Spark RDD Memory

Question

val rddMap = rdd.map(x=>x*x)
val rddReduce = rddMap.reduce((a:Int,b:Int) => a + b)

At which step will be the data loaded into the Memory of worker