The latest updates to Simics include accelerated simulation performance and expanded capabilities for enhancing access, collaboration, and automation for software developers. Additionally, the process for modelling, debugging, and model asset reuse has been further improved to deliver a simpler user experience. Other new improvements provide additional capabilities enabling agile and automated testing with shorter feedback loops.
Simulation performance has been enhanced for tightly coupled multi-core and multiprocessor systems on multi-core hosts. Testing on parallel workloads shows performance increases of up to seven times when simulating an eight-core Linux symmetric multiprocessing (SMP) target on an eight-core host. Coupled with its existing performance technologies, Simics can efficiently simulate ever more complex targets to deliver the benefits of simulation for any target system.
An important aspect of the offering, according to a WR spokesman, is that the capability scales; you can achieve the same degree of performance and flexibility in simulation of systems of any size. The simulation comes with a high degree of control including in the time dimension; you can halt on specified events and “rewind” the simulation to trace the source of unexpected behaviour.
Easier test automation comes from implified scripting interfaces and fault injection framework that allow easier use of Simics in test automation environments. A checkpoint server allows easier management of checkpoints across teams. Users can share the system state of their simulations with others.
A range of “Quick-start” platforms is provided: three quick-start platforms (Intel, ARM, and Power architectures) allows early prototyping and tooling integration. Enhanced support for SystemC models with a new SystemC library and improved debugging tools help teams get models up and running quickly.
Wind river sees a trend for IoT developers turning to Agile practices, which require constant on-demand access to the system, more efficient and faster collaboration between teams, and higher levels of automation. This release, WR says, represents a next-step in parallelism; support for multi-threading has been in place for some