Pogocache Outpaces Redis with Blistering Speed

Pogocache, a new open‑source caching system built in C by Josh Baker—known for creating Tile38—has entered general availability, promising breakthrough performance and seamless integration across multiple protocols. According to benchmarks on AWS c8g.8xlarge hardware, Pogocache delivers around 3.14 million queries per second, leaving behind Memcache, Redis, Dragonfly, Valkey, and Garnet.

This cache server distinguishes itself by supporting a range of existing communication standards—Memcache, Redis/Valkey, HTTP, and PostgreSQL wire protocol—so developers can leverage familiar tools like curl, psql, and Redis clients without rewriting code or managing multiple ports. The protocols are automatically detected, further simplifying adoption.

The architecture prioritises performance: an internally sharded hashmap architecture—often utilising thousands of shards—and Robin Hood hashing minimise contention and deliver efficient memory access even under heavy load. The tool can run either as a standalone server or be embedded directly into applications via a single C source file. In embedded mode, it bypasses network layers and can reportedly achieve over 100 million operations per second.

The 1.0 release, marking its general‑availability milestone, arrived in late July 2025, and version 1.1 has already rolled out—now with automatic background sweeps to evict expired keys. Licensed under AGPL‑3.0, the project values open‑source collaboration and transparency.

Reactions from the community reflect cautious optimism. Simon Willison—co‑creator of Django—praised the multi‑protocol flexibility, particularly the Postgres wire interface and compatibility with Redis and Memcached APIs. Yet voices like Cezar Henrique da Costa e Souza underscore the need for independent benchmarking and real‑world workload testing to confirm the impressive claims.

Beyond raw speed, Pogocache’s roadmap signals ambition: planned enhancements include domain‑specific integrations for web, SQL, geospatial data, shared‑memory access, inter‑process communication, and enterprise features such as distributed routing and failover support.

Behind the scenes, developer Josh Baker emphasises efficiency in design, focusing on minimal CPU cycles per request and optimised threading. In community exchanges, he noted aggressive profiling and tuning of both the hash structure and network stack to drive low‑latency, high‑throughput performance.

Pogocache’s emergence commands attention not just for its headline-grabbing numbers but for how well it integrates into existing ecosystems. Developers working across performance-sensitive environments—from real-time APIs and session stores to database caching layers—may find the familiar protocols and high throughput compelling.