The requirements of the defence industry differ from end user and enterprise computing needs in many ways. Commercial off-the-shelf (COTS) solutions are often deployed within a programme to leverage cost and performance benefits, although further modification may be needed to protect the unit in operation. That is the state of play today, but what about in a year, five years or even ten years from now?
Something quite interesting is happening in the enterprise, OEM and consumer computing markets. And it is going largely unnoticed.
Electronic components are evolving and becoming less susceptible to thermal changes. They are becoming more robust and increasingly capable of use outside the manufacturer’s intended applications. Gaming boards are starting to use high quality extended temperature components to enable overclocking, and in doing so, moving towards suitability for other applications. At the same time, enterprise solutions are being designed with longer service lives, so manufacturers are now able to provide extended warranty options to customers.
Hyperscale data centre users are pressurising manufacturers to use extended temperature components too, allowing the centre to run at a higher and more cost-effective thermal range, thus reducing cooling costs. SSD (solid state drive) technology is maturing and overcoming reliability challenges of the past, and processing capability continues to increase but now requires less energy and space. This reduces mass, adds functionality not previously possible and decreases cost against given performance metrics.
Is there an immediate benefit today for defence computing? Clearly there are some to be made, but the real one is perhaps still a few years away. Changes already beginning to appear are certainly advantageous, but they are also an unintended by-product of independent development streams. A keen eye can spot components and features on a board and know the likely suitability for an application. But without experience in testing for validation, this is only an assumption. In reality, very few commercial boards can be leveraged to offer what is really needed to deliver robust and reliable defence systems.
There is certainly a place for COTS hardware and computing in defence, as there is a place for highly engineered specialised computing solutions. But for an increasingly large number of applications, demands to operate COTS equipment in harsh environments were not part of the original manufacturer’s design intent. If improper consideration is given to the hardware selection and any necessary modifications, there is a real risk of unexpected or accelerated failures. Inevitably, this will lead to an escalation in total cost of ownership and a poor user experience.
So what is the solution? A robust billet chassis will meet certain needs. However, much like using a sledge hammer to crack a walnut, these highly engineered solutions are unnecessary for systems that need to be robust, but not to extreme levels.
An in-depth understanding of how to enhance commercial components and modify a commercial chassis will enable performance and cost advantages to be utilised with minimum dilution in demanding applications.
Issues such as shock, vibration, salt spray, cable separation, EMI filtering, IP ratings and cooling present significant challenges to COTS computers and associated networking. Specialised mounting systems, protection provided by the enclosures, isolation and shock protection can enable the successful use of COTS equipment. This will allow for competitive pricing, performance and availability to be leveraged – even in highly challenging environments.
Relationships with partners that excel in engineering solutions designed to bridge the divide between commercial and military off-the-shelf equipment is crucial. Through collaborative partnerships, the full potential of COTS hardware can be truly unlocked.
