.
This commit is contained in:
@ -27,8 +27,8 @@ TurboPFor: Fastest Integer Compression [ Finite Context Method FCM/DFCM floating point compression
|
||||
* Using **TurboPFor**, unsurpassed compression and more than 5 GB/s throughput
|
||||
* :new: **Time Series Compression**
|
||||
* **Fastest Gorilla** style compression with **variable byte** or **bit/io**.
|
||||
* Full range 32/64 bits, more efficient. > 5 GB/s compression and > 10 GB/s decompression
|
||||
* **Fastest Gorilla** style compression (incl. **RLE**) with **variable byte** or **bit/io**.
|
||||
* Full range 32/64 bits, better compression. Up to 15 GB/s compression and > 10 GB/s decompression
|
||||
* **Inverted Index ...do less, go fast!**
|
||||
* Direct Access to compressed *frequency* and *position* data w/ zero decompression
|
||||
* **Novel** **"Intersection w/ skip intervals"**, decompress the minimum necessary blocks (**~10-15%)!**.
|
||||
|
Reference in New Issue
Block a user