From 62863bc6fe99cad68cee2234bf19eab171e49529 Mon Sep 17 00:00:00 2001 From: powturbo Date: Sun, 21 Jun 2015 18:58:48 +0200 Subject: [PATCH] =?UTF-8?q?:new:=20Java+64=20bits=20lists=20for=20BitPacki?= =?UTF-8?q?ng,=20VSimple,=20VByte,=20Elias=20Fano,=E2=80=A6?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- README.md | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/README.md b/README.md index a87d537..381a39c 100644 --- a/README.md +++ b/README.md @@ -86,10 +86,11 @@ TurboPForDA,TurboForDA: Direct Access is normally used when accessing individual | 3.820.190.182| 15.97| 5.11| 118.81| 650.21|Simple16| | 3.958.888.197| 16.55| 5.30| 279.19| 618.60|[lz4](#lz4) 64K| | 4.521.326.518| 18.90| 6.05| 209.17| 824.26|Simple-8b| -| 4.683.323.301| 19.58| 6.27| 828.25| 991.34|**TurboVbyte**| +| 4.683.323.301| 19.58| 6.27| 828.97|1007.44|**TurboVbyte**| | 4.953.768.342| 20.71| 6.63|**1766.05**|**1943.87**|**TurboPackV**| | 4.953.768.342| 20.71| 6.63|1419.35|1512.86|**TurboPack**| | 5.203.353.057| 21.75| 6.96|1560.34|1806.60|SIMDPackD1 FPF| +| 6.074.995.117| 25.40| 8.13| 494.70| 729.97| blosc lz4 64K| | 6.221.886.390| 26.01| 8.32|1666.76|1737.72|**TurboFor**| | 6.221.886.390| 26.01| 8.32|1660.52| 565.25|**TurboForDA**| | 6.699.519.000| 28.01| 8.96| 472.01| 495.12|Vbyte FPF| @@ -162,7 +163,7 @@ using [900.000 multicore servers](https://www.cloudyn.com/blog/10-facts-didnt-kn - Data file Benchmark (file from [DocId data set](#DocId data set)) - ./icbench -c1 gov2.sortedVarintG8IU + ./icbench -c1 gov2.sorted ##### - Intersections: @@ -256,7 +257,7 @@ header files to use with documentation:
### References: + [FastPFor](https://github.com/lemire/FastPFor) + [Simdcomp](https://github.com/lemire/simdcomp): SIMDPack FPF, Vbyte FPF, VarintG8IU - + [Optimized Pfor-delta compression code](http://jinruhe.com): OptPFD/OptP4, Simple16 + + [Optimized Pfor-delta compression code](http://jinruhe.com): OptPFD/OptP4, Simple16 (limited to 28 bits integers) + [MaskedVByte](http://maskedvbyte.org/). See also: [Vectorized VByte Decoding](http://engineering.indeed.com/blog/2015/03/vectorized-vbyte-decoding-high-performance-vector-instructions/) + [Index Compression Using 64-Bit Words](http://people.eng.unimelb.edu.au/ammoffat/abstracts/am10spe.html): Simple-8b (speed optimized version tested) + [libfor](https://github.com/cruppstahl/for) @@ -269,4 +270,4 @@ header files to use with documentation:
- [On Inverted Index Compression for Search Engine Efficiency](http://www.dcs.gla.ac.uk/~craigm/publications/catena14compression.pdf) - [Google's Group Varint Encoding](http://static.googleusercontent.com/media/research.google.com/de//people/jeff/WSDM09-keynote.pdf) -Last update: 19 JUN 2015 +Last update: 21 JUN 2015