Testers praise RTOS performance

In independent tests conducted by Dedicated Systems, OSE received high marks in a number of key categories, including performance, fault tolerance, memory protection and interprocess communications.

Enea has reported independent test results for its OSE real-time operating system (RTOS).

Dedicated Systems conducted the tests, evaluating a broad range of key RTOS performance, reliability and functionality parameters.

OSE received high marks in a number of key categories, including performance, fault tolerance, memory protection and interprocess communications.

"The Dedicated Systems test results provide independent validation of OSE's position as the pre-eminent RTOS for distributed, high-availability applications", said Mike Christofferson, Director of Product Management at Enea.

"OSE delivers a unique combination of high performance, fault tolerance and flexibility that greatly simplifies the design of scalable, portable, distributed applications spanning multiple processors, operating systems, and interconnects".

"From a study of the OSE architecture, it is clear that this operating system is especially well suited for distributed fail safe systems", said Martin Timmerman, CEO of Dedicated Systems.

OSE also posted impressive results in Dedicated Systems' performance benchmarks, taking full advantage of the Freescale PowerPC processor's compute resources, and outpacing competitive RTOSes such as QNX in the majority of the tests.

OSE outstripped the competition in a wide range of performance tests, including thread creation, deletion and switching, semaphore creation, deletion, acquisition, and release, and mutex acquisition and release.

In many of the tests, OSE outperformed competitive RTOSes by an order of magnitude.

OSE received especially high marks for its transparent, direct message technology, which provides the foundation for OSE's Linx interprocess communications services.

"OSE's asynchronous message passing system is advanced and very useful", said Timmerman.

"OSE employs a true micro-kernel architecture".

"Fundamental to that architecture is OSE's asynchronous messaging service for communications between distributed threads, which provides a versatile medium for implementing advanced capabilities such as routing".

The Dedicated Systems report called particular attention to the transparency of OSE's message passing services, which makes communications between processes independent of where they reside, the target processor, the operating system, or the interconnect.

This transparency simplifies distributed application development by enabling processes running on multiple processors and operating systems to communicate in a seamless fashion, as if they were running on a single processor under the same operating system.

It also enables developers to reconfigure, scale, and upgrade their system with minimal changes to the application code.

OSE also earned high marks for its memory protection facilities, which not only enhance reliability, but greatly simplify the development and debug process.

"A nice feature of OSE", noted Timmerman, "is that the calls used to partition and set protection levels between different memory regions are independent of whether or not an MMU is present".

"This makes it easy to debug your system using an MMU (providing real memory protection) and commercialise it on a cheaper platform (without an MMU) without changing any application code - well done".

OSE is a compact, pre-emptive, memory-protected RTOS optimised for communications applications that demand the utmost in reliability, security, and availability.

Occupying less than 100Kbyte of memory, OSE enhances reliability and availability by utilising a processor's memory management unit to build a firewall between kernel and application processes.

OSE further enhances reliability and availability by providing automatic failure detection and supervision (health monitoring) for system processes and applications.

Linx is a scalable, high-performance, transparent message passing IPC service for complex, heterogeneous distributed systems utilising multiple operating systems and processors.

Linx employs a lightweight connection protocol that greatly enhances performance relative to competitive IPC protocols.

Relative to TIPC, for example, Linx provides, on average, 25% lower latency and 20% higher throughput for intra-node IPC, and 10% lower latency and 25% higher throughput for internode IPC.

Linx is the only IPC technology that scales from DSPs and microcontrollers to 64bit CPUs.

Linx is also operating system (ie Linux, OSE, and other RTOSs) and media/interconnect independent (ie Gigabit Ethernet, RapidIO, PCI and shared memory).

This flexibility and transparency enables Linx to accommodate diverse network configurations, from a single processor on a single blade, to large networks with complex cluster topologies deployed on hundreds of processors in a multirack system.

Back to top