Benchmarks about lzfse HOT 4 OPEN

There was this post last month, running the Compression library on an iPhone 7.
https://artandlogic.com/2017/02/save-time-energy-libcompression/

from lzfse.

That is interesting. What kind of data was this working with (binary, text, etc.)? Is it publicly available for one to test with? Would be interesting to also compare performance on desktops with some range of processors (e.g. different SSE and AVX instruction sets).

To give an idea of what I'm getting at, please take a look at these benchmarks. I'm sure there are strengths and weaknesses of the lzfse approach. In order to get a clear idea of what they are compared to other algorithms, it would be great to see where lzfse sits on similar graphs. This way one can more easily determine when lzfse is the right choice for a problem or another compression algorithm is better suited.

from lzfse.

FWICT I think squash, the library/benchmarking suite used to generate those results, can build a copy of lzfse. So I think it should be possible to include lzfse in the results. Have asked the project about including lzfse in the results.

xref: quixdb/squash#226

from lzfse.

Squash had a beta version with LZFSE from the public repository. The compression library in macOS and iOS is finely tuned for the hardware, and will be faster.

from lzfse.