Hadoop does not require a very high-end server with large memory and processing power. Due to this we can use any inexpensive system with average RAM and processor. Such kind of system is called commodity hardware.
Since there is parallel processing in Hadoop MapReduce, it is convenient to distribute a task among multiple servers
and then do the execution. It saves cost as well as it is much faster compared to other options.
Another benefit of using commodity hardware in Hadoop is scalability. Commodity hardware is readily available in market. Whenever we need to scale up our operations in Hadoop cluster we can obtain more commodity hardware. In case of high-end machines, we have to raise purchase orders and get them built on demand.