Table 1.

Performance analysis of various architectures for GHA training.

Architectures	Adders	Multipliers	Registers	Latency
proposed Architecture	O(q)	O(q)	O(mp)	max(q, 2bp + s)
[18]	O(mp)	O(mp)	O(mp)	max(q + 1,p + 1)
[19]	O(p)	O(p)	O(mp)	3m+p−1