You might know that Firebird supports various high-performance (HPC) computing workflows, but have you heard of high-throughput (HTC) computing?

HPC is useful when a single task requires a substantial amount of computing power to complete; for example, performing complex mathematical operations on a massive amount of data. HTC, on the other hand, is useful when you need to perform an independent discrete task on many items. For example, running a pattern identification routine against 10,000 separate images. Another common example is running a parameter sweep across a dataset, such as running the same calculation, adjusting a single variable thousands of times. In both examples, each iteration is a separate task, entirely independent of the others.

We’re expanding support for HTC workloads on Firebird. For larger operations, however, it is possible to run independent jobs on thousands of systems throughout the United States, including Firebird! (Incidentally, the mascot of HTC is the condor, because it can scavenge available compute resources – a natural companion for Firebird.)

If this sounds relevant to your work, please reach out to Jason Simms (jsimms1@swarthmore.edu).