Fujitsu Primergy CX1000 Built to Float Clouds

Executive Summary

Fujitsu’s PRIMERGY CX1000 has been designed from the ground up to support scale-out
infrastructures on an industry-standard platform. CX1000 simultaneously addresses the
four main challenges of large scale-out datacenters – i.e., performance, cost,
power/cooling, and footprint – by providing:

• Scale-out performance, delivered with replicable building blocks consisting of 38 servers in each CX1000 enclosure
• A lower cost of acquisition and operations through a minimalistic server design approach
• The Cool-Central™ cooling architecture, which can reduce power/heat energy costs by up to 20% and more
• A datacenter footprint that is up to 40% smaller than traditional systems, enabled by back-to-back setup of enclosures

To reduce costs, the compute nodes in the CX1000 do not have any redundancy functions. Instead, the system is designed to allow failed nodes to be replaced with great ease, leaving most of the responsibility for maintaining availability to software. The CoolCentral™ cooling mechanism in the CX1000 allows many CX1000 racks to be arranged throughout a datacenter in a much smaller footprint than traditional rack servers,
significantly improving the efficiency of scaling out on the platform.

Scale-out computing and x86 hardware play a central role in cloud computing, which promises to change the economics of IT by greatly increasing the flexibility of matching workloads with the resources they need. CX1000 clearly reflects the requirements of cloud computing in its design. While the distinctions between the CX1000 and Fujitsu’s other PRIMERGY platforms may appear to be subtle, they will have a major impact when multiplied over the large-scale deployments required by cloud computing.