RT-LAB provides tools for easy separation of the system model into subsystem models that can be executed on parallel target processors.

In this way, if you need to run a model in real-time that cannot be run on a single processor, RT-LAB provides a means of sharing the load over several processors.

RT-LAB is the first fully scalable simulation and control package that allows you to separate models for execution in parallel on a network of standard desktop PCs.

Driven by the demands of a mass market, users take advantage of rapid advancements in a wide range of readily available technologies at relatively low costs.

RT-LAB uses standard Ethernet as well as various PCI and PCIe expansion boards providing many I/O capabilities (digital, analog, communication protocols, etc.).

RT-LAB integrates with Opal-RT's OP5000 series, OP7000 series, and OP4500 series of hardware interface devices for nanosecond precision timing and real-time performance.

RT-LAB also supports cards from other leading manufacturers such as National Instruments, Softing, SBS, ConnecTech, Max Technologies, Pickering, Spectracom, and others.

At execution time, RT-LAB provides seamless support for inter-processor communication, using any combination of UDP/IP, Shared Memory and real-time low-latency communication of data between the target processors.

Users can also interact with the simulation in real-time from the host station using TCP/IP. The real-time communication is ensured by Dolphin Interconnect, yet another off-the-shelf solution.

RT-LAB runs on an open-source distribution of Linux optimized for real-time performances. 

For users that want to start their testing in soft real-time, RT-LAB also offers the possibility of using Windows as a target node operating system.

Within a time step, the system is doing more than computing the dynamic model. It also does administrative tasks, such as reading and writing I/O, updating the system clock, scheduling tasks, logging data, and handling communications. This restricts the amount of time available within a frame to compute the model values limiting the size of the model that can be computed on a single processor.

RT-LAB has reduced this overhead to a small percentage of raw hardware performance without losing functionality, thereby increasing the capacity to compute more complex models.

RT-LAB XHP (eXtra High Performance) mode allows very fast computation of the real-time model on the target system. This has allowed our customers to simulate complex systems over distributed processors, with analog and digital I/O, at cycle times below 10 microseconds.

The XHP Mode of RT-LAB can slash scheduling overhead to less than a microsecond, letting you use the full power of your system for the computation of highly dynamic models in real-time which is a solution for the developers who have the constant challenge of achieving accurate, high fidelity responses for the real-time simulation of increasingly complex models.

Even when the signals in a hardware-in-the-loop (HIL) system need only be updated in the hundreds of microseconds time frame, the model may need to be computed many times between each major time step in order to maintain numerical accuracy. The XHP Mode out-performs any other real-time system, and is particularly useful for modeling electrical systems, such as drive controls and power electronics.