This readme will be replaced

Squash

A provably fast and expensive hash algorithm

Squash Hash

The following things have to be done

• Security Analysis (Review)
• Benchmark on ARMv8, SSE4.2 compatible CPU and AMD GPUs.

Tests

Tests and sample results can be found in the Tests Branch. Head over there for further instructions.

Speed

On a Xeon E3-1225v2, the algorithm has an average speed of 58.5ns. For comparision, keccak takes about 800ns on average. Which results in 5.8 cpb for squash, 80 cpb for keccak - on an Ivy Bridge CPU.
On a Raspberry Pi 3b+, the calculation of one hash takes about 1534ns on average, which results in 33.5 cpb. For comparision, one keccak execution takes about 15467ns which results in 338 cpb.

Compilation

To get the optimal speed, it is recommended to compile using the -O3 -march=native (5x speed) flags aswell as -maes (9x speed) if supported by the CPU. Additionally it is recommended to use GCC instead of Clang. The compilation process will take about 20% less time and the execution of the hash will be about 25% faster. Differences in execution speed from GCC-6, GCC-7 and GCC-8 are very low, but GCC-7 consistantly gets the fastest results.

SquashPoW

Speed

The same Xeon E3-1225v2 as mentioned above has about 100kH/s (single-threaded, using DDR3-1333 RAM) using the 4GiB dataset, 11H/s using the light mode and a 64MiB cache.

Reference

A reference implementation for cryptonote based currencies can be seen in this pull request made for Clash, a Turtlecoin fork.