IDC forecasts that Big Data–related server shipments will increase from 6% of all servers shipped in EMEA in 2015 to 16% by 2019, and server values from $1 billion in 2015 to $2.7 billion by 2019.
Big Data storage capacity share of new shipments is expected to reach 20 exabytes by 2019, with a value of $2.7 billion.
According to our market sizing, 134,000 server units were shipped for Big Data purposes in 2015, and 764 petabytes of storage capacity deployed, with the majority being external storage.
While most current Big Data projects are starting off in companies' own datacenters, analytics workloads are increasingly being moved to the public cloud while sensitive data needs to remain on-premises in many cases for compliance reasons. IDC expects the public cloud infrastructure share of Big Data workloads to increase from 13% of server shipments in 2015 to 34% by 2019, and new storage capacity deployed on public cloud infrastructure to increase from 25% of Big Data workloads in 2015 to 55% by 2019. Most customers are expected to deploy some form of hybrid solution.
"Big Data and analytics have risen to the top of executives' and developers' agendas as the technology has evolved and mindsets are starting to change in organizations in EMEA," said Andreas Olah, senior research analyst, European Datacenters and Big Data, IDC. "The main challenge is not the data or its volume, but the ability to generate value from it. Many customers are still at the beginning of their journey and still don't know where to start. Others have high ambitions and clear ideas but are slowed down by increasing complexities and the lack of highly skilled data scientists and developers."
Big Data can no longer be ignored in European organizations in view of heightened competition from disruptive market entrants, Olah added. "While a lot of the focus is on choice of applications, it is crucial to get your infrastructure right to unlock and merge various data sources for value creation while avoiding running out of capacity or going over budget," he said. "Data and analytics-focused workloads require a different infrastructure setup than traditional applications with features such as in-memory computing, large storage pools attached to servers, linkages to cloud resources, and denser architectures for greater efficiency."