batch processing
/batch PRAH-ses-ing/ · noun · Development · Origin: 1950
Definitions
A method of running programs where jobs are collected into groups (batches) and executed sequentially without user interaction. Dominant in the mainframe era when computer time was expensive and shared, batch processing persists today in ETL pipelines, report generation, and overnight reconciliation jobs.
In plain English: Running a bunch of computer tasks in a group, one after another, without anyone needing to babysit them — still how many overnight business jobs work.
Example: The nightly batch processing run takes four hours to reconcile all transactions, so any bug in the pipeline means the finance team's morning is ruined.
Origin Story
Computing's original workflow: stack the cards and wait
**Batch processing** dates to the earliest days of computing in the 1950s. Programmers would prepare their jobs on **punch cards**, submit the deck to operators, and wait hours or days for results. The operator would feed batches of jobs sequentially into the computer.
The term "batch" comes from baking and manufacturing -- processing items in groups rather than individually. GM and North American Aviation developed the first batch processing operating systems for the IBM 701 in the early 1950s to maximize expensive computer utilization.
Interactive computing (timesharing) eventually replaced batch for most user tasks, but batch processing never died. ETL pipelines, nightly report generation, payroll processing, and MapReduce jobs are all batch processing. AWS Batch, Apache Spark, and cron jobs carry on the tradition.
Coined by: General Motors / North American Aviation (early batch systems)
Context: 1950s mainframe computing
Fun fact: At many 1960s universities, students got one batch submission per day. A single typo meant losing an entire day's work. This constraint bred a culture of extreme carefulness in code that today's 'move fast and break things' ethos has entirely abandoned.