High Performance Technical Computing
No matter their differences in performance of Itanium adoption, Early Adopters and Early Majority users can be grouped together in the type of computing that is being performed. This is High Performance Technical Computing, (HPTC for short), and it's best described as any form of computing where you are using the mathematical, analytical power of the processor.
The focus is on mathematical precision such as the multiplication of very large or small numbers with long numbers of decimal places and trying to get the solution in a speedy manner. Another way to describe it is in term of complexity: Illustrating a Windows background is not HPTC, but Pixar-style digital content creation is. Using Microsoft®[15] Paint would not count as processor intensive for HPTC, but drafting designs of industrial equipment in a major vendor's CAD application certainly would.
[15] Microsoft is a U.S. Registered trademark of Microsoft Corp.
The people performing technical computing are the ones who use computer cycles in gross quantities. The case studies done by the National Crash Analysis Center, (NCAC), in Washington is a great example of the demands that HPTC can place on a computer system. If you're going to simulate the crash of a couple of cars, you'll be looking at every minute element of the subject matter as it changes and goes through the crash test.
On current RISC based technology, this is a simulation that runs for an extremely long time to get the data crunched. According to one of the team members at the NCAC, on a 3-4 year old server it takes a solid month to run a single crash analysis. He got an Itanium based system in place and the exact same crash analysis reduced the calculation time to less than 48 hours.
NCAC is a typical high performance, technical computing client. They have problems that can be run today, but take extremely long run times. In the past, these clients might require a supercomputer to run these problems. However, even if they are one of the lucky few that can get access to one, using Itanium-based systems to work through these problems provides a much less expensive alternative.
It also illustrates how when computing power is available at a superior price-performance ratio, the additional power is immediately re-directed at solving more complex problems. In the case of NCAC, once it became less time-intensive to work on a single, straightforward head-on crash, they began to expand their work into more difficult studies.
For example, testing how off-center crashes might be simulated, or the multiple impacts of a car that collides with an object and then goes off the road. The additional power of the Itanium-based system allows them to more closely simulate a real-world problem. The end result is that we'll be driving safer cars in the future.