Fully matching application performance challenges and avoiding unwanted compromises
FRANCE – 3 December 2019 – Addressing the growing need to maximize performance, while keeping power budgets in check, Teledyne e2v has constructed a sophisticated methodology relating to its portfolio of microprocessors. Via this methodology the power consumption characteristics of QorIQ® T-Series and P-Series (Power Architecture®), as well as the ARM® based Layerscape devices supplied to its customers’ systems can be studied in detail, rather than relying only on datasheet information.
In a recent safety-critical aerospace customer project, Teledyne e2v was able to deliver T1042 multi-core microprocessors for which the power consumption was almost 50% lower on the final application — thereby dramatically reducing the constraints placed on the overall system.
The high levels of functionality needed by modern embedded designs means that engineers want to fit in as much processing capacity as possible. However, in many use cases (particularly in the avionics, military and space sectors) stringent power constraints must also be adhered to.
Published parametric data concerning power consumption will have certain margins included (based on device variation and worst case operating conditions). Often these will force engineers to be overly cautious — specifying microprocessors that offer less performance (or running devices at lower frequencies than they can actually attain), to ensure they remain within the maximum power consumption that their system will support.
Drawing on unrivalled knowledge in providing the market with high-reliability processing solutions, plus direct access to the native test programs from NXP (its original microprocessor manufacturing partner), Teledyne e2v is in a unique position when it comes to understanding devices’ power characteristics.
By combining all this with accurate replications of customers’ use cases and the actual application conditions that the microprocessors will be subjected to, comprehensive static and dynamic power consumption profiles can be obtained. Consequently, the company goes beyond the margins stated on datasheets and customers can source fully warranted power-optimized hardware.
By procuring only the most power efficient devices for the dedicated application, engineers can confidently boost their system designs’ operational performance figures. There are also significant benefits in terms of building additional headroom into systems (so they are future proofed for incorporating further features and functionality at a later stage) and the potential extending of working lifespans.
“When dealing with high-performance multi-core processors, there can be notable variation in static power from one device to another, while clocking frequency and process load will affect the dynamic power. Then there is the operational temperature to consider too,” explains Thomas Guillemain, Marketing & Business Development Manager at Teledyne e2v. “Through the visibility offered by this new service, we are now enabling far better optimization of our 32-bit and 64-bit microprocessors, so that maximum performance can be derived while keeping within the associated power consumption confines.”
