51
edits
SIMD/Types/float64x2: Difference between revisions
Uses for float64x2
Maikmerten (talk | contribs) (Preliminary definition for float64x2) |
Maikmerten (talk | contribs) (Uses for float64x2) |
||
| Line 1: | Line 1: | ||
The data type float64x2 represents 2 double-precision (64 bit) floating point values, packed into one 128 bit SIMD register. | The data type float64x2 represents 2 double-precision (64 bit) floating point values, packed into one 128 bit SIMD register. | ||
== Uses for float64x2 == | |||
* Full-precision implementations of, e.g., JPEG's [[SIMD/Uses/DCT |DCT]] may benefit from float64x2. | |||