Each CUDA-capable GPU node includes local DDR3 SDRAM as well as a 16-lane PCI Express? gen2 interface to the system backplane, providing maximum data throughput direct to GPU memory.

每个CUDA GPU节点包括本地的DDR3 SDRAM以及一个16通道PCI二代系统的背板接口，直接向GPU内存提供最大的数据流量。

youdao

The GPU memory could be considered to be a "cache" of data that the GPU is currently operating on, but many GPU algorithms are designed to operate on more data than can fit in the "cache".

的GPU内存可以被认为是一个“缓存”，GPU目前经营上的数据，但许多GPU算法的设计有更多的数据比操作可以在“缓存”。

youdao

A high-end PC GPU and a low-end mobile GPU can be literally hundreds of times performance difference apart. Same is true even on a single platform.

毫不夸张的说，高端PC和低端移动设备的GPU性能可能相差几百倍，甚至在同一个平台上也相差这么大。

youdao

While desktop GPU revenues were down slightly compared to last quarter, the notebook GPU segment exhibited strong growth.

FORBES: Nvidia Fights Off Intel And Should Be A $20 Stock

Dynamic Parallelism -- This capability enables GPU threads to dynamically spawn new threads, allowing the GPU to adapt dynamically to the data.

ENGADGET: NVIDIA outs a pair of Tesla GPUs to electrify your supercomputer

The NVIDIA Tesla K20 GPU is the new flagship of the Tesla GPU product family, designed for the most computationally intensive HPC environments.

ENGADGET: NVIDIA outs a pair of Tesla GPUs to electrify your supercomputer