And improvements are quite significant : on a pure in-memory comparison, 2 threads means (almost) twice faster, damn close to linear ramping.
Similar to LZ4, i've selected block division method due to its simplicity and parallel ramp-up capability. The blocks are a bit smaller, to preserve buffers memory.
Although more complex than LZ4, Zhuff is nonetheless fast enough to exhaust any mechanical HDD bandwidth with less than a single core. You will therefore need something faster to experience Zhuff speed on multi-cores systems, something like a RAM Drive.
Even then, Zhuff is likely to reach RAM Drive speed limits on quad-core systems, something i can't test myself, but would be glad to get reports on.
| version | threads | Compression Ratio | Speed | Decoding | |
| Zhuff | 0.7 | -t1 | 2.584 | 147 MB/s | 278 MB/s |
| Zhuff | 0.7 | -t2 | 2.584 | 285 MB/s | 550 MB/s |
You can download Zhuff on its new homepage.
