Contact us
中文
AI Booter-X8-MXM
Eight Myriad X MA2485 VPU
MXM form factor
Total 32Gb LP-DDR4 SDRAM on chips
Support OpenVINO™ toolkit
M/JPEG 4K at 60Hz encoder
H.264/H.265 4K at 30Hz encoder
Low power consumption
- Introduction
- Specification
- Order
- Download
AIBooster-X8-MXM MXM module integrated with Intel® Movidius™ VPUs drive the demanding workloads of modern computer vision and AI applications at ultra-low power. The dedicated Neural Compute Engine in Myriad X delivers more performance per Watt, and it helps AIBooster-X8-MXM achieves a perfect balance of power efficiency and high performance. Since AIBooster-X8-MXM is also designed for power constrained environments, it provides the ideal solution for device makers seeking to deploy advanced AI mobile applications at the edge.
| VPU Engine | Eight Myriad X MA2485 VPU Per VPU with - Up to 1 TOPS - 16 programmable 128-bit VLIW vector processors - CPUs 2x LEON 4 cores (RISC; SPARC V8) - On-chip accelerators 20+ image/vision processing accelerator Neural compute engine (DNN accelerator) - Neural network capability neural compute engine |
| Memory | Total 32Gb LP-DDR4 SDRAM on chips |
| Encoder | M/JPEG 4K at 60Hz encoder H.264/H.265 4K at 30Hz encoder |
| MXM (PCIe x4) | MXM (PCIe x4) |
| OS Support | Ubuntu 16.04.1/Kernel 4.10.0 Ubuntu 16.04.3/Kernel 4.14.20 Windows 10 Enterprise 64-bit |
| Feature Support | Intel® OpenVINO toolkit |
| Supported Network Topology | AlexNet, GoogleNet v1 & v2, Yolo Tiny* V1 & V2, Yolo V2, MobileNet-SSD, VGG-d , ResNet-18, Faster-RCNN |
| Dimensions | 113 x 82 mm |
| Power Consumption | < 50 W |
| Operating Temperature | -30°C to +85°C |
AIBooster-X8-MXM (P/N: 10VK00MXM00X0)
