If the observations are being read in parallel, then the order of observations in memory is somewhat randomized: the fastest thread of the moment gets it's block of observations to be the first in HPFOREST memory.  The Out-Of-Bag sample is chosen by in-memory observation number.   Different Out-Of-Bag samples produce different trees.   To check this theory, you can set THREADS=1 on the PERFORMANCE statement.   When HPFOREST runs on a cluster of machines (MPP mode, that is), then reproducibility is foiled by the system that gives random machine numbers to HPFOREST.   HPFOREST includes the machine number in some random choices so as not to have different machines doing identical things. ... View more