Denodo performance gains are most apparent when tables from disparate data sources are combined. Denodo leverages the strengths of different underlying data sources by delegating to each of them the work that they do best, and then combining the results in such a way as to minimize what could slow them down, notably the transfer of large amounts of data over the network. For relational database data sources like Oracle, Denodo will delegate many joins and aggregations, for example, which they do efficiently. Denodo also can also delegate to massively parallel processing systems (MPP) like Spark and Impala operations like joins and aggregations which they process particularly well. Denodo can also automatically cache views if needed, which can be especially useful for slower data sources, like certain web services.
The Denodo data virtualization layer which performs these connections and optimizations does add some overhead, so if you are selecting data from one table in one single data source like Oracle, the access time will be slower (since in this simple scenario, there is nothing for Denodo to transform or optimize and you are adding Denodo as an additional layer). I suggest you try some queries which combine your data in more complex ways and which represent other real-life use cases. These may be a better gauge of the performance gains you can expect in a production scenario, and will give you an idea of the performance capabilities of the platform.
Also, keep in mind that Denodo can provide other benefits of data virtualization beyond performance, such as centralized access to multiple data sources and a standardized security and governance layer. It also allows you to easily deliver multiple access methods to consuming applications (JDBC, ODBC, and web services).
For more information about performance considerations when building a Logical Data Warehouse, and a detailed explanation of many of Denodo's optimization strategies, check out this:
And for information about the Denodo query optimizer, how it works and how to configure it, check out this:
Hope this helps!