Next Page | Prev Page | Main Index | Home |
MPbase is just not quite like anything you may have come
across before. It has many characteristics that are truly unique.
It also has many characteristics that it shares with current data
handling environments. Most people will look at a single MPbase
characteristic and say "Oh that means MPbase is just
like ________." From this myopic, "single characteristic"
point of view they will normally be correct.
However, it is important to consider that at its core MPbase
is a unique entity. It is neither fair nor safe to make assumptions
based on single matching characteristics. Normally certain characteristics
are reliable indicators of others. In other cases certain characteristics
are mutually exclusive. Most of these conditions do not hold true
in the MPbase environment.
The following paragraphs describe many of the major characteristics
of MPbase. At first read they may seem to be a set of impossible
combinations. It is possible! This is due to the unique internal
architecture of MPbase. The only possible way to describe
MPbase is with a large and inaccurate set of "sort-of-likes."
Maybe someday there will be some "just-likes." But for
now "sort-of-likes" will have to do.
MPbase internally is sort of an
Update-efficient, highly compressed, content-addressable, multidimensional, data-intelligent, relational, hierarchical, object-oriented, platform-independent, self-organizing, massively parallel, self-tuning, fault-tolerant, network-centric, hypercube.
Also it is sort of
Index-free while at the same time being just one big index. The
fastest possible structure for access and/or update. Best used
for transaction-based processing and/or best for batch processing.
MPbase is precisely none-of-the-above. But it is sort of
all-of-the-above. In addition, the more data and tasks you give
MPbase the better it works. The better the compression
works. The better the update processing works. The better the
batch processing works. The better the transaction processing
works. The farther MPbase gets behind, the faster it will
catch up.
The combination of characteristics in MPbase only make
sense from the inside looking out. From the outside looking in
there seems to be no possible way that this combination of characteristics
could possibly come from one system. Internally MPbase
has changed all the rules. Therefore all of the normal assumptions
need to be revisited. Most of the limits to current information
processing are self-imposed. It is time to lose those limitations
and start using the full power of the underlying hardware.
Next Page | Prev Page | Main Index | Home | © 1998-2004 NPSI |