What may surprise business owners, program managers, and systems engineers is that technology may not be the most important aspect when making form-factor decisions. The success or failure of a new product design may depend on the actual business model of the commercial-off-the-shelf (COTS) provider. In recent years companies have moved away from designing in house (for reasons I will discuss in a moment). An alternative to designing in-house is outsourcing to companies that already have a proven COTS platform. There are clear advantages in leveraging COTS: faster time to market, lower engineering costs, lower risk, and immediate development of the software utilizing a COTS development platform (as opposed to multiple months until first prototype boards are ready). The COTS marketplace may be segmented into two main approaches: single board computer (SBC) and system-on-module (SOM).
Companies who historically designed products in-house are being forced to look at alternative solutions due to many factors in the evolving market. The cost, complexity, schedule, and risks of designing custom electronics have caused a business-driven trend towards the utilization of COTS SBCs. This includes leveraging an existing COTS design to create modified COTS circuitry. The speeds of the latest processors, memory densities, and high speed interfaces (e.g. USB 3.0, PCIe, 10/100/1000 Ethernet, HDMI, SATA) drive up microprocessor board design costs and design complexity. In addition, these designs require: larger engineering teams, more expensive and complex tools (schematics, layout, and signal integrity simulation), longer product design cycles, and higher prototype costs.
SOMs are designed to be inexpensive, adhere to standard interfaces that may constrain system architectures, lack the full set of system peripheral circuitry, and require a carrier board to complete peripheral implementation and enable system integration. The carrier board design partially offsets the advantages of using a module.
The business model for SOM module vendors favors high volume clients that require little technical support. Modules conform to standards (such as QSeven), and therefore constrain board area, pin-out, and offer little means of product differentiation between vendors. Consequently, differentiation is based mostly on price, which leads to low profit margins and low overhead for SOM vendors– minimizing their engineering support for customers. Thus, it’s their high volume customers (10k+ units) who are likely to receive high priority engineering support when needed. Support may include the up-front expertise in finalizing carrier card architecture and design or the resolution of problems once in production. A company with lower volumes does not align with the module vendor’s business case, and thus must invest their own engineering teams to resolve issues. This is often a hidden cost of using SOMs. To complicate this further, many off-shore SOM manufacturers directly target large volume opportunities in North America. This leads to inefficiencies in or outright problems with communications and support both in the engineering and production phases. This elevates risk and has often resulted in complete failure.
SBCs include the processor and memory that are present on SOMs and add integrated peripherals and I/O. This includes support for features, such as battery charging/monitoring, system power management, and display/touch screen interfaces. SBCs are free from the feature-constraining standards, which enables them to provide higher feature density and optimized system architectures, which yield lowest system power and cost. The SBC can often be purchased for use as a standalone board or mounted on a custom carrier that implements specialized circuitry. SBC architecture also provides a straight forward path to cost and feature-optimized modified COTS devices. With modified COTS, clients can leverage the COTS design and have a product built to their exact product requirements.
The business model for SBC manufacturers targets customers with low to mid volume. Unlike SOM vendors basing their business models on standards and price, SBC form-factor freedom enables product differentiation through features, target market characteristics, and support. The SBC manufacturer offers direct engineering support to the lower volume customers as well as engineering design services that offer a simple migration path to semi-custom (modified COTS) designs. The business model of the lower volume product developer and the SBC vendor complement each other. This alignment increases on-time and on-budget project completion with minimal problems during new product introduction.
Understanding business models of SOM vs SBC vendors is key to a successful modern electronic product design. When selecting a vendor, be sure you understand the level of support, and the underlying incentives the vendor has for earning your business. Aligning your interests with your vendor’s interests goes a long way towards the success of your product – and ultimately your business.