NVIDIA working with Microsoft on GPU computing
NVIDIA have been working with Microsoft recently to promote their ‘Tesla’ GPUs for high performance parallel computing applications using the Windows HPC Server 2008 operating system.
“The coupling of GPUs and CPUs illustrates the enormous power and opportunity of multicore co-processing,” said Dan Reed, corporate vice president of Extreme Computing at Microsoft. “NVIDIA’s work with Microsoft and the Windows HPC Server platform is helping enable scientists and researchers in many fields achieve supercomputer performance on diverse applications.”
NVIDIA Research developed several GPU-enabled applications on the Windows HPC Server 2008 platform, such as a ray tracing application that can be used for advanced photo-realistic modelling of automobiles. Related to this, NVIDIA worked with Microsoft Research to install a large Tesla GPU computing cluster and is studying applications that are optimized for the GPU.
“The combination of GPUs and the Windows platform has been a great benefit to our VMD (Visual Molecular Dynamics) user community, bringing advanced molecular visualization and analysis capabilities to thousands of users,” said John Stone, senior research programmer at the University of Illinois Urbana-Champaign. “As we move toward even larger biomolecular structures, GPUs will become increasingly important as they bring even more computational power to bear on what will be highly parallelizable computational problems.”
“The scientific community was one of the first to realize the potential of the GPU to transform its work, observing speedups ranging from 20 to 200 times while using a range of compute-intensive applications,” said Andy Keane, general manager of NVIDIA’s Tesla business. “Researchers are increasingly using Windows on workstations and in data centres due to strong development tools like Microsoft Visual Studio, its ease of system management and its lower total cost of ownership.”
It’s good to see Microsoft finally stepping up to the plate when it comes to GPGPU development, while we’re definitely quite a way away from being able to use the immense power of enthusiast cards to speed up all applications, in a way similar to a CPU upgrade, support for using the massive raw computational power of the GPU is definitely improving. With Microsoft now officially on board we should expect to see the GPGPU features in Windows HPC Server 2008 trickle down into the consumer versions of Windows as time passes.
If you’re interested in NVIDIA’s Tesla architecture you can find out more here
on their official site. While the cards are based on their desktop parts don’t expect to be playing Crysis on any of them any time soon, as these cards are only made to crunch numbers, even down to the lack of any VGA connectors.