DataXchg: A Portable Data Exchange Toolkit for Heterogeneous Computing
Zhu Wang, Chonlei Mei, Hai Jiang · 2010
In distributed systems, data generated on one machine might not be able to be used by others directly because of the incompatibility issue caused by data type, endianness, size and padding situation. Data conversion procedure is indispensable, especially in open systems. So far, there is no widely accepted communication toolkit to support heterogeneous data exchange, especially for aggregate type data, in high performance computing community. Most time, programmers have to handle data formats manually or pass scalar type data to standard libraries for symmetric data conversion. In order to achieve high programmability and efficiency in both homogeneous and heterogeneous open systems, a portable data exchange toolkit, DataXchg, is proposed. Data items are represented by binary tags and data conversion only happens in heterogeneous environments. It eases programmers' coding work so that all data types, aggregate and scalar, will be exchanged easily between computers. Experimental results have demonstrated its effectiveness and performance gains in terms of productivity and data conversion speed.