GPU Computing wades in to the Mainstream
The idea that most successful technologies become invisible doesn’t yet apply to GPU computing, but its getting there. This week there was a handful of major HPC system announcements based on GPU equipped platforms, but you wouldn’t have known that from the headlines. After knowing about what is GPU No longer the interloper in high performance computing, GPU are beginning to fade in to the background, just like every other mainstream HPC technology. The CPU in the system were even more invisible though, they were not even mentioned. Today though all clusters and workload managers support GPU computing to one extent or another. They have to, given the increasing level of penetration of GPUs in HPC clusters. The idea is to help automate the management of the GPU resources in the cluster so that the system admins don’t have to treat these CPU GPU machines like exotic animals.
By the way even these CPU GPU machines are becoming more common place. Naming convention for them has not quite settled. In mobile graphics technology some are calling them hybrid systems, while other are referring to them as heterogeneous machines. But the preferable is the latter one since hybrid implies a mixing of DNA, which people take to mean the processor’s transistors. Since the GPUs and CPUs are still discrete entities, heterogeneous seems the better nomenclature here. Even the AMD Fusion Chips and future Project Denver processors from NVIDIA, which mix CPU and GPU components on-chip, still seem more heterogeneous than hybrid. But we have a feeling when GPUs are integrated to this level and more importantly when applications are oblivious to the mix of underlying computational units, we will just be calling them processors again. That’s what happens when technology becomes invisible.
But way back in 2009, there were only two designers of graphics card, ATI and NVIDIA. Currently they are regarded as two major players in computer graphics technology. They both have also developed some parallel computing platform. A very common misconception is that these companies actually manufactures graphics cards. In fact, these companies only design the architectures and specifications for the cards to be manufactured. These companies then lease right to produce the cards to other countries such as EVGA, BFG and XFX which physically produce the cards and distribute them. The interface used by a graphics card is the determining factor when ascertaining what is graphics card, an option for any specific computer. The current market leader is the PCI – e 2.0 or PCI express 2.0 interface. Though older systems may still poses PCI and AGP graphical interfaces, these technologies have largely been outmoded and graphics cards are no longer being produced for them.
Processing your request, Please wait....
