Deploying Edge AI: From Concept to Production (Download)
It seems that the edge has become the topic du jour lately, but it’s not as new as most of us think. But before we get into the weeds with this paper, let’s take a trip down memory lane; back to 1978 to see how TI engineers developed the first voice-synthesized DSP embedded device, the Speak and Spell, and how it led to today’s advanced embedded devices.
From the Speak and Spell to the present day, we have developed advanced AI-enabled microcontrollers. These next-generation devices will revolutionize the capabilities of edge networks. We now have edge devices that offer heterogeneous compute, multimodal on-device inference, and robotics-grade real-time performance. These devices are an elegant solution that will conquer many of the critical constraints that have been holding back the evolution of the edge.
