News at a Glance...
Nuremberg, Germany, Embedded World, February 26, 2008 QNX Software Systems today published the source code for QNX® Neutrino® adaptive partitioning, an innovative software technology that enhances system security, simplifies software integration, and brings the benefits of time and memory partitioning to multi-core designs.
QNX Neutrino adaptive partitioning allows developers to build secure compartments around their applications and assign a guaranteed amount of memory and CPU time to each compartment. As a result, applications can meet their realtime deadlines and deliver critical services in almost any scenario, even if malware or a denial of service (DoS) attack attempts to monopolize system resources.
QNX Neutrino adaptive partitioning uses a flexible, patent-pending scheduler that makes the best use of fluctuating processor demands. If the applications in a partition do not use all of their budgeted CPU cycles, the scheduler will dynamically reallocate those cycles to partitions that can benefit from the extra processing time. This approach is far more efficient than traditional partitioning solutions, whose fixed schedulers always waste these spare CPU cycles, even when other partitions could use them.
Conventional partitioning solutions protect applications, but compromise on performance. QNX adaptive partitioning, on the other hand, provides the protection of time partitioning while permitting full use of processor capacity, said Kerry Johnson, director of product management at QNX Software Systems. No other solution offers this combination of service availability and performance.
First partitioning solution for multi-core processors
QNX adaptive partitioning also offers the first secure memory and CPU partitioning capability for multi-core systems. For instance, a partition can span multiple cores, allowing partitioned applications to leverage the full power of multi-core processors.
In comparison, conventional partitioning schemes are limited to uniprocessor designs. As a result, developers who wish to protect their systems from untrusted applications or rogue attacks must attempt to partition each core separately, usually by proliferating multiple copies of the OS on each core.
Faster software integration
Besides offering greater security, QNX adaptive partitioning can dramatically reduce the time required to integrate complex software applications. Often, problems show up during the system integration phase, when competition for memory or CPU cycles among applications causes task starvation, a condition that can take weeks or months to resolve. QNX adaptive partitioning prevents task starvation by allowing developers to verify whether each subsystem is working within its memory and/or CPU budget. Problems are caught early, long before the integration phase, when they cost least to correct.
Available for download
QNX is releasing the source code to QNX adaptive partitioning as part of its new hybrid software model. To download the source, QNX community members can visit Foundry27 (www.foundry27.com), the QNX developer portal.
About QNX Software Systems
QNX Software Systems, a Harman International company (NYSE: HAR), is the industry leader in realtime, embedded OS technology. The component-based architectures of the QNX® Neutrino® RTOS, QNX Momentics® development suite, and QNX Aviage middleware together provide the industrys most reliable and scalable framework for building innovative, high-performance embedded systems. Global leaders such as Cisco, Daimler, General Electric, Lockheed Martin, and Siemens depend on QNX technology for network routers, medical instruments, vehicle telematics units, security and defense systems, industrial robotics, and other mission- or life-critical applications. Founded in 1980, QNX Software Systems is headquartered in Ottawa, Canada, and distributes products in over 100 countries worldwide.
QNX Software Systems
+1 613 591-0931
QNX, Aviage, Momentics, and Neutrino are trademarks of QNX Software Systems GmbH & Co. KG, registered in certain jurisdictions, and are used under license. All other trademarks and trade names belong to their respective owners.