Edge AI Acceleration Modules

The EAI series provides low power plug-in AI modules and GPU cards to help accelerate deep learning inference on the edge

Hailo-8™ M.2 AI Acceleration Module

26 TOPS AI Computing delivered by Hailo-8 VPU
M.2 2242 w/ optional 2280 extensions
-20~65°C operating temp. w/ optional heatsink
Typical 2.5W Low power consumption
Rich AI Dev. Toolkits including Model Library, Model Converter, Runtime SDK Integration

MXM 3.1 Type A Intel Arc A370M Embedded GPU Card with DP 1.4a

Intel Arc A370M with 1550 MHz Base Clock, GDDR6 4GB 64 bit Memory, up to 4.198 TFLOPS
TGP 40W Low consumption with optional thermal kit, support 0-60°C working temperature
Reserved DP 1.4a x4 displays up to 8K UHD resolution
Advantech Edge AI SDK and Intel OpenVINO support
DirectX12, OpenGL 4.6, OpenCL 3.0, H.264/H.265 (HEVC)/AV1 hardware encode/decode support

PCIe x16 Intel Arc A370M Embedded GPU Card with 2x HDMI 2.0b + 2x DP 1.4a

Intel Arc A370M with 1550 MHz Base Clock, GDDR6 4GB 64 bit Memory, up to 4.198 TFLOPS
High Resolution 8K DP 1.4a x2 + 4K HDMI 2.0b x2, up to 4 Displays
0-60°C working temperature support w/ auto smart fan
Advantech Edge AI SDK and OpenVINO support
DirectX12, OpenGL 4.6, OpenCL 3.0, AV1/H.264/H.265 (HEVC) hardware encode/decode support

PCIe x16 Intel Arc A380E Embedded GPU Card with 4x DP 1.4a

Intel Arc A380E with 2000 MHz Base Clock, GDDR6 6GB 96 bit Memory, up to 5.018 TFLOPS
4x DP1.4a up to 8K UHD resolution
0-60°C working temperature support w/ auto smart fan
Advantech Edge AI SDK and Intel OpenVINO support
AV1, DirectX12, OpenGL 4.6, OpenCL 3.0, H.264/H.265 (HEVC) hardware encode/decode support

Hailo-8™ PCIe AI Acceleration Card

52 TOPS delivered by 2x Hailo-8 VPU
PCIe x16, HHHL design with single slot bracket
-40~70°C operating temp. support
Optimized cooler design with smart fan
Rich AI Development Toolkits for Runtime SDK integration, Model build, Model Convert, and Pre-trained AI application