14 DAY TRIAL // RenderStream has just announced that its high performance VDACTr8 and VDAC4x2 workstations can now be configured with up to eight AMD Radeon HD 6970, or a maximum of four HD 6990 graphics cards, in order to deliver a staggering 21.6 teraflops.
Both of these machines are built around a dual socket motherboard which supports Intel's powerful X5600-series Xeon processors and come configured with 12GB of system memory.
The rest of the hardware specifications of these systems are also pretty much identical, and the main feature that differentiates between the two is the number of graphics cards that are supported, since the VDACTr8 can be configured with up to eight Radeon HD 6970, while the VDAC4x2 is “limited” at four HD 6990 GPUs.
Of the two, the VDACTr8 is the more powerful solution as the performance of the machine has been measured at 21.6 teraflops by the company.
Compared to the previous version of the VDACTr8, which used eight Nvidia GTX 580 graphics cards, the AMD-powered machine is about 71 percent more powerful, which allows it to process GPGPU-enabled applications at far greater speeds.
In fact, RenderStream has measured that a RenderStream equipped with eight Radeon HD 6970 GPUs can evaluate over 45 billion solutions per second in the integer-based oclHashCat benchmark, versus 18 billion for the GTX 580 based systems
“OpenCL is rapidly a maturing programming language that, with it’s continued development, begins to rival CUDA, enabling high performance computing with AMD GPUs,” said Joe Pizzini, the owner of RenderStream.
“GPGPU high-performance computing using AMD GPUs show great potential for information security, medical imaging, computer graphics and rendering,server side rendering, finite-difference-time-domain (FDTD), electro-magnetics, physics, bio-science and EDA.
“These systems need to be seriously considered in any development plans for GPGPU acceleration,” concluded the company's rep.
No details regarding the pricing of these new machines have been made public until now.