Sensor processing has entered a new arena with the multitude of high performance and/or low cost sensors. From advanced military sensors (i.e., electro-optical, infrared, LIDAR, radar, sonar, etc.) to mobile devices with cameras, GPS, gyroscopes and accelerometers, sensor fusion has created a whole new set of applications and augmented reality that greatly benefits the user. In many cases, the sensor platform is space and power constrained. An additional key factor is that to achieve the maximum benefit, the sensing and processing needs to be taking place almost all the time. For equipment that runs on a battery, this can be a non-starter…until now. HyperX processors can be used as a co-processor to a GPP or application processor to off-load the sensor processing and augmented reality algorithms. Since each HyperX core is very low power and some algorithms may only require a few cores at a time, this is an extremely power efficient means of doing persistent sensor processing. Another great example is voice or gesture recognition. One or more cores may need to be always on in order to listen or look for instructions, while the rest of the cores can be in standby mode ready to power up in the event that an instruction is received. Using a GPP for tasks such as this is highly power inefficient.
Other examples include medical imaging. Right now, the time it takes to process an extremely high resolution image (i.e., X-ray, CT scan, PET, MRI, etc.) is far too long and results in inefficiencies and a delayed diagnosis. Using HyperX processors can significantly reduce the time for processing of medical images, resulting in more timely diagnoses and thereby increase patient throughput.
Contact us to discuss how the HyperX processor’s high performance, low power sensor processing can give you and your products a competitive advantage in real-time.