## Profile

## Stream

### Yann Collet

Shared publicly -**Huffman revisited part 5 : combining multi-streams with multi-symbols**

In previous article , a method to create a fast multi-symbols Huffman decoder has been described. The research was using single bitstream encoding, for simplicity. However, earlier investigation proved that using multiple bitstreams was good for speed on m...

### Yann Collet

Shared publicly -**Huffman revisited, Part 4 : Multi-bytes decoding**

In most Huffman implementations I'm aware of, decoding symbols is achieved in a serial fashion, one-symbol-after-another. Decoding fast is not that trivial, but it has been well studied already. Eventually, the one symbol per decoding operation becomes its...

### Yann Collet

Shared publicly -**Fuzz testing Zstandard**

An advance issue that any production-grade codec must face is the ability to deal with erroneous data. Such requirement tends to come at a second development stage, since it's already difficult enough to make an algorithm work under "normal conditions". Be...

### Yann Collet

Shared publicly -**Accessing unaligned memory**

Thanks to Herman Brule , I recently received an access to real ARM hardware systems, in order to test C code and tune them for performance. It proved a great experience, with lots of learnings. It started with the finding that xxhash speed was rubbish on A...

### Yann Collet

Shared publicly -**Huffman revisited - Part 2 : the Decoder**

The first attempt to decompress the Huffman bitStream created by an huff0 version modified to use FSE bitStream ended up in brutal disenchanting. While the decoding itself worked fine, the resulting speed was a mere 180 MB/s . OK, in absolute, it looks r...

### Yann Collet

Shared publicly -**Huffman revisited part 5 : combining multi-streams with multi-symbols**

In previous article , a method to create a fast multi-symbols Huffman decoder has been described. The research was using single bitstream encoding, for simplicity. However, earlier investigation proved that using multiple bitstreams was good for speed on m...

### Yann Collet

Shared publicly -### Yann Collet

Shared publicly -**Huffman revisited - Part 3 - Depth limited tree**

A secondary issue that most real-world Huffman implementations must deal with is tree depth limitation. Huffman construction doesn't limit the depth. If it would, it would no longer be "optimal". Granted, the maximum depth of an Huffman tree is bounded by...

### Yann Collet

Shared publicly -**Huffman revisited - part 1**

Huffman compression is a well known entropic compression technique since the 1950's. It's optimal , in the sense there is no better construction if one accept the limitation of using an integer number of bits per symbol, a constraint that can severely l...

- L'Ecole Nationale Supérieure de Télécommunications de BretagneTélécommunications, 1993 - 1996

- OrangeMarketing M2M, 2009 - present

Huffman revisited part 5 : combining multi-streams with multi-symbols fastcompression.blogspot.com In previous article, a method to create a fast multi-symbols Huffman decoder has been described. The research was using single bitstream enc |

Huffman revisited, Part 4 : Multi-bytes decoding fastcompression.blogspot.com In most Huffman implementations I'm aware of, decoding symbols is achieved in a serial fashion, one-symbol-after-another. Decoding fast is n |

Accessing unaligned memory fastcompression.blogspot.com Thanks to Herman Brule, I recently received an access to real ARM hardware systems, in order to test C code and tune them for performance. I |

Huffman revisited - part 1 fastcompression.blogspot.com Huffman compression is a well known entropic compression technique since the 1950's. It's optimal, in the sense there is no better construct |

Huffman revisited - Part 2 : the Decoder fastcompression.blogspot.com The first attempt to decompress the Huffman bitStream created by anhuff0 version modified to use FSE bitStream ended up in brutal disenchant |

Strict Aliasing dbp-consulting.com So What's The Problem? There's a lot of confusion about strict aliasing rules. The main source of people's confusion is that there are two d |

Bloom Filters - Explained ~ Tech-Effigy techeffigytutorials.blogspot.com The Bloom filter is a space efficient, probabilistic data structure, designed to test the membership of elements to a set. The trade-off for |

ZSTD - A stronger compression algorithm fastcompression.blogspot.com ZSTD, short for Z-Standard, is a new lossless compression algorithm, aiming at providing both great compression ratio and speed for your sta |

Performance of various general compression algorithms - some of them are... java-performance.info This article compares several Java general compression algorithm implementations - you can compress your data at over 300 Mb/sec using some |

Wifi Analyzer market.android.com Turns your android phone into a Wi-Fi analyzer!! Shows the Wi-Fi channels around you. Helps you to find a less crowded channel for your wire |

bourse aux jouets - Hauts-de-Seine - 92 bourse-aux-jouets.org découvrez les dates des bourses aux jouets dans les Hauts-de-Seine - 92 |

Counting bytes fast - little trick from FSE fastcompression.blogspot.com An apparently trivial and uninteresting task nonetheless received some special optimization care within FSE : counting the bytes (or 2-bytes |

Let's move on fastcompression.blogspot.com I've received an answer from Don Bailey. He blames the situation on a lack of communication. OK. In an attempt to bring the discussion to a |

Streaming API for LZ4 fastcompression.blogspot.com For quite some time, the LZ4 Streaming API project has been started and delayed, as other priorities stepped in the way. To be fair, one imp |

Taking advantage of unequalities to provide better compression fastcompression.blogspot.com When starting investigation on ANS properties, in November 2013, I stumbled upon the fact that positions in the table are not equivalent. Ba |

Ultra-fast normalization fastcompression.blogspot.com Today's objective is to use the lessons learned when defining the perfect normalization algorithm to design a new, better, fast normalizatio |

Perfect Normalization fastcompression.blogspot.com People keeping an eye on the github repository of FSE may have noticed the apparition a new function, called FSE_normalizeCountHC(). Let's d |

Better normalization, for better compression fastcompression.blogspot.com A secondary issue to deal with when implementing an FSE entropy coder is the requirement to normalize statistics. Basically, it consists in |

FSE tricks - Memory efficient subrange maps fastcompression.blogspot.com With last blog post, we were left with a description of the FSE compression algorithm. However, the code associated was a bit cryptic. So le |