Defense budget negotiations begin next week in the US House of Representatives, and the ongoing sequestration is once again driving deep cuts across the military – including major cuts to technology spending. These harsh cuts are drawing loud criticism from across the political spectrum, but few observers expect the situation to change soon. And the situation in the US is hardly unique. Global military budgets are also under fire, with weak economies and austerity measures putting pressure on spending worldwide.
Despite these tight budgets, armed forces worldwide are still need advanced technology to fulfill their missions. This creates a tough challenge for designers, who are being asked to create cutting-edge solutions on a shoestring budget. To deal with this new reality, engineers need to carefully optimize their development processes. Key areas to consider include commercial off-the-shelf (COTS) hardware and portable software – two approaches that can cut costs not only today but in the future.
Let’s start with a look at the hardware. COTS boards have already gained widespread popularity, but I think it is worth considering a move beyond boards to deployment-ready systems. These systems can provide major schedule and risk benefits by pre-integrating COTS boards, storage, power supplies, and other components. To illustrate this point, look at the testing requirements for a conventional system. Qualification can require 20-25 individual tests, each of which can cost up to $3,000 USD and last up to three days – a major undertaking, to say the least.
Deployment-ready systems avoid much of this testing, and can minimize other integration, optimization, and qualification burdens. What’s more, you don’t have to compromise technical specs to take get these benefits. For example, the CRS 48.5 from GE Intelligent Platforms is available with the latest 4th generation Intel® Core processor product family, which provides cutting-edge performance and power efficiency. Check out GE’s COTS Rugged Systems page for more info on this and other COTS systems. And to see how pre-integrated systems can be deployed in UAV applications, I recommend ADLINK’s recent article.
Figure 1. The GE Intelligent Platforms CRS 48.5 is built on cutting-edge tech.
Of course, hardware design is only part of the development process. Software development presents its own challenges, particularly in high-performance applications like signal and sensor processing. These applications typically require multi-processor architectures and intense performance optimization – two demands that can strain development budgets and timelines. Fortunately, COTS development tools and middleware can meet the challenge.
One good example here is the recently-upgraded Advanced Multiprocessor Integrated Software (AXIS) development environment from GE Intelligent Platforms. The tool set includes a variety of features that aid multi-processor development, such a GUI for visualizing and tuning processes across multi-threaded, multi-core, and multi-node platforms. AXIS 6.0 adds support for the 4th generation Intel Core processor and support for the Open MPI Message Passing Interface (MPI).
The Open MPI support is particularly noteworthy because it enhances some of the key benefits of COTS hardware, namely better code re-use and portability. Because it is a widely accepted standard, Open MPI-based software can be more readily re-used on new hardware and ported across different vendors’ hardware. For example, Mecury System’s tools also support Open MPI.
Similar benefits are available from other widely-used middleware like Gedae* and Data-Distribution Service for Real-Time Systems (DDS) – for more on this and related development topics, I recommend reading the Curtiss-Wright paper Five Budget-Busting Cost Drivers that Threaten Sensor Processing System Design.
GE’s upgraded tools also include the latest version of AXISLib, a set of optimized DSP and math libraries. These libraries recently added support for the Intel® Advanced Vector Extensions 2.0 (Intel® AVX 2.0), providing an up to 2X boost in signal processing performance. It is hard to overstate the value of libraries like AXISLib. DSP code optimization is one of the most challenging tasks a programmer can face. Using a math library lets you avoid the optimization time sink and all the headaches that go with it.
We’ve written a lot about Intel AVX 2.0 and the libraries that support it. But if you missed our earlier comments, a good place to start is the Intel AVX 2.0 primer written by N. A. Software. As illustrated in Figure 2, this article lays out the details of the architecture and the performance improvements you can expect. For a few more details and a quick look at the Curtiss-Wright Continuum Vector libraries, see our Intel AVX 2.0 imaging blog.
Figure 2. Intel® AVX 2.0 provides up to 2X faster performance.
The hardware, tools, and libraries I’ve described here are only a few of the options available from the Intel® Internet of Things Solutions Alliance. From modular components to market-ready systems, Intel and the 250+ global member companies of the Alliance provide scalable, interoperable solutions that accelerate deployment of intelligent devices and end-to-end analytics. Close collaboration with Intel and each other enables Alliance members to innovate with the latest technologies, helping developers deliver first-in-market solutions.
You can find more of the Alliance’s mil/aero products in our Solutions Directory. And of course you can browse this community for more insights on designing advanced solutions on a tight budget. Let me know if there is a topic you’d like to see us address!
ADLINK is a Premier member of the Intel® Internet of Things Solutions Alliance. GE Intelligent Platforms is an Associate member of the Alliance. Curtiss-Wright Controls Defense Solutions, Gedae, Inc., and Mercury Systems, Inc. are General members.
Roving Reporter (Intel Contractor), Intel® Internet of Things Solutions Alliance
Editor-In-Chief, Embedded Innovator magazine
Follow me on Twitter: @kentonwilliston