论文部分内容阅读
自从 GPU 于2001年进入可编程时代以来,“能否把 GPU 丰富的资源应用到通用计算中去”一直是业界讨论的话题,毕竟通用计算设计的 CPU 实力有限,而在进行纯 CPU 操作的时候,有浮点专长的 GPU 在一边睡大觉着实是一种浪费。早在 DX8时代,就有研究机构尝试用 Ti4200显卡进行数据库操作,进入 DX9时代以后,此类研究就更多了,不过所有的尝试,都不像一个月前斯坦福大学的 Folding@HOME 那样贴近广大用户,可以毫不夸张的说,在 X1000显卡上实现 Folding@home 是GPGPU 走向大众应用的一个里程碑。
Since the GPU entered the programmable age in 2001, “the ability to apply GPU-rich resources to general-purpose computing” has been a topic of discussion in the industry since, after all, the CPU power of the general-purpose computing design is limited, and when performing pure CPU operations , A GPU with floating-point expertise is really a waste of sleep. As early as the era of the DX8, there are research institutes trying to use the Ti4200 graphics card for database operations into the DX9 era, such studies more, but all attempts, unlike the Stanford University Folding @ HOME a month ago, close to the general Users, it is no exaggeration to say that achieving Folding @ home on the X1000 graphics card is a milestone for GPGPU applications to the public.