The Myriad X is the latest generation in the Movidius VPU line created specifically for embedded visual intelligence and for inferences. “Allowing the devices to have a visual intelligence like that of a human represents the first leap forward in computing. With Myriad X we are redefining what a VPU means in terms of delivering as much power as possible for artificial intelligence and computer vision and all with the thermal and power limitations of modern wireless devices,” said Remi El-Ouazzane, the vice president and general manager of Movidius at Intel New Technology Group. With the presentation of the Neural Compute Engine, the Myriad X architecture will be able to provide 1 TOPS of computer performance in inferences from deep neural networks. It is the first Chip System (SoC) that incorporates a specialized Neural Compute Engine to accelerate the inferences of deep learning at the local level. However, now if we talk about the Neural Compute Engine, then Myriad X combines visual processing, images and deep learning inferences in real-time with 128-bit programmable VLIW vector processors, more configurable MIPI lanes, improved acceleration for vision and 5 MB Homogeneous on-chip memory that allows up to 450GB per second of internal bandwidth, minimizing latency and reducing power consumption due to reduced off-chip data transfer. The VPU Movidius Myriad X is a unit designed to provide more autonomous performance in a wide range of products, including drones, robotics, smart cameras and virtual reality. So, what do you think about this extraordinary and high-tech SoC from Intel? Simply share your views and thoughts in the comment section below.
Δ
