AAEON - UP AI Core X mPCIe, 30 mm Heat Sink

The UP AI Core X is a complete product line of neural network accelerator units for edge devices. Whether the automation you create is meant to identify and evaluate items in a shopping cart, alert airport security personnel to unattended luggage, or monitor traffic jams for an autonomous vehicle - with AI Core X, you can integrate the real-time computing power you need right where you need it.

Cloud computing is suitable for many things, but not for real-time edge applications with strict latency, bandwidth, security, and reliability constraints. Sometimes, the only solution is to accommodate data processing alongside data collection. This means cloud computing is not feasible. Fortunately, there is another option. With up to four trillion operations per second (TOPS) and a Neural Compute Engine that can deliver up to one TOPS with only a few watts of power, AI Core X is perfect for heavy tasks in any edge application.

The AI Core X is based on the recently introduced Intel Movidius Myriad X, a third-generation Vision Processing Unit (VPU) that is the first of its kind to include a Neural Compute Engine - a dedicated hardware accelerator for neural networks, trainable with industry-standard tools.

AI Core X is available in various form factors. You can choose between MiniCard/mPCIe, M.2 2230, M.2 2242, or M.2 2280.

It is compatible with UP Squared Boards and all SBCs with a miniPCIe interface.
Model: UP AI Core X
VPU: Intel Movidius Myriad X 2485
Number of VPUs: 1
Form Factor: mPCIe
Dimensions: 30 x 51 mm
Supported Frameworks: Caffe, TensorFlow
Memory: 4 GB LPDDR4
Cooling: Fanless Heat Sink
System Requirements: x86_64...