PICMG, a leading consortium for the development of open embedded computing specifications, has announced major extensions to the COM-HPC Computer-on-Module (COM) standard with the ratification of COM-HPC 1.15 for Functional Safety (FuSa). COM-HPC 1.15 supports workload consolidated high-performance mixed-criticality for real-time edge servers and clients that must execute not only safety controls but also vision, AI and IIoT gateway logic or even TSN based 5G communication.
COM-HPC 1.15 is a set of safety extensions that expand the FuSa capabilities of “safety island” blocks on modern chipsets out to the broader system. Available on all COM-HPC form factors – including the upcoming COM-HPC Client Mini – COM-HPC FuSa extensions define a dedicated SPI signal that connects health and status monitoring features of such blocks to a FuSa “safety controller” located on a COM-HPC carrier board where any findings can be processed for external use.
The COM-HPC 1.15 architecture thus enables the creation of multicore embedded mixed-criticality systems by providing a direct path to redundancy and fail-safe process implementation for developers of robotics and industrial machine controls, train and wayside controls, autonomous vehicles, avionics, and other critical systems.
“With the small size definition of the upcoming Mini specification and the FuSa extensions, COM-HPC covers all use cases I can think of,” says Christian Eder, Chairman of the COM-HPC technical committee and Director Product Marketing at congatec. “COM-HPC is the most complete computer module definition ever. I expect extremely fast growth for scalable and compute-power-hungry embedded applications based on COM-HPC technology.”
The COM-HPC 1.15 specification effort is sponsored by ADLINK, congatec, and Kontron.
For more information on the COM-HPC FuSa specification, visit www.picmg.org/openstandards/com-hpc, or purchase the specification for $750 from www.picmg.org/product/com-hpc-module-base-specification-revision-1-15.
More on PICMG’s range of open, modular computing standards can be found at https://www.picmg.org.