VME: Kontron is producing VPX offerings these days. In light of the hubbub surrounding the outside-VSO OpenVPX working group, which side of the fence does Kontron stand on – VSO or OpenVPX? Why?
SPARRVIK: Kontron has and always will be 100 percent focused on our customers and what works best for them. To that point, we surround ourselves with the best suppliers and partners, such as the VITA Standards Organization. Kontron has authored several open standards. In addition, we support and will continue to support relevant open standards adopted by the proper industry authorities because we know our customers have come to depend on organizations such as VITA.
VME: What effect do you think the OpenVPX working group will have on VPX’s future? On VSO? On the industry as a whole?
SPARRVIK: Again, our focus is on the customer. Lately, the military industry focus has been on the battle between VSO and OpenVPX, and not on the relevant benefits of VPX over other competing architectures. We are confident that other players in the market will see the value of working together in joint effort to address VPX and its future. VPX has a bright future in the defense arena, and we believe the industry will move forward together to best address customers’ needs.
VME: Naysayers have been pronouncing VME nearly dead for years. Will VME ever be fully replaced by VPX, SFFs, or some other technology – or will it remain alive and kicking?
SPARRVIK: Kontron considers VPX the next-generation VME, so we absolutely disagree with the assumption that VME is dying. If you are talking about legacy three-row and five-row VME, we do not consider this configuration on life support either. In fact, we are seeing many legacy military programs specifically ask for VME to upgrade their existing systems. The cost of replacing legacy VME chassis, I/O cards, and software precludes this choice for many large programs. That said, there is far more competition for VME than ever.
VME: Which other three VITA standards (besides VPX) do you predict will gain market traction in the next 5 to 10 years, and why?
SPARRVIK: Three standards that come to mind are VITA 47, VITA 48, and VITA 57. VITA 47 is the common conduction-cooled and ruggedization standard for all suppliers. This standard is vitally important for the customer because it ensures a uniform process in board-level screening from vendors. Customers can now easily compare boards from different vendors who have all been screened using the same methods and standards.
VITA 48 REDI is another standard that will gain prominence in the coming years. Since this is a mechanical standard and not electrical and sets the stage for heat dissipation and maintenance, we believe that this is another vital tool that will standardize mechanical specifications that meet customer and application requirements.
The last is the VITA 57 FMC specification. FPGAs are gaining prominence as multidimensional tools offering an inherent value to the customer for long-term benefits and control of their projects and programs. This standard allows customers to maintain their IP even if the FPGA goes End Of Life [EOL]. Imagine having control over your IP for 15 to 20 years. Imagine how much money that would save the federal government. We think the FMC spec might wind up being even more popular than the VITA 42 XMC specification.
VME: Which ONE technology will most affect mission-critical systems within the next decade, and why?
SPARRVIK: As we look ahead, we see technology getting smaller, faster, and more power efficient. This is due to the advancements in multicore technology. As a Premier member of Intel’s Embedded Communications Alliance (ECA), we’ve seen firsthand the advantages that Intel-based multicore products bring to the table. Customers are looking for the most power in the smallest footprint. Many products are designed with space constraints in mind, and that means we have to pack as much power into each board and system as we can. Multicore technology is also extremely reliable, and that is a “must-have” for any mission-critical application. These multicore technologies are allowing customers to develop products, such as UAVs, that keep humans out of harm’s way. These types of mission-critical applications demand the most efficient processing available, and multicore is addressing this extremely well.
VME: What are the biggest challenges facing your VME customers these days?
SPARRVIK: Sustainability of large, big-ticket military programs is at the forefront of the current budget realities. However, we see that many of these programs present small opportunities for legacy three- and five-row VME. While the new big-ticket programs have been migrating to VPX, the older programs absolutely need the funding to update and upgrade vehicles and planes that must be replaced when coming back from multiple theaters of combat. Older aircraft must have their electronics upgraded, and some of the electronics are decades old. This maintenance cannot be delayed. In other words, if you focus too heavily on just the large military programs, suppliers are missing true growth opportunities for upgrade programs over the next few years.
VME: Kontron recently released VPX boards featuring the PowerPC and Intel Core 2 Duo processors, and offers other form factors featuring the Atom processor. Are there any drawbacks to the Atom, or is it the panacea for all that ails board designers?
SPARRVIK: As I mentioned earlier, our Premier status within Intel’s ECA allows us to work with cutting-edge processor technology in advance and has given us a firsthand view of the excitement stemming from the launch of the Intel Atom. It is a unique product that is generating a lot of interest (and business) for low-power applications in small form factor lines. So your question about the Atom and VPX is a good one.
To explain, the Intel Atom is extremely well-targeted for embedded systems, and that is why it has seen so much success in small form factor lines. However, the high-speed I/O capabilities of the Intel Atom (eMenlow) platform have been shown to fall short of what defense users expect and need in a high-end SBC form factor such as VPX (or VME and CompactPCI). For example, a typical customer expectation in any of these form factors is that the SBC can communicate with one or more GbE connections at full speed, and that communication does not impact the throughput of other I/O devices such as SATA, PCI, or VMEbus. A survey on the existing Atom-based boards on the market reveals that the chipset simply does not offer enough high-speed ports to support the required bandwidth. So we see a huge potential for the Intel Atom overall, but just not in these form factors today. Of course, we will reevaluate as the high-speed I/O capabilities of this platform continue to evolve.
VME: What are the top two issues facing defense industry suppliers like Kontron in the next five years?
SPARRVIK: From our point of view, what Kontron and all defense suppliers face is the challenge to constantly adapt products and services to a changing market. The defense computing industry used to be dominated by VME and PowerPC sold to numerous agents in the programs procurement food chain. Today, however, defense contractors are designing more complex systems that require that multiple standards run on numerous form factors. Kontron’s strategy is to replicate commercial product success in the defense market, and that has positioned us extremely well. So really, today’s defense industry suppliers need to keep ahead of customer expectations in terms of service, support, and technology that match life cycle demands.
Kontron 858-677-0877 • www.kontron.com