Power electronics in data center hardware can make or break the ability to enable an implementation for success or not. A common industry practice is to use simplified assumptions for loading and power conversion efficiency to provide approximate calculations of power utilization. Unfortunately, implementers find huge errors and variability between predicted and actual power consumption, which yield very costly (in CAPEX, OPEX, and time) headaches for many stakeholders. This is due to oversimplification of the highly convoluted and transient nature of power loading in the data center.
This entry‐ to intermediate‐level tutorial will provide an in‐depth investigation into what drives power requirements in data center equipment. The first part will focus on key attributes of data center hardware system power budgets including major loads and their power profiles, correlation between such attributes and how they impact overall power demand (i.e. – how redundancy impacts budget), and specific examples (i.e. – blade server, white‐box, high‐end network switch).
The second part will focus on rolling‐up what was learned about power budgets at the system‐level into the rack, aisle, and complete data center levels to provide the full picture of the data center power solution all the way from the load to the building inputs. This will include many application‐specific examples (i.e. – virtualization, dynamic power allocation, etc.) representative of solutions for today and tomorrow (i.e. – IoT, Cloud Computing, etc.).
Who should attend?
Engineers involved in enterprise power solutions, Program Managers, Data Center Managers, enterprise/networking system designers, SW/FW Engineers, Mechanical Engineers, Technical Sales & Marketing Personnel, and anyone involved with OEM market penetration in the enterprise/networking space.
A prize will be raffled off at the event, all attendees are eligible.