A Compute Unified System Architecture for Graphics Clusters Incorporating Data Locality
A Compute Unified System Architecture for Graphics Clusters Incorporating Data Locality
复制标题
结合数据局部性的图形集群计算统一系统架构
DOI:
10.1109/tvcg.2008.188
复制
发表时间:
2009
影响因子:
5.2
通讯作者:
T. Ertl
中科院分区:
文献类型:
--
作者:
Müller;S. Frey;M. Strengert;C. Dachsbacher;T. Ertl
We present a development environment for distributed GPU computing targeted for multi-GPU systems, as well as graphics clusters. Our system is based on CUDA and logically extends its parallel programming model for graphics processors to higher levels of parallelism, namely, the PCI bus and network interconnects. While the extended API mimics the full function set of current graphics hardware-including the concept of global memory-on all distribution layers, the underlying communication mechanisms are handled transparently for the application developer. To allow for high scalability, in particular for network-interconnected environments, we introduce an automatic GPU-accelerated scheduling mechanism that is aware of data locality. This way, the overall amount of transmitted data can be heavily reduced, which leads to better GPU utilization and faster execution. We evaluate the performance and scalability of our system for bus and especially network-level parallelism on typical multi-GPU systems and graphics clusters.
登录
查看更多内容
DOI:
--
发表时间:
2001
期刊:
影响因子:
--
作者:
Ping An;Alin Jula;S. Rus;Steven Saunders;Timothy J. Smith;Gabriel Tanase;Nathan L. Thomas;N. Amato;Lawrence Rauchwerger
通讯作者:
Lawrence Rauchwerger
影响因子:
2.5
作者:
Zhe Fan;Feng Qiu;A. Kaufman
通讯作者:
Zhe Fan;Feng Qiu;A. Kaufman
DOI:
--
发表时间:
2008
期刊:
EGPGV@Eurographics
影响因子:
--
作者:
M. Strengert;C. Müller;Carsten Dachsbacher;T. Ertl
通讯作者:
T. Ertl
DOI:
--
发表时间:
1992
期刊:
Future generations computer systems
影响因子:
--
作者:
H. Bal
通讯作者:
H. Bal
影响因子:
2.2
作者:
M. Rinard;D. Scales;M. Lam
通讯作者:
M. Lam