High performance computing (HPC) around the world is being used to analyse big data, as well as to drive the move towards cloud computing.
These are among the findings of the latest International Data Corporation (IDC) worldwide study of high performance computing end-user sites, representing 905 HPC systems, nearly double the 488 systems profiled in the previous version of the study.
Highlights from the study’s six reports include:
* Processors/co-processors/accelerators report – 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 rather than production computing. Industrial/commercial firms tend to buy fewer of these devices but use more of them in production environments.
* High performance data analysis report – 67% of the sites in the 2013 study said 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 forecasts that revenue for high performance data analysis (HPDA) servers will grow robustly during the 2012 to 2017 forecast period, increasing from $743,8-million in 2012 to nearly $1,4-billion in 2017. HPDA storage revenue will near $1-billion by 2017.
* Storage/interconnects report – The 2013 end-user study also confirmed IDC supply-side research finding that storage is the fastest-growing technology area at HPC sites. By 2017, IDC expects HPC storage revenue to increase to a record $6-billion.
That $6-billion figure would equal the value of the worldwide HPC server market in the year 2000. Within the surveyed sites’ primary HPC systems, Ethernet variants predominated and InfiniBand was a strong second. The percentages of each varied in the sites’ other HPC systems.
* Cloud computing report – 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.
* Applications software report – 64,4% of the respondents’ codes are running on one node or less, 13,3% of the codes run on just a single core, only 5,2% of the applications are being run on more than 1,000 cores, and just 0,9% scale to 10,000 or more cores. IDC forecasts that HPC application software spending will reach $4,8-billion by 2017 and will command a higher percentage of HPC budgets.
* Systems software report – The study confirms that the expanding sizes and complexity of HPC systems, along with their need to operate in new environments, poses substantial challenges for HPC management software (middleware). IDC forecasts that spending on HPC systems software will expand to exceed $1,5-billion by 2017.
“The most surprising findings of the 2013 study are 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, and the steady growth in cloud computing for HPC,” says Earl Josephe, programme vice-president: technical computing at IDC.