B&C: VIA has become famous in the consumer market during the second half of the 90s and the beginning of 2000, first as a chipsets manufacturer, and, after the Cyrix and Centaur acquisitions, as a low-end CPU manufacturer. Could you please explain the reasons behind the choice of VIA to leave those markets and why it chose to remain only in the business of the Embedded products and All-In-One Mini ITX and Nano boards?
TA: Even during the days when we were a leader in the PC chipset market, we had a very clear understanding that low power and increased integration of components at both the silicon and motherboard level would be key requirements for both PCs and other connected devices. By concentrating on these two key areas, we were able to build a strong position in the embedded market and decided to focus on fast growing segments such as thin clients and digital signage. As we have seen over the last couple over years, the PC market has become extremely unhealthy with overall sales and prices declining and minimal opportunities for innovation and differentiation. While we do still sell into the PC market, we prefer to focus our energies on high potential new markets that place value on our strengths not just in hardware but also increasingly in software.
B&C: VIA was the first company which has believed in the commercial success of Mini ITX boards within the consumer market with the various Epia, but the high sale price has prevented a huge diffusion. Afterwards Intel has recognized the potential of this type of boards, and Intel’s solutions, based on the Atom CPU, have supplanted VIA’s products from the market. Today VIA, to not do the same mistake, is selling the VAB boards based on ARM SoC and the Springboard platforms. Both of these products have an excellent price/quality ratio, so that they can compete with the other cards based on ARM SoC, such as the Raspberry Pi and Cubieboard. Unfortunately it is very difficult, if not impossible, to find the VIA boards in shops, unlike its competitors, with all that this entails in terms of sales. In Italy, for example, the shipment of a Springboard costs $30, which is the 30 % of the cost of the board. Are you planning, in the near future, some more aggressive marketing campaign? It’s a shame that such good offers are not sold for logistical inefficiencies.
TA: We continue to sell a large volume of Mini-ITX boards globally by providing real value to our customers through strong customer support, software and hardware development and customization services, and longevity support for up to seven years. This is a much healthier way of doing business than simply dropping prices to the floor and it means we can build strong partnerships with our customers that enable them to innovate and differentiate their products. One of the main reasons why the PC market has slowed down so dramatically is that PC OEMs and makers were given very little room for innovation and thus were only able to compete on price. As a result, the market has become much less attractive because of the lack of exciting new products.
Raspberry Pi have done an incredible job of reigniting interest in hardware among enthusiasts worldwide. Watching the excitement they have generated takes me back to the days of the launch of the Mini-ITX! There are a lot of exciting projects being carried out with Raspberry Pi, but one of the key challenges many people are facing is how to turn these great ideas into actual products. One of our key goals with Springboard is to provide a platform that will allow people to turn their ideas into actual products that they can sell on the market. That means providing comprehensive Android and Linux software development packages as well as the board itself. It also means offering extensive customer support as well as specialist software and hardware customization services to help them make the product ready for the manufacturing. One of the major complaints about a lot of hardware projects that are launched on sites such as Kickstarter is that they almost invariably get delayed because teams underestomate the complexities of readying a product for manufacturing. With Springboard our aim is to streamline this process.
We do plan on promoting VIA Springboard more in the market place, but we belive we offer a strong value proposition with the board and dont plan on changing the pricing.
B&C: I had the chance to try out the development platform based on the Springboard VAB-600 card, and I found it really good. On the forums and the chan discussions, however, I noticed that a criticism often is made against this type of cards: the chronic lack of at least two SATA ports and two Ethernet ports. It is true that boards are dedicated to developers, but many would like to buy these cheap ARM cards to build a small home server or NAS (through software such as OpenMediaVault), but often the inability to use one or more HDD holds many buyers; so they are forced to fall back on a Mini-ITX based on a x86 Atom CPU, cheaper and easier to find (these cards are, however, very limited, having a maximum of two SATA ports and one Ethernet port). Have you explored the market to see if this type of solution can be successful? If so, are you planning the construction of any product at a good price (you begin to find some NAS x86 below 300 €) for this niche market?
TA: I’m glad you found Springboard to be good in your testing! As you mention in your question, the home server market is already served quite well by low cost x86 platforms. Our hope with Springboard is that people will use the platform to create new types of connected devices.
B&C: Talking about the competitors, the Raspberry Pi is having a considerable success as living room media player, thanks to the excellent GNU/Linux OpenELEC distribution. Has VIA ever thought about making a Pico-ITX board exclusively dedicated to this market?
TA: We don’t see Raspberry Pi as a competitor. We would be delighted if enthusiasts used our Springboard platform for living room media player applications but currently we don’t plan on launching any platforms specifically dedicated to this market.
B&C: Intel is desperately trying to enter the Mobile market with the Atom SoCs, but without great success. Has VIA ever thought to do the same, creating a SoC based on the x86 ISA (because of the x86 license)? Or is aware that the x86 architecture will never beat ARM in this context? Possessing the subsidiary WonderMedia, scorer of their existing ARM SoC, in principle it would be quite possible for VIA.
TA: The short answer to this question is: no! We are focused on developing ARM SoCs and platforms. Just before Embedded World in February, we launched our latest VIA Elite E1000 SoC with the VIA VAB-1000 board optimized for graphic intensive Android performance. When we first launched ARM boards into the embedded market, many of our customers saw them as low cost alternatives to existing x86 boards but they quickly found that they could not be used as direct replacements. As a general rule, x86 is great for general purpose computing platforms whereas ARM is great for low power application-specific systems such as smart phones, tablets, and set top boxes.
B&C: VIA has recently announced a partnership with Mozilla for a Firefox OS optimization. Does VIA intend to enter the mobile market - thanks to WonderMedia, again - or will VIA’s role be limited to simple development platforms?
TA: We have been working with Mozilla for a while now on our APC, APC Rock, and APC Paper platforms and see interesting opportunities for various applications such as browser based PCs and Smart TVs. At Mobile World Congress, we showed our first generation of VIA Vixen reference tablets featuring Firefox OS and experienced a lot of positive interest in them. We plan to continue expanding the partnership with Mozilla.
B&C: On a different note, VIA has made its mark as a producer of excellent audio chip, with Vinyl family. Over the years it has managed to carve out a nice market share of motherboards, against Realtek, and in the market of discrete sound cards, to the detriment of Creative Labs. Although graphically we are making significant steps forward, but currently one of the biggest flaws of smartphones and tablets is the quality of the audio. Given that gaming will play an increasingly important role with these terminals, and the audio plays an important role during the gaming experience, are there some plans about the prospect to bring your technologies on these platforms in the future? Even through the sale of licenses to several manufacturers of ARM SoC.
TA: Currently our focus remains on the continued development of audio chips and codecs for the PC and peripherals market. We currently don’t have any plans to move into any other segments.
B&C: On Bits and Chips we have published an historical article on the Jalapeno Cyrix CPU, which, however, has never seen the market. VIA, when it bought Cyrix, gave up to Jalapeno CPU and to MediaGX CPUs (National Semiconductor, later, has sold these CPUs to AMD), taking for itself the projects of traditional CPU based on Core MII and MIII. If Via had bought the CPU MediaGX projects, perhaps It would have been able to present the first APU more than 10 years ago, instead of continuing the development of the classic x86 CPUs. Looking at the success the AMD APUs and ARM SoCs are having today, has VIA nothing to complain about this choice?
TA: A very interesting question! At the time we bought Cyrix and Centaur we didn’t have any experience in the x86 processor market at all so we decided to focus on discrete processors to begin with. That was a big enough challenge in itself. Taking on an even more complex product like the MediaGX would have been a huge drain on resources.
B&C: Turning to the foundries and fabless companies, different sources and many journalists talk about the prospect that the 28nm production node will remain the most affordable for many more years to come. Which processes are used by VIA at the time, and where (TSMC, UMC, GF, etc.)? In addition, to remain competitive, if any were needed, is VIA planning to continue to focus on actual production processes (40nm, 32nm, 28nm)? Or will VIA try the gamble, for example using a newer process (like Altera did with the 14nm of Intel)?
TA: We have a great relationship with our existing foundry partners; so I don’t see us gambling right now.
B&C: In conclusion, I have a question about the market in the near future. The Cloud Computing market will grow a lot, and it will focus on energy saving (Low Power Server, use of SoC instead of the classic CPU, use of SSD in place of the traditional HDD, etc.). This sector, to date, has not a leader. VIA for many years, more precisely since 2001, supports the Green Computing and, as we have already said, has over ten years of experience in the production of low-power embedded systems and cards Mini ITX and Pico for the consumer market. What role will VIA play in the near future in the Server and Embedded markets? Will be a star or a supporting actor?
TA: I think we would be delighted to win an Oscar for best supporting actor. VIA is not a consumer brand and our goal is to help our customers to come out with great products by proving the best platforms and services, including software and hardware customization.
For traditional embedded markets, we will continue to focus on delivering low power best of breed platforms and systems based on ARM and x86 to meet our customers’ needs. At the same time, we are also introducing exciting new products for markets such as digital signage. Did you know that all the top movie theaters in Taiwan now feature massive VIA Video Walls that show the latest movie trailers on multiple screens? Or that Taipei Zoo recently installed a VIA Video Wall to show visitors the new baby panda and his mother while they were lining up to see them? I’m hoping that some time we’ll be able to install VIA Video Walls in Milan to showcase the latest Italian fashion. Perhaps then we can become a star – but never a diva.
B&C: Thanks Tiziano, it has been a pleasure talking with you!
TA: Thank you! This has been fun.