Resource Requirements

The Vestec Speech Engine utilizes an efficient client-server architecture that can scale to support several dozen ports, given appropriate computational resources. The memory and CPU load for each speech port is relatively small and an ordinary PC can easily support several ports.

(One “port” refers to one “channel” of speech recognition. So, a 4-port setup means simultaneous speech recognition on four communication channels.)

The following table presents sample memory and CPU usage of Vestec engine for different number of speech ports. The test was conducted on a machine with Intel Pentium D 2.80GHz CPU, containing 1GB RAM, and running Linux Debian 5 operating system. The grammar consisted of 500 distinct keywords while the processed audio was 1.6 seconds long. Each test ran for 140 seconds, with each interaction taking 14 seconds.

NUMBER OF PORTS	MEMORY USAGE (MB)	CPU USAGE (%)
1	54.6	1.74%
2	100.8	3.45%
4	193.2	6.45%
8	378.0	13.11%