https://github.com/ifd3f/bigfloat
High-performance, high-precision 80-bit floating point library
https://github.com/ifd3f/bigfloat
bigfloat cpp intrinsics
Last synced: 7 months ago
JSON representation
High-performance, high-precision 80-bit floating point library
- Host: GitHub
- URL: https://github.com/ifd3f/bigfloat
- Owner: ifd3f
- Created: 2020-08-19T06:33:21.000Z (about 5 years ago)
- Default Branch: master
- Last Pushed: 2020-12-08T02:18:01.000Z (almost 5 years ago)
- Last Synced: 2025-01-11T03:48:52.466Z (9 months ago)
- Topics: bigfloat, cpp, intrinsics
- Language: C++
- Homepage:
- Size: 326 KB
- Stars: 0
- Watchers: 2
- Forks: 0
- Open Issues: 3
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Bigfloat Library 
High-performance, high-precision floating point library
## Useful Links
- [Automated bigfloat benchmark times](https://astrid.tech/bigfloat/dev/bench/)
- [What Is Astrid Working On?](https://github.com/Plenglin/bigfloat/projects/1) (tm)
## Useful References
- [Intel Intrinsics Guide](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#techs=MMX,SSE,SSE2,SSE3,SSSE3,SSE4_1,SSE4_2,AVX,AVX2,FMA), filtered with extensions that exist on most gamers' computers based on the [Steam Hardware Survey](https://store.steampowered.com/hwsurvey)
- [C++ Optimization Strategies and Techniques (slightly outdated)](https://www.tantalon.com/pete/cppopt/main.htm)
- [Coding Practices which enable the compiler/optimizer to make a faster program (StackOverflow)](https://stackoverflow.com/q/2074099/12947037)
- [What Every Computer Scientist Should Know About Floating-Point Arithmetic](https://docs.oracle.com/cd/E19957-01/806-3568/ncg_goldberg.html)
## Packed Format
Likely would be used when sending over networks and such.
| Field | Start | End | Count | Notes |
|----------:|------:|----:|------:|:------|
| Exponent | 79 | 64 | 16 | Bias of `1023` (same as IEEE double precision)
| Sign | 63 | 63 | 1 | `0` is positive, `1` is negative
| Mantissa | 62 | 0 | 63 | Technically 64 bits, with a `1` for the MSB
| **Total** | 79 | 0 | 80 | This is 10 bytes, or 2 more than a double.