https://github.com/fast-pack/simdcompressionandintersection
A C++ library to compress and intersect sorted lists of integers using SIMD instructions
https://github.com/fast-pack/simdcompressionandintersection
algorithms compression integer-compression intersection simd simd-instructions
Last synced: about 1 month ago
JSON representation
A C++ library to compress and intersect sorted lists of integers using SIMD instructions
- Host: GitHub
- URL: https://github.com/fast-pack/simdcompressionandintersection
- Owner: fast-pack
- License: apache-2.0
- Created: 2013-12-16T18:51:31.000Z (over 12 years ago)
- Default Branch: master
- Last Pushed: 2023-07-13T18:31:57.000Z (about 3 years ago)
- Last Synced: 2025-04-02T23:58:01.268Z (over 1 year ago)
- Topics: algorithms, compression, integer-compression, intersection, simd, simd-instructions
- Language: C++
- Size: 1.33 MB
- Stars: 431
- Watchers: 29
- Forks: 59
- Open Issues: 4
-
Metadata Files:
- Readme: README.md
- Changelog: CHANGELOG
- License: LICENSE
Awesome Lists containing this project
README
SIMDCompressionAndIntersection
======================
[](https://github.com/lemire/SIMDCompressionAndIntersection/actions/workflows/ubuntu22.yml)
[](https://github.com/lemire/SIMDCompressionAndIntersection/actions/workflows/visualstudio.yml)
As the name suggests, this is a C/C++ library for fast
compression and intersection of lists of sorted integers using
SIMD instructions. The library focuses on innovative techniques
and very fast schemes, with particular attention to differential
coding. It introduces new SIMD intersections schemes such as
SIMD Galloping.
This library can decode at least 4 billions of compressed integers per second on most
desktop or laptop processors. That is, it can decompress data at a rate of 15 GB/s.
This is significantly faster than generic codecs like gzip, LZO, Snappy or LZ4.
Authors: Leonid Boystov, Nathan Kurz, Daniel Lemire,
Owen Kaser, Andrew Consroe, Shlomi Vaknin, Christoph Rupp, Bradley Grainger, and others.
Documentation
-----------------------
* Daniel Lemire, Nathan Kurz, Christoph Rupp, Stream VByte: Faster Byte-Oriented Integer Compression, Information Processing Letters 130, 2018 https://arxiv.org/abs/1709.08990
* Daniel Lemire, Leonid Boytsov, Nathan Kurz, SIMD Compression and the Intersection of Sorted Integers, Software Practice & Experience 46 (6), 2016 http://arxiv.org/abs/1401.6399
* Daniel Lemire and Leonid Boytsov, Decoding billions of integers per second through vectorization, Software Practice & Experience 45 (1), 2015. http://arxiv.org/abs/1209.2137 http://onlinelibrary.wiley.com/doi/10.1002/spe.2203/abstract
* Jeff Plaisance, Nathan Kurz, Daniel Lemire, Vectorized VByte Decoding, International Symposium on Web Algorithms 2015, 2015. http://arxiv.org/abs/1503.07387
* Wayne Xin Zhao, Xudong Zhang, Daniel Lemire, Dongdong Shan, Jian-Yun Nie, Hongfei Yan, Ji-Rong Wen, A General SIMD-based Approach to Accelerating Compression Algorithms, ACM Transactions on Information Systems 33 (3), 2015. http://arxiv.org/abs/1502.01916
This work has also inspired other work such as...
* P. Damme, D. Habich, J. Hildebrandt, W. Lehner, Lightweight Data Compression Algorithms: An Experimental Survey (Experiments and Analyses), EDBT 2017 http://openproceedings.org/2017/conf/edbt/paper-146.pdf
* P. Damme, D. Habich, J. Hildebrandt, W. Lehner, Insights into the Comparative Evaluation of Lightweight Data Compression Algorithms, EDBT 2017 http://openproceedings.org/2017/conf/edbt/paper-414.pdf
* T. D. Wu, Bitpacking techniques for indexing genomes: I. Hash tables, Algorithms for Molecular Biology 11 (5), 2016. http://almob.biomedcentral.com/articles/10.1186/s13015-016-0069-5
* Jianguo Wang, Chunbin Lin, Yannis Papakonstantinou, Steven Swanson, An Experimental Study of Bitmap Compression vs. Inverted List Compression, SIGMOD 2017 http://db.ucsd.edu/wp-content/uploads/2017/03/sidm338-wangA.pdf
Simple demo
------------------------
Check out example.cpp
You can run it like so (e.g., under Linux or macOS):
Usage (Linux, macOS and similar systems)
------------------------
```
make
./unit
```
A static library file is built as `libSIMDCompressionAndIntersection.a` which you can use in your
own projects along with our header files located in the `include` subdirectory.
You may also build and run our example:
```
make example
./example
```
To run tests, you can do
```
./testcodecs
```
(follow the instructions)
Building and installing with CMake
------------------------
A CMake build is also provided. It works on x86/x64 (SSE/AVX) and on 64-bit ARM
(the SSE intrinsics are mapped to ARM NEON), and it builds the same library and
tests as the Makefile.
```
cmake -S . -B build
cmake --build build
ctest --test-dir build # runs the unit tests
```
To install the library, headers and a CMake package configuration:
```
cmake -S . -B build -DCMAKE_INSTALL_PREFIX=/your/prefix -DSIMDCOMP_BUILD_TESTS=OFF
cmake --build build
cmake --install build
```
Downstream CMake projects can then locate it with `find_package` and link the
imported target:
```cmake
find_package(SIMDCompressionAndIntersection REQUIRED)
target_link_libraries(yourapp PRIVATE
SIMDCompressionAndIntersection::SIMDCompressionAndIntersection)
```
The installed headers live under `/include/SIMDCompressionAndIntersection`
and are added to your include path by the imported target, so `#include
` works directly. Useful options: `-DSIMDCOMP_BUILD_TESTS=OFF`
to skip the tests/benchmarks and `-DSIMDCOMP_INSTALL=OFF` to disable install
rules (handy when consuming the project via `add_subdirectory`).
Usage (Windows users)
------------------------
Windows users wishing to build using Visual Studio should go into a Developer Powershell, which is
accessible through the menus in the Visual Studio interface, and run the following from the
directory of the project:
```
nmake -f .\makefile.vc
.\example.exe
.\unit.exe
```
Under Windows, the static library is built as the file `simdcomp_a.lib` which you can use in your
own projects, along with our header files located in the `include` subdirectory.
For a simple C library
----------------------
This library is a C++ research library. For something simpler,
written in C, see:
https://github.com/lemire/simdcomp
Comparison with the FastPFOR C++ library
-----------------------------------------
The FastPFOR C++ Library available at https://github.com/lemire/FastPFor
implements some of the same compression schemes except that
it is not optimized for the compression of sorted lists of integers.
Other recommended libraries
---------------------------------
* libvbyte: A fast implementation for varbyte 32bit/64bit integer compression https://github.com/cruppstahl/libvbyte
* TurboPFor is a C library that offers lots of interesting optimizations. Well worth checking! (GPL license) https://github.com/powturbo/TurboPFor
* Oroch is a C++ library that offers a usable API (MIT license) https://github.com/ademakov/Oroch
* CSharpFastPFOR: A C# integer compression library https://github.com/Genbox/CSharpFastPFOR
* JavaFastPFOR: A java integer compression library https://github.com/lemire/JavaFastPFOR
* Encoding: Integer Compression Libraries for Go https://github.com/zhenjl/encoding
* FrameOfReference is a C++ library dedicated to frame-of-reference (FOR) compression: https://github.com/lemire/FrameOfReference
Licensing
------------------------
Apache License, Version 2.0
As far as the authors know, this work is patent-free.
Requirements
------------------------
On x86/x64, a CPU (AMD or Intel) with support for SSE2 (Pentium 4 or better) is
required while a CPU with SSE 4.1* (Penryn [2007] processors or better) is
recommended.
On 64-bit ARM (AArch64, e.g. Apple Silicon and ARM servers), the SSE intrinsics
are mapped to ARM NEON via `include/neon_sse.h`, so no x86 hardware is needed.
NEON is baseline on AArch64, so no special compiler flag is required.
A recent GCC (4.7 or better), Clang, Intel or Visual C++ compiler.
On x86, a processor supporting AVX (Intel or AMD) is assumed by the default
makefile (but AVX is not required, see below).
Tested on Linux, MacOS and Windows, on both x64 and ARM64. It should be portable
to other platforms.
*- On x86, the default makefile might assume AVX support, but AVX is not
required. For GCC compilers, you might need the -msse2 flag, but you will not
need the -mavx flag. On ARM64 the makefile automatically drops the x86 `-mavx`
flag.
For advanced benchmarking, please see
advancedbenchmarking/README.md
where there is additional information as well as links to real data sets.
Acknowledgement
------------------------
Thanks to Kelly Sommers for useful feedback.
This work was supported by NSERC grant number 26143.