Kernkompression

Kernkompression, or kernel compression, denotes the set of techniques used by an operating system kernel to store data in compressed form in memory or on disk in order to reduce the overall memory footprint and improve I/O efficiency. It can operate on data structures kept by the kernel, such as the page cache and swap space, and can also apply to file-system data managed by the kernel. The approach is usually transparent to ordinary applications, which continue to access data as usual while the kernel handles compression and decompression.

In memory management, kernkompression reduces RAM usage by storing pages and caches in compressed form, allowing a larger active data set to fit into limited memory. It often complements swap, lowering disk I/O when swapping becomes necessary. Real implementations include zram and zswap within the Linux kernel, and similar memory compression features in Windows and macOS.

File-system level compression applies the same concept at the storage layer, compressing data before writing it to permanent storage and decompressing it on read. This can save disk space at the cost of additional CPU work and potential latency. Examples include in-kernel support in file systems like Btrfs and APFS, and compression capabilities in NTFS on Windows.

Common algorithms used in kernkompression are based on LZ-family schemes such as LZ77/LZ78, LZ4, and Zstandard, sometimes combined with entropy coding. Effectiveness depends on data characteristics; highly random data yields modest gains and can incur CPU overhead. Trade-offs include CPU cost, latency, memory bandwidth, and energy use, making kernkompression most beneficial for workloads with redundancy and predictable access patterns.

Historically, kernel-based compression arose from memory management research and has been adopted in several operating systems to improve usable memory and reduce swap pressure. It remains an area of ongoing optimization as hardware memory and storage hierarchies evolve.