The proportion of high performance computing (HPC) sites employing co-processors and accelerators more than doubled during the past two years, and a surprising two thirds of HPC sites are now performing big data analysis as part of their HPC workloads. This is according to the International Data Corporation (IDC) Worldwide Study of HPC End-User Sites.
The 2013 study included sites representing 905 HPC systems, more than double the 488 systems profiled in the 2011 version of the study, IDC announced at the International Supercomputing Conference in Leipzig.
Highlights from the study include:
* The proportion of sites employing co-processors or accelerators in their HPC systems jumped from 28,2% in the 2011 version of the study to 76,9% in 2013. Co-processors/accelerators advanced from slightly more than 1% of all processor parts in 2011 to 3,4% in 2013, with Intel Xeon Phi co-processors and NVidia GPUs running neck and neck for leadership, and FPGAs in a respectable third-place position.
The use of co-processors and accelerators is still wider than it is deep, meaning that these newer devices have entered many more sites but are often still used for exploratory purposes.
* 67% of the sites in the 2013 study say they perform big data analysis on their HPC systems, with 30% of the available computing cycles devoted on average to big data analysis work. IDC did not ask about big data analysis in the 2011 version of the study.
* The 2013 end-user study also confirmed the IDC supply-side research finding that storage is the fastest-growing technology area at HPC sites.
* The proportion of sites exploiting cloud computing to address parts of their HPC workloads rose from 13,8% in 2011 to 23,5% in 2013, with public and private cloud use about equally represented among the 2013 sites.
“The most surprising finding of the 2013 study is the substantially increased penetration of co-processors and accelerators at HPC sites around the world, along with the large proportion of sites that are applying big data technologies and methods to their problems,” says Earl Joseph, programme VP: technical computing at IDC.