Jon Toor, vice president of marketing at Xsigo (, says:

Cost: When you include the resources at both ends of the cable – the switch ports and adapter cards – a single cable can represent up to $4000 in CAPEX.
Downtime exposure: Even with the most rigorous best practices observed, a cable is still the physical device most likely to fail and one of the hardest to troubleshoot.
Cooling: Efficient cooling requires unrestricted airflow. A wall of cables impairs that, causing equipment to run hotter and cooling systems to work harder.

Reduce the number of cables to begin with. Virtual I/O technologies can reduce that by about 70%. Before now, each server connection required a separate server cable. That would mean about 8 to 18 cables per server in a virtualized server configuration. Today, you can reduce that to 2 cables with virtual I/O and retain the bandwidth, redundancy, and your I/O connections. You can also change software configuration to prevent pulling cables and reducing the risk of misconfiguration.

When deploying new servers or upgrading network / storage connectivity, consider infrastructure consolidation options that incorporate virtual I/O. Virtual I/O does for the infrastructure exactly what server virtualization does for the servers themselves: it reduces the number of physical components by 70%, cuts connectivity CAPEX by 50%, and lets you complete moves, adds, and changes 100X more quickly. Three years ago, this technology was new on the market, but today it’s proven in the world’s largest data centers, delivering substantial cost and efficiency gains.