Benchmarking SSDs: Compressible vs. Incompressible Data

Benchmarking Solid State Drives (SSDs) requires testing with both compressible and incompressible data to accurately gauge real-world performance. This article explores the fundamental differences between compressible and incompressible files, how SSD controllers manipulate data to boost speeds, why manufacturers favor specific test results for marketing, and how utilizing both testing methods provides a complete picture of a drive's true speed and lifespan.

Defining Compressible and Incompressible Data

To understand SSD testing, you must first understand the data being written. Compressible data contains repeating patterns and redundancies that algorithms can easily shrink. Common examples include plain text files, HTML code, spreadsheets, and many operating system files.

Incompressible data is information that has already been compressed or is entirely random, meaning an algorithm cannot shrink it any further. Examples include MP4 videos, JPEG images, MP3 audio files, ZIP archives, and encrypted data.

The Role of the SSD Controller

The SSD controller is the processor inside the drive that manages how data is read from and written to the NAND flash memory. Several SSD controllers feature built-in, real-time data compression.

When you save a file, the controller intercepts the data and attempts to shrink it before writing it to the storage chips. If you save a 10GB compressible file, the controller might shrink it down to 4GB. Because the drive only has to physically write 4GB of data, the transfer finishes much faster. The computer, however, calculates the transfer speed as if the full 10GB was written, resulting in massive benchmark scores.

Exposing True Hardware Limits

Testing with only one type of data gives a skewed perspective of an SSD's capabilities.

Benchmarking software that uses compressible data (like ATTO Disk Benchmark) demonstrates the absolute best-case scenario for a drive. Storage manufacturers frequently use these specific tests to advertise "up to" speeds on their packaging because the numbers are higher.

Benchmarking software that uses incompressible data (like AS SSD or specific settings in CrystalDiskMark) bypasses the controller's ability to shrink files. The SSD is forced to write every single byte of data to the NAND flash. This exposes the true, raw physical write speed of the memory chips, which is often significantly slower than the speeds achieved with compressible data.

Matching Benchmarks to Real-World Workloads

Testing both data types is necessary because everyday computing involves a mixture of both.

If you use your computer primarily for office work, web browsing, and software development, your workload is heavily reliant on compressible data. A drive that excels in compressible benchmarks will feel very fast for these tasks.

Conversely, if you are a video editor working with raw footage, a photographer transferring thousands of images, or a gamer installing massive, pre-compressed game archives, your workload consists almost entirely of incompressible data. Relying solely on compressible benchmark results would lead you to buy a drive that bottlenecks under heavy media workloads.

Understanding Drive Endurance

Data compression also impacts the physical lifespan of an SSD. NAND flash memory can only be written to a finite number of times before it degrades.

When an SSD successfully compresses data, it writes less physical information to the drive. This reduces wear and tear on the memory cells, a concept known as reducing Write Amplification. By analyzing how a drive handles both compressible and incompressible data, users—especially those in enterprise environments—can better predict the drive's longevity and endurance based on the specific types of files they plan to store.