by Anton Shilov
07/07/2008 | 06:45 PM
Developers of graphics processing units (GPUs) have been talking about the possibility of general-purpose (GP) computing on graphics chips for over six years now, however, very few actual have been achieved in the direction. Intel Corp. claims that GPGPU hardly has any future and serious intentions to promote it, such as Nvidia Corp.’s compute unified device architecture (CUDA), will never be successful.
“The problem that we’ve seen over and over and over again in the computing industry is that there’s a cool new idea, and it promises a 10x or 20x performance improvements, but you’ve just got to go through this little orifice called a new programming model. Those orifices have always been insurmountable as long as the general purpose computing models evolve into the future,” said Patrick P. Gelsinger, senior vice president and general manager of Intel’s digital enterprise group, reports Custom PC web-site.
Intel knows what it is talking about: it has Itanium microprocessors based the explicitly parallel instruction computing (EPIC) paradigm. Such processors may truly deliver extreme performance, but this requires substantial optimization of software, something, which is why Intel Itanium processors cannot reach desktops or even mainstream workstations.
Since Intel’s forthcoming discrete graphics processing unit code-named Larrabee has many x86 cores inside, Mr. Gelsinger believes that Nvidia’s compute unified device architecture (CUDA) and Advanced Micro Devices close to metal (CTM) initiatives would eventually vanish into oblivion.
“We expect things like CUDA and CTM will end up in the same interesting footnotes in the history of computing annals – they had great promise and there were a few applications that were able to take advantage of them, but generally an evolutionary compatible computing model, such as we’re proposing with Larrabee, we expect will be the right answer long term,” Mr. Gelsinger added.