Meta's CacheLib Sees New Release After Two Year Hiatus For Helping With High DRAM Prices

Meta released CacheLib 2026.05.25 on Monday, its first new version in two years, to help services cope with surging DRAM prices driven by the AI boom. The open-source caching engine, originally launched in 2021 to offset memory costs using non-volatile drives, now returns amid astronomical DRAM pricing compared to 2021 levels.

Meta's CacheLib Sees New Release After Two Year Hiatus For Helping With High DRAM Prices Back in 2021 CacheLib was originally started to help with caching on non-volatile memory drives to augment their existing DRAM-based caching of services at Facebook/Meta. With memory pricing being where its at over the past year, CacheLib has now seen its first new released version since June 2024. There isn't any public change-log or concise release notes as of writing for today's CacheLib 2026.05.25 release, but those wanting to try it out can find it at Facebook open-sourced CacheLib as a new caching engine https://www.phoronix.com/news/Facebook-Open-Source-CacheLib . Back in 2021 it was done to help scale services with non-volatile memory caching to offset increasing DRAM costs at the time. Now in 2026, DRAM memory prices are astronomical compared to 2021 pricing given the AI surge. And, surprisingly, Meta is out with a new CacheLib release after being absent the past two years.CacheLib was originally started to help with caching on non-volatile memory drives to augment their existing DRAM-based caching of services at Facebook/Meta. With memory pricing being where its at over the past year, CacheLib has now seen its first new released version since June 2024. There isn't any public change-log or concise release notes as of writing for today's CacheLib 2026.05.25 release, but those wanting to try it out can find it at GitHub https://github.com/facebook/CacheLib/releases/tag/v2026.05.25.00 and CacheLib.org https://cachelib.org/ . The modern CacheLib continues to describe itself as a " pluggable caching engine to build and scale high performance cache services ."