By delivering what was once server-room HPC horsepower on a workstation-class machine, Watters said engineering groups can be assured of faster turnaround on simulation jobs. For example, allowing them to understand designs better before sending them out for certification testing while being to affect changes at the design engineer's seat rather than having to go through the protracted cycles of coordinating an iterative loop with an external simulation organization.
The coupling of the technology also keeps real-world photo realism and real-world physics turned on during all stages of the design cycle, Watters said.
If you're an auto stylist, you don't want to work in pastel colors, you want the actual design and assembly work done with photorealism turned on. When photorealism is on all the time, people make better design decisions.
Same can be said for a designer studying tradeoffs between esthetics and functionality for a wheel design. With the added horsepower of the Kepler-based Telsa/Quadro combination, they can get dynamic feedback on the structural load of the wheel design as they are working, inviting the exploration of lighter weight materials.
The increased number of cores supported by the Kepler architecture, in addition to other improvements, up the performance of the new Maximus architecture for tasks like simulation on an order of a fourfold increase, depending on the specific operation, Watters said. Not only does this increase allow for faster simulation and test cycles, but it also enables engineering groups to prove out more designs with simulation in the same amount of time.
NVIDIA's usual OEMs are supporting the new Maximus architecture, so expect to see workstations from Hewlett-Packard, Dell, Lenovo, and Fujitsu, among others. Leading CAD and CAE vendors such as ANSYS, Autodesk, Bunkspeed, Adobe, Dassault Systemes, and The MathWorks have also certified their applications to run on the new Maximus generation. GPUs based on the new Kepler-powered Maximus will be available in October.
For industrial control applications, or even a simple assembly line, that machine can go almost 24/7 without a break. But what happens when the task is a little more complex? That’s where the “smart” machine would come in. The smart machine is one that has some simple (or complex in some cases) processing capability to be able to adapt to changing conditions. Such machines are suited for a host of applications, including automotive, aerospace, defense, medical, computers and electronics, telecommunications, consumer goods, and so on. This discussion will examine what’s possible with smart machines, and what tradeoffs need to be made to implement such a solution.