USB ports consume power – not a great amount of power until you consider the cost of powering 70,000 servers, each with the USB port active.
Combined with an active PCIE and SAS Storage controller, the power bill for those 70,000 servers over three years is USD 2.2 million (assuming $.012 per kW hour). This is the kind of insight you get when managing hyperscale datacenters operations – the smallest settings become significant over such a large footprint. .
Here are the reasons why you should care about the inclusion of innovative software that boosts your software-defined infrastructure and addresses these kinds of power-related issues in your datacenter and datacenter system software.
You don’t want to waste money
As the example above shows, you need to optimize settings with scale to save money. Modeling shows potential TCO savings of over 50 percent by empowering personnel and lowering costs in required services, software and hardware.
You don’t want to guess
Good decisions are always dependent on good data. You need software designed to discover existing hardware and associated configurations across an entire datacenter footprint. This enables big data analysis and automated management with scale.
You want to to monetize your investment quickly
Once you buy equipment, you want it operational in order to start payback on the investment with the smallest possible lead time. Today equipment is ordered, and then it sits on docks waiting for humans to rack and stack, cable, configure and make available.
By automating the discovery, configuration and operation via the command and control software, the moment a rack is online it is operational and available for allocation.
These are the problems that the San Francisco-based company NodePrime was formed to solve. We announced the intention to acquire NodePrime on the 5th of April.
Taking big ideas and scaling them
NodePrime is a small company with 15 or so employees. But they are only small in number – their ideas are big. And Ericsson has always taken big ideas, industrialized them, scaled them and taken them to over 180 countries.
We have been working with NodePrime to integrate their software into our Hyperscale Datacenter System 8000 software for over a year. The software has been designed to run datacenters – the kinds of datacenters that Twitter, Facebook and Amazon run, that have millions of servers not hundreds and that we call hyperscale, delivering the highest asset utilization and highest efficiency.
Why are you not hyperscale today? We ask the same question….