Ara240 16GB M.2 Module

[aman_12]

Which quantization precisions are officially supported in Ara240 16GB M.2 Module ?(INT4,8,16)

[yipingwang]

According to Ara240 Discrete Neural Processing Unit Data Sheet

Precision	Officially documented support
INT4	No documented support found
INT8	Supported
INT16	Supported
INT32	Supported, though not in your INT4/8/16 list