George A. Constantinides

MASE simplifies and speeds up ML accelerator development for modern machine learning models.

A new method enhances deep learning models on FPGAs for faster processing.

The eGPU enhances parallel processing in FPGAs with flexible architecture.

A new technique for optimizing large language models while maintaining performance.

A look at the performance of soft GPGPUs versus IP cores in FFT tasks.

ROVER simplifies RTL design by automating complex optimization processes.

A new standard improves neural network speed and energy efficiency in computing.

A new method optimizes lookup tables using 'don't care' conditions.