The gigaflops behind "Avatar" Print
Written by Harris Georgiou   
Sunday, 03 January 2010 00:00

First of all, happy new year, good health and good kick-off start to all.

One of the major features of this holiday season was the opening of the film "Avatar" (2009), an innovative 3D animated film with many CGI state-of-the-art technologies during its development. It's not the first time a blockbuster is by far better and more realistic than the previous top, in this case probably the second "Transformers" film (2009), also rich in heavy-duty CGI features. In the case of "Avatar", the power behind the breath-taking 3D rendered animation is a combined effort of the processing power of two studios, Weta Digital and ILM.

Weta's studios used HP's Cluster Platform 3000BL, specifically the BL2x220, L54xx servers, each with dual Intel EM64T Xeon L54xx (Harpertown) at 2.5 GHz and 32-64 GB of memory. This particular HP system is one of the latest "blade" sries and it has a peak performance of 10 Gflops. Thus, the combined processing power of the platform, a cluster of nodes containing (roughly) 38,000 processors in total and interconnected via gigabit Ethernet, can be estimated to more than 300,000 Gflops of raw speed. Nevertheless, Weta Digital and ILM worked in parallel in order to support the new motion-capture virtual camera that Avatar's director James Cameron used during the full-3D stereoscopic shooting, as well as many more months of endless rendering for producing the spectacular final result.

Na'vi character from the film 'Avatar' (20th Century Fox (c) 2009)

Na'vi character from the film 'Avatar' (20th Century Fox (c) 2009)

Related links:

 

NOTE: The site is currently in testing (beta) version and several pages are still under heavy construction or empty. Please stay tunned for frequent updates and of course feel free to send feedback on any bugs, broken links, etc.

Last Updated on Wednesday, 27 April 2011 20:47