Today, Nvidia CEO Jen-Hsun Huang kicked off the GPU Technology Conference (GTC) with his keynote address. In this presentation, Huang gave some exciting news indeed, including some details about Nvidia's next-generation graphics card cores and a CUDA C compiler for x86 systems.
With AMD's next generation Southern Islands GPUs set to come out soon, it's about time we learned a bit about what's coming next out of Nvidia's camp. Continuing their current code-naming scheme of famous inventors, Nvidia's next-generation graphics cores will be named Kepler and Maxwell.
With power usage and heat being one of the main concerns of the current Fermi cores, one of the main focuses with these new generations is efficiency. In the slide showing the core roadmap, the other axes aside from the year was performance-per-watt.
Looking at this graph we see that Kepler, expected some time next year in the 28nm process, looks to be three to four times more efficient with around six double-precision gigaflops per watt as opposed to Fermi's one and a half. Maxwell, expected two years later in 2013 possibly utilizing a 22nm process, looks to take this well beyond with an impressive sixteen double-precision gigaflops per watt. Huang also noted that Kepler will be implementing a number of unnamed technologies to lower the cursed CPU bottleneck in graphics card performance.
On the software side of things, Huang announced in collaboration with The Portland Group (PGI) the release of the CUDA x86 C compiler. This will allow developers to create CUDA-based application to run on just about any current PC and server system without the need of a Nvidia graphics card. In PGI's official press release, representatives from the two companies made the following comments:
"CUDA C for x86 is a perfect complement to CUDA Fortran and PGI’s optimizing parallel Fortran and C compilers for multi-core x86," said Douglas Miles, director, The Portland Group. "It’s another important element in our on-going strategy of providing HPC programmers with development tools that give PGI users a full range of options for optimizing compute-intensive applications, while allowing them to leverage the latest technical innovations from AMD, Intel and NVIDIA."
"In less than three years, CUDA has become the most widely used massively parallel programming model," said Sanford Russell, general manager of GPU Computing software at NVIDIA. "With the CUDA for x86 CPU compiler, PGI is responding to the need of developers who want to use a single parallel programming model to target many core GPUs and multi-core CPUs."
Discuss in our Forums