Author Archives: Richard Startin

Vectorised Algorithms in Java

There has been a Cambrian explosion of JVM data technologies in recent years. It’s all very exciting, but is the JVM really competitive with C in this area? I would argue that there is a reason Apache Arrow is polyglot, and it’s not just interoperability with Python. To pick on one project impressive enough to


The conventional wisdom is that the stages of a task should be pipelined, so you don’t need to wait for the completion of one stage before the next is started. It surprises me that it seems you can sometimes do better when performing each stage of a pipeline in a short batch. Useful optimisation opportunities

Data Driven Logic

I really don’t like reading or writing blocks of if-else statements. They make my eyes glaze over. Rumour has it that processors don’t like executing them either, though that’s less true now than it once was. There are two problems with these blocks of statements, and neither one of them is performance: They are hard

Garbage Collector Code Artifacts: Card Marking

In the JVM, lots of evidence of garbage collection mechanics can be seen from JIT compiler output. This may be obvious if you think of garbage collection as a task of book-keeping: the various auxiliary data structures used to track inter-region or inter-generational references, relied on for faster marking, need to be kept up to

Parallel Bitmap Aggregation

A bitmap index represents predicates over records as sets consisting of the integer identities of each record satisfying each predicate. This representation is actually a few decades out of date, and systems like Pilosa use much more sophisticated data structures, and Sybase had even more on offer back in the 90s. But the chances are,