networking space?

Grout: The key to the AdvancedTCA transition is the requirement of “telco grade” and the scalability/availability of the solutions. Traditionally, this becomes an issue as the deployment grows and the established service providers get involved. This is now happening and we expect this transition to continue. AdvancedTCA is unlikely to penetrate every blade server opportunity among cable and Internet providers. The equipment that can reside in a non-telco facility will continue to be “commodity-level” blade servers, or rack mounted servers.

I think the advantage cable and Internet providers have enjoyed is not necessarily using commodity-level blade servers, but rather they have adopted the principle of open and off-the shelf solutions. This has not been the case in the world of telecommunications. With AdvancedTCA, carriers can now take advantage of enterprise technologies (such as Ethernet, Linux, Intel processors, etc.) that are packaged in a five-nines solution and available in a uniform way from many providers. The real savings and leverage point is open, modular, standards-based solutions packaged into a highly reliable solution. It is unlikely that service providers will compromise on network reliability, but using AdvancedTCA they can take advantage of standards-based solutions that maintain high-level service requirements.

RTC: What importance do you think that standards such as ATCA, COM Express and MicroTCA, to name a few, will have going forward in the embedded computing business, primarily in communications? For example, will standards such as ATCA enable smaller players to participate in large projects (much like the COTS initiative has done in the military arena) or will they tend to restrict participation to a relatively few large companies?

Grout: Standards play a very important role in the embedded industry, which is why RadiSys is an executive member of PICMG and the SA Forum. Standards are essential to drive down costs and enable outsourcing through a broad ecosystem of suppliers, reducing single source scenarios. However, standards don’t solve all problems. Within most standards there are still options among which individual implementations must choose—the fabric in AdvancedTCA is an obvious example. Because of this, standards-based products don’t simply snap together like Legos to become a working system. Complete system architecture, integration and validation play a key role in the successful deployment of standards-based products and this is a key capability that RadiSys delivers as an integrated platform provider. While the standards open up industry-wide participation by companies of all sizes at the building block level, there will likely be a smaller number of larger companies with the architecture, integration and validation, and supply chain management capability to deliver complete system platforms.

RTC: Switched Fabrics: Do you anticipate that the strong acceptance of PCI Express as a significant factor in connectivity will bring about a major revolution in embedded computing? Does Radisys currently use any switched fabrics (other than Ethernet in PICMG 2.16 and ATCA) in any of its products or does it intend to?

Grout: PCI Express is a critically important, standard interconnect technology. It greatly enhances the ability to connect myriad processing and I/O elements in a standard manner with scalable performance and deterministic extensibility. Initially, it will be evolutionary, replacing the ubiquitous PCI bus in embedded board and system designs. In the future it will become an enabler for modular computing platforms, leveraging multi-vendor building block solutions. RadiSys supports several types of switch fabrics in current and planned products including Gb and 10 Gb Ethernet, PCI Express and Advanced Switching Interconnect, as well as several proprietary switch fabrics for specific applications.

RTC: The issue of China is a complex one as we tend to treat it as a large market opportunity as well as a powerful and potentially more aggressive competitor, even a threat. How do you balance the competitiveness of China’s manufacturing—which many U.S. companies are taking advantage of—with the need to address the country with the largest population in the world as a customer?

Grout: China is an incredibly vibrant market that represents a huge opportunity for RadiSys. We have a three-tiered, long-term strategy to address this region. First and foremost, we are very excited by the growing market opportunity, and to serve this market well, we believe it is important to develop and build products in China. To support our strategy, we recently opened a design center in Shanghai and have established relationships with local manufacturing and distribution partners. In addition to supporting our strategy to serve China-based customers, we derive advantage by tapping into the impressive design and manufacturing capabilities that exist there. Certainly there is higher complexity and potentially higher risk when working in any new geography, but we have a long-term and meaningful commitment to the region.

RTC: The recent acquisition of Force by Motorola represents a large piece of consolidation and puts considerable “force” behind ATCA. Do you expect this to foster growth in the market generally, or simply damage competitiveness and drive the market to commodity status?

Grout: Motorola’s move is further validation of the AdvancedTCA market. Some analysts are expecting the market to grow to multiple billions by 2007, and with relatively few players in the market this equals tremendous opportunity for RadiSys.

Even during the telecommunications downturn we continued to invest in R&D and today that investment is paying off. This year companies will begin serious evaluation of AdvancedTCA products and begin incorporating them into their designs in late-2005. The embedded space is going through a transformation from one-off board level designs to more modular and more integrated solutions. This is very exciting. To do this well, there will likely be meaningful consolidation in the industry. For these types of solutions to really penetrate the market, there must be 2-3 highly credible providers, or choices for the users. The industry is beginning to unfold in a way that can support this shift.

In terms of commoditization, we believe there will be many ways to competitively differentiate our solutions versus those available from others. We can compete and win on features, performance, time-to-market, manageability, scalability and upgradeability, etc. So, standards like AdvancedTCA allow for a massive increase in the TAM available to RadiSys, and the competencies of system architecture, development, integration, validation and deployment give us many ways to differentiate, compete and win.