Stealth’s Unmatched Performance

Fundamentally, compression has four major performance metrics:

  • Compression Ratio: the ratio describing how much data is compressed (higher is better)

  • Compression Speed: how quickly data can be compressed, or how much data can be compressed per unit time (higher is better)

  • Decompression Speed: similar to above, for decompression (higher is better)

  • Memory (RAM) Usage: the amount of physical memory required to perform compression and decompression (lower is better)

When these four are maximized (or minimized, in the case of RAM), efficiency increases. Balancing these four metrics, along with a few other factors, is what grants Stealth its industry-leading efficiency.

If we normalize the compression ratios of six leading algorithms, and compare them to Stealth, we can easily demonstrate our performance advantage on compression speed, decompression speed, and RAM usage:

Compression Speed

Decompression Speed

Memory Usage


What does it all mean?

Simply put, greater speed and lower memory usage equals a smaller footprint. In other words, Stealth can perform the same compression task as another algorithm, while using significantly less resources, which can result in sizable savings.

What does this mean for your specific use case?


Notes:

  1. All tests performed using enwik8 from the Large Text Compression Benchmark

  2. Tests performed on a single AWS EC2 c5.2xlarge VM running Windows Server 2016, single-threaded

  3. Time measurements taken using ptime.exe, ±0.0025 ms accuracy

  4. We do not claim this test to have any scientific bearing, it is solely shared for demonstrative purposes