MemryX
MemryX provides low-power, high-performance edge AI acceleration to run real-time neural network inference on devices. The company implements proprietary compute-at-memory technology and a dataflow architecture in its MX3 family accelerators, including MX3 Edge and MX3 M.2. Their fabless semiconductor products are compatible with host processors such as x86, Arm, and RISC-V and support AI frameworks including TensorFlow, PyTorch, and ONNX. MemryX targets applications in video management, industrial automation, smart vision, autonomous systems, and other edge-focused industries and offers developer resources for model compilation and deployment.
MemryX
MemryX provides low-power, high-performance edge AI acceleration to run real-time neural network inference on devices. The company implements proprietary compute-at-memory technology and a dataflow architecture in its MX3 family accelerators, including MX3 Edge and MX3 M.2. Their fabless semiconductor products are compatible with host processors such as x86, Arm, and RISC-V and support AI frameworks including TensorFlow, PyTorch, and ONNX. MemryX targets applications in video management, industrial automation, smart vision, autonomous systems, and other edge-focused industries and offers developer resources for model compilation and deployment.