Nvidia has made several announcements at its SC2011 supercomputing conference, aimed at expanding its presence within the HPC, workstation, and graphics processing markets.
Perhaps the chipmakers' most significant move was the launch of Maximus, a new workstation architecture combining its Quadro GPUs and Tesla C2075 companion processors for high-end graphic visualization.
The offering is the first of its kind to simultaneously enable interactive graphic design and design rendering or simulation -- two processes that, historically, had to be carried out separately within a workstation, Nvidia said.
"In the past, you had to either design or simulate; you couldn’t do both at the same time," said David Watters, senior director, marketing, manufacturing and design industries at Nvidia. "Each process was very slow if you tried to do both, so performance seriously suffered. This [Maximus] eliminates this choice. You can do both design and simulation on the same workstation."
This simultaneity is enabled through Maximus’ use of both Nvidia Quadro GPUs, fueling the 3D graphics design and visualization piece, and its Tesla processors, handling the parallel computing necessary for design rendering or simulation.
The coming together of design and simulation is especially valuable to manufacturers looking to optimize product design and durability testing, said Watters.
"You can design the interior of a car or a bike and see very compellingly what that product will look like in real life," Watters said. You can see how a product reacts to certain lighting, how "real-world" physics may work with or against its design, or how easily its materials may wrinkle or tear.
"It could take days before knowing if it [a product] is structurally sound," Watters explained. "Now, with computational simultaneity, you can know instantaneously."
Apart from designers, Maximus workstation architecture presents a new opportunity to VARs, Watters said, as the product’s appeal lies within an already channel-driven market.
"One of the big things about the manufacturing space is that it’s powered and fueled so much by regional VARs and specialty VARs, so they can absolutely take advantage here," Watters said.
Maximus will also carve out a space for VARs within the SMB market, the Nvidia spokesman noted.
For smaller manufacturers working with only 10 or 15 workstations, the optimization of those machines is critical. As simultaneous design and simulation means freeing up precious resources, the SMB space is "one of the bigger sweet spots for Maximus" sales.
In addition to Maximus, Nvidia used SC2011 as a platform to introduce OpenACC, a new open standard for parallel computing programming, along with its collaboration with the Barcelona Supercomputing Center (BSC) to deploy a new hybrid supercomputer fueled by its Tegra ARM CPUs and CUDA GPUs.