__int128_t and __float128 in GNU

Usage of __int128_t:

• supports really big intermediate numbers, from -2^128 + 1 to 2^128 - 1
• it has 128 bits

Usage of __float128 :

• can go up to 20 (or 30, if I remember correctly) decimal places
• it has 128 bits

Problems of __int128_t and __float128:

• can not print a integer of __int128_t
• can not read a integer of __int128_t

Comparing __int128_t and long long

• __int128_t consumes 3 - 5 times more memory and time then long long
• __int128_t consumes 3 - 5 times more time and time then long long

Comparing __float128 and long double :

• __float128 consumes 3 - 5 times more memory and time then long double
• __float128 consumes 3 - 5 times more time and time then long double

A little exercice before the end :

What is the difference __m128d and __int128 ?

Comment below and thanks for reading !