With 30 times the machine learning throughput as compared to a CPU alone, the functionality of the NPU now puts the onerous on other chip manufacturers to process deep, more inferred content.
Industrial and IoT streams, designs, and concepts will see a surge in ability and an enhanced time to market, as well as reduced bills-of-materials. For in-home automation and smart cities, a more "thoughtful" machine is on the horizon. The entire process should also become easier for the average user in terms of speech-to-action recognition. In the industrial space, more in-depth instruction sets, with higher levels of inference will result in safer, more secure, higher-order tasks. The smart factory could also gain a more robust and precise performance set, with an eye on how such a direct impact on the factory and warehouse systems will play out in real time. Such change can see a "calculated" plant floor - that changes with product parameters rather than needing to retool. This could be a huge advancement in how products are manufactured and stored.
Next week, tune in for the IoT Technology showcase, where we examine the latest advances and products. |