Compared to other popular interchange format approaches like TIF or SSF, the PIF/PBF approach seems to be the most promising, since on the one hand it provides an elaborate human-readable ASCII form suitable for intersite data exchange, and on the other hand a compact and optimized random-access binary format suitable as a TCAD framework binary intertool format.
Conventional file-based approaches like TIF [Axel93] and SSF [Hopp93] employ an ASCII format which does not allow random access, i.e. a TCAD tool has to scan the whole file for the data it wants to read (which is problematic considering the large amounts of data required for three-dimensional simulations), whereas in a binary PLB the tool directly accesses the needed data through a symbol in the hash table; a process requiring not more than one pointer dereference. In knowledge of the deficiencies of an ASCII intertool exchange format, the TIF and SSF format's keywords are kept short and are not more than single letters, or even worse, just numbers, thus violating the goal of human-readability, which is advantageously met by the well-structured and verbose PIF ASCII form.
The DATEX format developed by the ``Institut für Integrierte Systeme'' at the Swiss Federal Institute of Technology exhibits an ASCII syntax based on keywords associated with data values in curly braces. Keywords for defining coordinate transformations, tensor product and unstructured element grids, as well as arbitrary data sets which are used to define, e.g., doping profiles. The format may be extended by adding keywords to the list of reserved keywords. A ``binary form'' is employed too by compressing the ASCII files; a tool has to decompress the file prior to reading and after writing, since data can not be accessed directly in the compressed file. Therefore random access is not possible with this format, which seems to be problematic since this format explicitely supports three-dimensional geometries and grids, and thus has to cope with large data amounts.