In the current paradigm, when a company needs to run a critical component of their business logic at optimal speed in the cloud, they have three options:

1. Scale-up (by upgrading processors at the node level)
2. Scale-out (by adding nodes)
3. Manually implement code optimisations specific to the underlying hardware.

E2Data proposes an end-to-end solution for Big Data deployments that will fully exploit and advance the state-of-the-art in infrastructure services by delivering more performance from fewer resources. The E2Data stack will achieve this by dynamically profiling, compiling and optimising code for execution on chosen devices, such as CPUs, GPUs, FPGAs, and others. By removing the need for developers to craft specific device code in languages like CUDA or OpenCL, E2Data will create substantial savings in developer time, while still exploiting the power of diverse device architectures, such as are currently offered by Microsoft, Amazon and others.

Over the course of the project, testing will be conducted on both high-performing x86 and low-power ARM cluster architectures representing realistic execution scenarios in four resource-demanding applications : finance, healthcare, green architecture, and security.