Network Computing is part of the Informa Tech Division of Informa PLC
SAN DIEGO -- Storage Networking World -- Hollywood movie maker Pixar, famous for movies such as Toy Story and Finding Nemo, has been wrestling with a massive data challenge during the production of its latest release, Cars, according to Greg Brandeau, Pixar's VP of technology.
During his keynote here today, Brandeau explained that Cars, which is slated for a June release, has put more strain on his internal systems than any other movie, swallowing up a colossal 2,300 CPU years over the course of the last five years. In other words, in Brandeau's view, a single CPU would have to run for 2,300 years in order to do all the number crunching for this movie.
"Cars is the most ambitious film we have ever made, the frames are way more complex. We used 300 times more compute power to make Cars than Toy Story," he explained.
Brandeau added that the rendering system used for animation was creaking under the strain of the new release, prompting a rethink of the firm's storage strategy. "We realized that our rendering time was taking too long. Frames were taking 10 hours [to render] that should have taken one hour."
According to Brandeau, it was not the rendering system, which is based on Dell servers, that was causing the slowdown. On closer inspection, an NFS file system feeding data to the Dell boxes was found to be the source of the problem. "Our central file system was getting hammered in a way it had never been hammered before. The NFS caches couldn't go fast enough -- they did not have enough RAM on them."
Recommended For You
From infrastructure to app delivery, from data to applications, it’s past time to modernize your practices, processes, and providers to ensure you’re able to take advantage of AI and whatever comes next.
What skills do network managers really need to properly secure industrial networks? What new protocols, frameworks, and regulations are important? And what conferences and certifications can help? Here are five tips to get started.
A full-stack approach to retail edge offers retailers a way to optimize operations and adapt to changes in a post-pandemic world.