The
SPD Engine is designed for high-performance data delivery. It enables an application rapid access
to SAS data for processing. The SPD Engine delivers data to applications rapidly because
it organizes the data into a streamlined file format that takes advantage of multiple
CPUs to perform parallel input functions.
The SPD Engine uses threads to read blocks of data very rapidly and in parallel. Tasks are performed in conjunction
with an operating system that enables threads to execute on any of the computer’s
available CPUs. Although threaded Read tasks are an important part of the SPD Engine
functionality, the real power of the SPD Engine comes from how it structures SAS
data. The SPD Engine organizes data into a file format that includes partitioning
the data. This data structure permits threads, running in parallel, to perform Read
tasks efficiently.
The SPD Engine is a high-speed alternative to the default Base SAS engine for processing
very large
data sets. It reads data sets that contain billions of observations. For example,
this includes data sets that expand beyond the size limit imposed by some operating
systems and data sets that SAS analytic software and procedures must process faster.
The SPD Engine boosts performance in the following ways:
-
support for hundreds of gigabytes
of data
-
scalability on symmetric multiprocessor (
SMP) computers and massively parallel processor (MPP) computers
-
parallel WHERE selections
-
-
-
parallel data delivery to applications
-
automatic sorting on BY statements
The SPD Engine runs on UNIX, Windows, and z/OS (zFS file system only).