From WikiChip
Difference between revisions of "microsoft/msfp8"
< microsoft

(msfp8)
(No difference)

Revision as of 22:59, 23 August 2019

MSFP8-MSFP11 (Microsoft floating-point 8-11) is a series of number encoding formats occupying 8 to 11 bits representing a floating-point number. MSFP8 is equivalent to a standard half-precision floating-point value with a truncated mantissa field. It is the 8-bit equivalent of the bfloat16 number format and is used for the very same purpose of hardware accelerating machine learning algorithms. MSFP was first proposed and implemented by Microsoft.