Motorola Mobility’s Moto Z2 Force and Moto X4 are the first two smartphones to utilize Qualcomm’s Neural Processing Engine, a software development kit (SDK) launched in mid-summer as a method of helping developers run neural network models trained with the help of TensorFlow or Caffee(2) on Snapdragon chips, regardless of whether they’re seeking to utilize the CPU, GPU, or the DSP of the company’s devices, or some combination of the thereof. The San Diego, California-based tech giant officially confirmed the presence of its technology in Motorola’s recently released Android smartphones on Tuesday, suggesting that many more original equipment manufacturers will soon follow suit and start leveraging the Neural Processing Engine in order to implement machine learning and general artificial intelligence technologies into their offerings in a more efficient manner.
Qualcomm’s solution is compatible with all of the company’s contemporary chips, having been implemented by Motorola into both the Snapdragon 630-powered Moto X4 and the Moto Z2 Force which comes with the firm’s premium Snapdragon 835. Landmark Detection is the first user-facing feature of Motorola’s latest mobile offerings enabled by the Neural Processing Engine, Qualcomm said, adding that the solution behind this self-explanatory name allows people to identify various landmarks through the camera lenses installed on their devices. Refer to the GIF animation below for a short demo of the newly implemented technology.
As AI services are steadily gaining popularity, more manufacturers are expected to start leveraging technologies specifically dedicated to managing neural networks, an inherent component of machine learning. Qualcomm’s Neural Processing Engine technically isn’t a hardware solution, though it’s still tied to the company’s latest mobile chips and isn’t directly transferable to another SoC category. However, the technology wasn’t designed exclusively for smartphones and tablets, with Qualcomm predicting that a variety of industries ranging from healthcare and Internet of Things to manufacturing and self-driving vehicles could take advantage of its solution once AI applications become more ubiquitous. The company didn’t specifically name any other partners that are presently in the process of leveraging the Neural Processing Engine but more such entities should be revealed in the near future as Qualcomm’s AI push keeps intensifying.