Floating-point arithmetic

• IEEE 754 - IEEE Standard for Floating-Point Arithmetic
  • https://0.30000000000000004.com/ ↗
  • NaN
  • signed zeros
  • infinities
• bfloat16 ↗. The bfloat16 format was developed by Google Brain, an artificial intelligence research group at Google
• posit aka Unums (universal numbers) 3

See also:

• http://www.johngustafson.net/pdfs/BeatingFloatingPoint.pdf ↗
• https://ieeexplore.ieee.org/document/9446981 ↗
• https://ieeexplore.ieee.org/document/9764927 ↗
• https://github.com/stillwater-sc/universal ↗