Corporate Blogs

Low power graphics for high power applications

By Dan Hulme - Last updated: Thursday, July 14, 2016

Graphics-processing-unitWhen you think of mobile GPUs (graphics processing units), you’re probably thinking of smartphones. Graphically intensive mobile games and increasing screen densities have driven the demand for low-power GPUs and video encode/decode hardware. With the availability of this power, demands are being driven further with new applications such as mobile VR (virtual reality). For example, Google announced last month that their Tango project to help smartphones understand physical space and motion has found its way into its first consumer hardware. The new kinds of AR (augmented reality) and vision applications are powered not only by extra sensing capabilities, but Qualcomm’s Snapdragon 652 system-on-chip with a beefy Adreno 510 GPU.

The smartphone industry isn’t the only one to use these new graphics capabilities to improve their products. A decade ago, the most graphics-intensive device most people carried around with them was their in-car satnav. Faced with stiff competition from the likes of Google Maps, the personal navigation device sector uses mobile GPUs to give a more sophisticated user interface, showing junction layouts and buildings in 3D to make it easier to relate the display with the view through the windscreen. The whole automotive industry is moving in the same direction. Some manufacturers are already demonstrating all-software dashboards, with all the dials and indicator being drawn on one big screen, in the automotive equivalent of the glass cockpit.

In the office, printer manufacturers are adopting mobile GPUs to make printers easier to use. New printers can show a step-by-step 3D animation of how to fix a paper jam or change the toner cartridge, to reduce the rate of maintenance calls and total cost of ownership. In normal use, the GPU stays in a low-power mode, helping the device meet its energy efficiency targets. Even SCADA systems in big distribution centres are showing visualisations of throughput on the real-world layout of the system, so operators can quickly identify bottlenecks or unused capacity. Even where power is not a constraint, mobile GPUs on systems-on-chip (SoCs) are often easier to integrate, quicker to bring to market, and orders of magnitude cheaper per unit than bigger GPUs intended for desktop PCs.

Often the companies wanting to add 3D graphics to their products have no in-house expertise about 3D rendering, programming GPUs, or integrating SoCs with GPUs, so they come to Cambridge Consultants. We can fill a gap in their expertise, or we can bring together experts in all the needed disciplines to design a whole product. Just as consumer products have driven the demand for mobile GPUs, they’ve also helped Cambridge Consultants to build up a wealth of experience in making human-machine interfaces user-friendly and visually attractive: experience we can use just as well on your next industrial automation, construction, or office product.


Media


AuthorDan Hulme


Subscribe to blog feeds:


Recent Posts


Posts by categories