Google Caffeine, the search giant’s secret project, is officially in testing mode and going live in its first datacenter very soon. The developer’s preview of the new search tool in the Google Sandbox was launched back in August but little feedback has been heard about the pre-beta release.

The new search engine is engineered so that the average user won’t notice any changes; however, power users may detect a slight difference, which is why there was a power-use preview released first.

“Fundamentally, it’s a rewrite of our indexing system,” said Matt Cutts, a principal engineer at Google. “If you want to have a good search engine you have to crawl really well, you have to index really well and then you have to serve it up really fast.”

Google Caffeine will set a powerful foundation for Google’s future indexing changes. To put the project into perspective, the new search tool is comparable to the Big Daddy update that happened in late 2005. Once it reaches the right level of quality it will most likely expand out to other datacenters.

The following differences were found by web developers and power searchers who tested Google Caffeine.

• More results show on the results page
• Sites appear to have higher page rankings
• Search results save to html files in the sandbox rather than a pdf file
• Search results appear faster
• Search results are more relevant
• Sites that link from news, media, social networks, and content-oriented links are ranking higher than sites that have directory links

The full Caffeine roll out is expected to happen in early 2010. Matt Cutts talks about Google Caffeine on his blog, if you are interested in learning more.