Traditional data integration products use the Extract-Transform-Load (ETL) architecture to run the data transformations. In this architecture, a dedicated proprietary engine is in charge of the data processing. This engine introduces additional complexity and cost in the data integration picture, and frequently becomes a performance bottleneck.
The ELT (Extract-Load and Transform) architecture used in Semarchy Convergence for Data Integration™ removes the need for this middle-tier engine. It leverages the data processing engines in place to run processes generated and optimized for these engines.
By removing the cost of the hardware, software, maintenance and tuning skills required for the ETL server, the ELT architecture guarantees the best TCO. This architecture scales naturally with the source and target data stores. It guarantees the best performance and scalability for data integration at any time.