Semiqa
SEMIQA provides energy-efficient AI processing by running computations in the analog domain. It uses ANNET, an Analog Neural Network architecture with aMEM, an ACU, a MU, and dMEM to execute neural operations without extra digital conversions. The system is ONNX-compatible, enabling integration with existing AI workflows. Target customers include edge computing, real-time analytics, voice recognition, marine, satellite, and automotive ADAS applications.
Semiqa
SEMIQA provides energy-efficient AI processing by running computations in the analog domain. It uses ANNET, an Analog Neural Network architecture with aMEM, an ACU, a MU, and dMEM to execute neural operations without extra digital conversions. The system is ONNX-compatible, enabling integration with existing AI workflows. Target customers include edge computing, real-time analytics, voice recognition, marine, satellite, and automotive ADAS applications.