 Cloud computing is an essential part of modern technology and it is having a growing impact on our daily life. It is at the core of AI, mobile services, data analytics and the Internet of Things. As these technologies evolve, the cloud services they rely on need to evolve to meet their growing needs. Here at IBM Research, we are rethinking the very foundations on which the cloud is built. We have developed a design and prototype of this aggregated data center. A traditional data center is, put simply, a large collection of servers, each offering certain amounts of resources such as CPU and memory. Applications deployed to the cloud are decomposed into their basic processes, each having its own CPU and memory requirements. These processes must be started on data center servers that have enough capacity to host them. Because CPU and memory are physically confined together within server boundaries, a process cannot draw resources from more than one server. That can lead to fragmentation of spare CPU and memory, leaving unused capacity unavailable to incoming requests. This leads to significant loss in economies of scale. A higher energy footprint and ultimately higher costs for the cloud user. Our concept is to break the physical boundaries of data center servers by disaggregating CPUs and memory into separate physical entities. This transforms the aggregated data center memory to a single resource pool from which any CPU can draw resources. As requests arrive, the system can connect cores to memory on the fly, building platforms that match exactly the requirements. The advantages of memory disaggregation are manifold. First of all, with the same amount of resources, we can serve many more requests. Secondly, memory-centric application can now use all the memory available in the data center instead of being confined to the memory of a single server. Finally, memory and CPU can now be upgraded independently, bringing more agility and modularity to the cloud service provider. Our experimental results so far have been extremely promising and we're working towards improving our technology by building a larger-scale prototype that will pave the way towards the adoption of this technology. We're working on Dreadbox, which is a Horizon 2020 European Commission funded project. IBM Research in Ireland is leading the project, working with 10 industrial and academic institutions across the European Union. As cloud permeates our modern, always-connected world, the impact of this type of breakthrough in cloud's fundamental technology could be really far-reaching. By helping to build a more efficient cloud, we're in line with IBM's commitment to help thousands of businesses who rely on the cloud to build faster, better, and more advanced services for their clients in the world.