RedHawk KVM-RT Introduced

Concurrent Real-Time, a global provider of high-performance Linux® solutions, today announced RedHawk KVM-RT™, a new virtual real-time environment based on the Kernel Virtual Machine (KVM) hypervisor. With RedHawk Linux as the host operating system, a KVM-RT installation provides guaranteed real-time performance in virtual machines. Multiple RedHawk real-time guests and multiple Windows and non-real-time Linux guests are supported. KVM-RT can achieve worst case response times of less than 10 microseconds on certified hardware platforms.

RedHawk Linux is currently the industry’s leading real-time Linux operating system for both x86 and ARM64 systems. It is widely used in mission-critical applications including simulation, training, data acquisition and process control. The release of KVM-RT allows real-time multiprocessing applications to better utilize available computing resources.

“KVM-RT is the ideal solution when one or more virtual machines need hard-real-time performance. Time-critical applications that traditionally used dedicated real-time boxes can now share hardware platforms with non-real-time.” said Ken Jackson, Concurrent Real-Time President and CEO. “Our testing indicates that KVM-RT achieves far better real-time performance than is possible with other hypervisor solutions.”

KVM is a Type 1 hypervisor that is included with RedHawk Linux. RedHawk KVM-RT leverages the unique features of the RedHawk host for memory management, process scheduling, device access and I/O. KVM-RT creates individual threads for each virtual CPU allocated to a virtual machine. Real-time performance is guaranteed through RedHawk’s shielding and scheduling mechanisms. I/O devices can be completely passed through to specific guests or para-virtualized to reduce the idle time of devices. Para-virtualized devices reduce the amount of necessary hardware without compromising throughput.

A KVM-RT configuration tool is available to assist the user in allocating a platform’s processor cores, memory, peripherals and other hardware resources to each virtual machine.