• monkyyy (5/5) Oct 12 80-bit floats are irrelevant; I rarely even use 64 bits.
• Richard (Rikki) Andrew Cattermole (3/11) Oct 12 Relevant for 16bit floats:

monkyyy <crazymonkyyy gmail.com> writes:

80-bit floats are irrelevant; I rarely even use 64 bits.

The industry is going **smaller** floats with ai

https://en.wikipedia.org/wiki/Bfloat16_floating-point_format

I dont know about the google vs ieee debate, but a 16 bit float 
would probably be more useful then the talk of a 80 bit float.

"Richard (Rikki) Andrew Cattermole" <richard cattermole.co.nz> writes:

On 13/10/2025 7:40 AM, monkyyy wrote:
 80-bit floats are irrelevant; I rarely even use 64 bits.
 
 The industry is going **smaller** floats with ai
 
 https://en.wikipedia.org/wiki/Bfloat16_floating-point_format
 
 I dont know about the google vs ieee debate, but a 16 bit float would 
 probably be more useful then the talk of a 80 bit float.

Relevant for 16bit floats: 
https://forum.dlang.org/post/vm4v9c$ofk$1 digitalmars.com