I put up a benchmark graph showing how Whirlycache performs with 20 concurrent threads doing 50% read activity earlier today on the Whirlycache site. Here’s the graphic for your viewing pleasure. I also added a caveat about the usefulness of benchmarks right on the Whirlycache homepage, so please read that before you blast me here.
I can summarize: when you have a lot of threads going against a particular cache implementation containing a lot of elements, Whirlycache will basically always come out on top. Obviously, it doesn’t have all of the features that some of the other caches have, so there’s a tradeoff that may be important to you.