Understanding NVIDIA GPGPU Hardware
نویسنده
چکیده
This chapter presents NVIDIA general purpose graphical processing unit (GPGPU) architecture, by detailing both hardware and software concepts. The evolution of GPGPUs from the beginning to the most modern GPGPUs is presented in order to illustrate the trends that motivate the changes that occurred during this evolution. This allows us to anticipate future changes as well as to identify the stable features on which programmers can rely. This chapter starts with a brief history of these chips, then details architectural elements such as the GPGPU core structuration, the memory hierarchy and the hardware scheduling. Software concepts are also presented such as thread organization and correct usage of scheduling.
منابع مشابه
Genetically Improved CUDA C++ Software
Genetic Programming (GP) may dramatically increase the performance of software written by domain experts. GP and autotuning are used to optimise and refactor legacy GPGPU C code for modern parallel graphics hardware and software. Speed ups of more than six times on recent nVidia GPU cards are reported compared to the original kernel on the same hardware.
متن کاملBoosting Java Performance Using GPGPUs
Heterogeneous programming has started becoming the norm in order to achieve better performance by running portions of code on the most appropriate hardware resource. Currently, significant engineering efforts are undertaken in order to enable existing programming languages to perform heterogeneous execution mainly on GPUs. In this paper we describe Jacc, an experimental framework which allows d...
متن کاملAccelerating Data-Serial Applications on Data-Parallel GPGPUs: A Systems Approach
The general-purpose graphics processing unit (GPGPU) continues to make significant strides in high-end computing by delivering unprecedented performance at a commodity price. However, the many-core architecture of the GPGPU currently allows only data-parallel applications to extract the full potential out of the hardware. Applications that require frequent synchronization during their execution...
متن کاملShifter: Fast and consistent HPC workflows using containers
In this work we describe the experiences of building and deploying containers using Docker and Shifter, respectively. We present basic benchmarking tests that show the performance portability of certain workflows as well as performance results from the deployment of widely used nontrivial scientific applications. Furthermore, we discuss the resulting workflows through use cases that cover the c...
متن کاملGPGPU Computing
Since the first idea of using GPU to general purpose computing, things have evolved over the years and now there are several approaches to GPU programming. GPU computing practically began with the introduction of CUDA (Compute Unified Device Architecture) by NVIDIA and Stream by AMD. These are APIs designed by the GPU vendors to be used together with the hardware that they provide. A new emergi...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2013