Created
April 15, 2013 14:59
-
-
Save cabo/5388730 to your computer and use it in GitHub Desktop.
Convert an IEEE 754 Half-precision (16-bit) float into a native float in 11 instructions and 1 branch.
Portable to IEEE 754 Single-precision implementations (i.e., everything post-VAX).
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| float decode_half(int half) { | |
| union { uint32_t u; float f; } val; | |
| val.u = (half & 0x7fff) << 13 | (half & 0x8000) << 16; | |
| if ((half & 0x7c00) != 0x7c00) | |
| return val.f * 0x1p112; | |
| val.u |= 0x7f800000; | |
| return val.f; | |
| } |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
please note that this doesn't work on platforms where sizeof(int) < 4, because the shifts will overflow.
also, it involves a floating point multiplication, so may not be the optimal for platforms without fpu.