From: minforth@gmx.net   
      
   Am 10.07.2025 um 00:10 schrieb Paul Rubin:   
   > dxf  writes:   
   >> As for SSE2 it wouldn't exist if industry didn't consider   
   >> double-precision adequate.   
   >   
   > SSE2 is/was first and foremost a vectorizing extension, and it has been   
   > superseded quite a few times, indicating it was never all that   
   > adequate.  I don't know whether any of its successors support extended   
   > precision though.   
      
   You don't need 64-bit doubles for signal or image processing.   
   Most vector/matrix operations on streaming data don't require   
   them either. Whether SSE2 is adequate or not to handle such data   
   depends on the application. "Industry" can manage well with 32-bit   
   floats or even smaller with non-standard number formats.   
      
   The AVX extension introduced YMM registers that can do simultaneous   
   math on four 64-bit double-precision floating-point numbers.   
   The intended application domain was scientific computing.   
      
   The determining factors are data througput and storage space.   
   Today, with GPUs, speed and power consumption, driven by AI.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)