Multiple Images to Processed Images - Process hundreds of images with a single workflow.

Understanding batch operations

Batch processing is the practice of applying the same operation to a collection of files in a single run rather than processing them one at a time. The concept dates to the earliest mainframe computers, which queued punch card jobs into batches to maximize CPU utilization. Modern desktop batch tools follow the same principle: define an operation, point it at a list of files, and let the computer work through them while the user does something else.

Effective batch processing depends on three things: parallelism, error isolation, and progress reporting. Parallelism means running multiple operations simultaneously to use all available CPU cores; a 10-core machine can process 10 files in parallel and finish a batch in roughly one tenth of the time of serial processing. Error isolation means a single failed file does not abort the entire batch; failures are recorded and the remaining files continue processing. Progress reporting tells the user how many files are done, how many remain, and the estimated time to completion.

Batch operations on local hardware are usually I/O bound rather than CPU bound, especially for image and PDF conversions. Reading and writing files is slow compared to the computation, so the optimal level of parallelism is usually equal to the number of CPU cores or slightly higher. Going beyond that does not improve throughput because the disk becomes saturated. SSDs handle parallel I/O much better than spinning disks, which is why batch performance on modern Macs and PCs has improved dramatically since the SSD transition.