Full System Simulation

Executive Summary

There’s never been a moment in time quite like the one we’re in. As our connected systems grow exponentially complex, there is a need to accelerate development teams like never before.

Full-system simulation is a true paradigm shift in the DevSecOps space that offers the environment needed to do just that. A simulation environment empowers developers and programmers to collaborate, reverse execute, and record results in such a way that DevOps teams are able to save 90 percent of testing time and scale exponentially. This rapidly growing area of technology is sure to impact businesses of all sizes in the near future.

We know that hardware-based testing is not going to cut it long-term for organizations that want to remain competitive.The question before developers is a simple one, and it’s a matter of investment: Will their organizations be impacted positively because they adopted simulation-based testing over hardware-based testing, or impacted negatively because their competitors did?

The Landscape of Programming & Development Demands a New Approach

To a DevSecOps professional, time is so much more than a precious commodity.

It’s also a measure of speed, performance, and one’s ability to stay competitive in today’s complex and connected landscape. Time, and your ability to leverage it well, will ultimately determine your organization’s ability to scale, and the speed at which it can do it.

But what if you could take everything you’ve come to accept about time constraints on your teams and turn it upside down, allowing you to conduct up to 80 percent of your tedious hardware-based software testing 90 percent faster?

This is now possible with full-system simulators, also known as digital twins, that readily replace the hardware-based testing that is currently hindering your team’s ability to scale. At Wind River®, we’ve been perfecting the art of full-system simulators using Wind River Simics, and the impact on our DevSecOps team’s ability to test, innovate, and scale has been astounding.

And while a full-system simulator may sound like a luxury reserved for enterprise-level organizations, businesses of all sizes will need to adopt this style of testing in the coming years. Customer and company expectations for speed and accuracy in today’s world of complex, multilayered architectures will demand it due to the limitations of hardware-based testing. It’s only a matter of time.

What Is Full System Simulation?

Full-system simulation is a true paradigm shift in the DevSecOps space, and the impact on an organization’s ability to iterate and scale cannot be overstated.

Having a simulated digital twin of your hardware allows you to run the majority of your software tests in a simulated representative hardware environment, reducing your team’s reliance on manual and time-consuming hardware-based testing. Since repetition in a tried-and-true environment is necessary–especially in deterministic simulation–this ability to iterate and troubleshoot rapidly in a simulation of any complex system becomes game changing, to say the least.

Deterministic simulation and why it matters 

Simulation with determinism is when we seek to achieve the same results every time we run and record something several times, creating a baseline. Once this baseline is created, we can create intentionally bad, glitchy, or difficult testing environments and situations to test software. We can aim to trick or break the software in a controlled way as we exercise parts of its code.

This means teams can go from running single hardware-based tests to multiple tests at a time. With full system simulations, you move from, say, one RF, to multiple at the same time, and scale it using simulation technology. You can also gain insights into the interaction between different processors. For instance, by leveraging this technology, you can go from one CPU to an entire simulation of the complete base station, and simulations can be recorded and shared with colleagues in various locations.

Test rapidly, reverse execute errors, and record all results 

With hardware-based testing, it’s not uncommon to reach an error and not know where in your binary the problematic code is. With Simics simulator, it’s easy to reverse-execute errors to see what led to the failed condition, recording the results.

We can also cause these errors intentionally, creating intentionally difficult environments in order to simulate various scenarios.

In either scenario, whether the failed condition was intentional or accidental, we can grow our understanding of the interaction between software and representative hardware.

Automating regression tests

Because we can create a digital twin, it’s easy to set up automated regression tests to run as frequently as needed. This becomes critical as more nimble DevSecOps teams spin up new iterations of their software or roll out frequent updates. And if anything about the hardware or prototype hardware changes, adjusting your digital twin is easy.

Why Choose Simics

Simulation testing gives DevSecOps teams the one thing they always need more of—time.

Wind River has spent the last several years perfecting an effective and efficient methodology for simulation-based testing, and we found that we were able to bring 60–80 percent of our hardware-based tests into a simulated environment, reducing our time spent on testing by up to 90 percent. And because we recognize collaboration as a key to successfully and efficiently developing software, it is critical to ensure Simics is easily accessible by your whole team.

We’re excited for this new world of “Simulation as a Service,” and Simics is uniquely designed to help DevOps save time when it comes to software testing. In fact, Wind River customers report that they’re saving 80% to 90% of their time after adopting simulation into their workflow.

Ready to learn more about how our full-system simulation platform can help your business?

Let’s connect, or check out our webinar, Using Simulation Technology to Launch a Robotic System Program.