To help to determine the best approach for your needs, the tables below compare performance metrics relating to the Java API. They show the detection speed for a single request, the memory usage and the startup time for each data file type, depending on the mode of operation used.

Results stated are from a quad core Xeon 2.2GHz.

Please note that results may vary based on the version of JVM and Java compiler used to build the source files. The following configuration was used to produce these results: JDK 1.7.0_71 and JVM Java HotSpot(TM) 64-Bit Server VM 24.71-b01

Modes of Operation
Memory	Initializes the data set into memory
File (Stream)	Only loads the essential headers
Array	Initializes a byte array representation of the data file

Below is a table of performance metrics relating to the Java Pattern API. It shows the detection speed for a single request for each data set and where applicable also the mode of operation used.

Time Per Detection (ms)
	Lite	Premium	Enterprise
Memory	0.05	0.07	0.07
File (Stream)	0.07	0.09	0.11
Array	0.06	0.08	0.10

Average Memory Usage (Mb)
	Lite	Premium	Enterprise
Memory	238	512	661
File (Stream)	26	26	28
Array	64	124	158

Startup Time (ms)
	Lite	Premium	Enterprise
Memory	5548	6675	7522
File (Stream)	224	216	209
Array	325	225	224

Below is a table of performance metrics relating to the Java Hash Trie API. It shows the detection speed for a single request for each data set and where applicable also the mode of operation used.

Time Per Detection (ms)
	Lite	Enterprise
Memory	0.00204	0.00206
File (Stream)	0.0069	0.0068

Average Memory Usage (Mb)
		Lite	Enterprise
Memory	JVM	25Mb	25Mb
	outside JVM	120Mb	144Mb
File (Stream)	JVM	25Mb	25Mb
	outside JVM	from 1Mb with caching

Startup Time (ms)
	Lite	Enterprise
Memory	280	310
File (Stream)	10	11