status-im
/
leopard
mirror of https://github.com/status-im/leopard.git
2
0
Fork 0
Code Issues Projects Releases Wiki Activity

Update Benchmarks.md

This commit is contained in:
catid 2017-05-31 23:03:05 -07:00 committed by GitHub
parent 7082e5b34c
commit 47444a9795
1 changed files with 14 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

25
Benchmarks.md
View File

@ -1,18 +1,18 @@
# Benchmarks: # Benchmarks:
On my (few year old) laptop: On my (few year old) laptop using AVX2 instruction set. I'm not being very rigorous here. The point is that it's really fast.
Some example performance measurements: Some example performance measurements:
``` ```
Leopard Encoder(0.256 MB in 100 pieces, 10 losses): Input=4990.25 MB/s, Output=499.025 MB/s Leopard Encoder(0.256 MB in 100 pieces, 10 losses): Input=5333.33 MB/s, Output=533.333 MB/s
Leopard Decoder(0.256 MB in 100 pieces, 10 losses): Input=1407.37 MB/s, Output=140.737 MB/s Leopard Decoder(0.256 MB in 100 pieces, 10 losses): Input=1695.36 MB/s, Output=169.536 MB/s
Leopard Encoder(0.256 MB in 100 pieces, 20 losses): Input=3560.5 MB/s, Output=712.1 MB/s Leopard Encoder(0.256 MB in 100 pieces, 20 losses): Input=3878.79 MB/s, Output=775.758 MB/s
Leopard Decoder(0.256 MB in 100 pieces, 20 losses): Input=632.88 MB/s, Output=126.576 MB/s Leopard Decoder(0.256 MB in 100 pieces, 20 losses): Input=833.876 MB/s, Output=166.775 MB/s
Leopard Encoder(8.192 MB in 128 pieces, 128 losses): Input=1266.13 MB/s, Output=1266.13 MB/s Leopard Encoder(8.192 MB in 128 pieces, 128 losses): Input=1964.98 MB/s, Output=1964.98 MB/s
Leopard Decoder(8.192 MB in 128 pieces, 128 losses): Input=482.243 MB/s, Output=482.243 MB/s Leopard Decoder(8.192 MB in 128 pieces, 128 losses): Input=600.542 MB/s, Output=600.542 MB/s
Leopard Encoder(2.56 MB in 1000 pieces, 200 losses): Input=1942.34 MB/s, Output=388.467 MB/s Leopard Encoder(2.56 MB in 1000 pieces, 200 losses): Input=1942.34 MB/s, Output=388.467 MB/s
Leopard Decoder(2.56 MB in 1000 pieces, 200 losses): Input=367.109 MB/s, Output=73.4219 MB/s Leopard Decoder(2.56 MB in 1000 pieces, 200 losses): Input=367.109 MB/s, Output=73.4219 MB/s
@ -41,6 +41,9 @@ The results are all from libraries I've written over the past few years. They a
``` ```
For 64KB data chunks: For 64KB data chunks:
Leopard Encoder(8.192 MB in 128 pieces, 128 losses): Input=1964.98 MB/s, Output=1964.98 MB/s
Leopard Decoder(8.192 MB in 128 pieces, 128 losses): Input=600.542 MB/s, Output=600.542 MB/s
CM256 Encoder: 64000 bytes k = 128 m = 128 : 82194.7 usec, 99.6658 MBps CM256 Encoder: 64000 bytes k = 128 m = 128 : 82194.7 usec, 99.6658 MBps
CM256 Decoder: 64000 bytes k = 128 m = 128 : 78279.5 usec, 104.651 MBps CM256 Decoder: 64000 bytes k = 128 m = 128 : 78279.5 usec, 104.651 MBps
@ -56,12 +59,12 @@ FEC-AL Decoder(8.192 MB in 128 pieces, 128 losses): Input=121.093 MB/s, Output=1
For 128 data pieces of input and 128 data pieces of redundancy: For 128 data pieces of input and 128 data pieces of redundancy:
+ Fastest to encode: Leopard (1.26 GB/s) + Fastest to encode: Leopard (1.96 GB/s)
+ Distant second-place: WH256 (660 MB/s), FEC-AL (515 MB/s) + Distant second-place: WH256 (660 MB/s), FEC-AL (515 MB/s)
+ Slowest encoders: Longhair, CM256 + Slowest encoders: Longhair, CM256
+ Fastest to decode: WH256 (830 MB/s) + Fastest to decode: WH256 (830 MB/s)
+ Distant second-place: Leopard (480 MB/s) + Distant second-place: Leopard (600 MB/s)
+ Slowest decoders: FEC-AL, CM256, Longhair + Slowest decoders: FEC-AL, CM256, Longhair
There are a lot of variables that affect when each of these libraries should be used. There are a lot of variables that affect when each of these libraries should be used.
@ -136,8 +139,8 @@ http://github.com/catid/siamese
##### Leopard-RS *new*: O(K Log M) FFT MDS Reed-Solomon codec ##### Leopard-RS *new*: O(K Log M) FFT MDS Reed-Solomon codec
Encodes at 1.2 GB/s, and decodes at 480 MB/s for this case. Encodes at 2 GB/s, and decodes at 600 MB/s for this case.
12x faster than existing MDS approaches to encode, and almost 5x faster to decode. 20x faster than existing MDS approaches to encode, and 6x faster to decode.
This uses a recent result from 2014 introducing a novel polynomial basis permitting FFT over fast Galois fields. This uses a recent result from 2014 introducing a novel polynomial basis permitting FFT over fast Galois fields.
This is an MDS Reed-Solomon similar to Jerasure, Zfec, ISA-L, etc, but much faster. This is an MDS Reed-Solomon similar to Jerasure, Zfec, ISA-L, etc, but much faster.
It requires SSSE3 or newer Intel instruction sets for this speed. Otherwise it runs much slower. It requires SSSE3 or newer Intel instruction sets for this speed. Otherwise it runs much slower.