Abstract
Scaling of large datacenters requires a significant restructuring of the internal networking. This is needed to ensure higher utilization of the resources (computation, memory and storage) and to minimize the latency in between different servers/racks/pots.
Solution for introducing new technology and new topologies, and providing integrated control and orchestration, is the topic of this talk based on ongoing research in the EU research project cosign.
Revideret
17 maj 2024