These news roundups tend to have a strong focus on new innovations and bleeding-edge technology, which is great if you’re planning a new data center network deployment.
But what about the vast majority of organizations that have an existing infrastructure to maintain, while staying competitive and keeping up with new developments? This week’s articles all relate to the question of improving efficiency and capacity through integrating new technology into existing systems. This approach means extending the life of legacy infrastructure, while architecting new network and data center infrastructure to open up future opportunities.
DISA News | The DOD’s plans are a great example of an organization that is currently burdened with a lot of legacy systems and installations using a hybrid approach to gradually improve efficiency and expense, according to the restrictions and needs that apply to each area of improvement.
The three tiers of this plan are: Traditional Data Centers, for those workloads which will “always require hardware to function”; On-Premise Private Cloud, for the department’s most sensitive data; and Off-Premise Commercial Cloud Services for less sensitive data and applications. Off-Premise Cloud Services is the area where they expect the biggest savings and efficiencies, but also presents the challenge of establishing appropriate security procedures.
The Register | Verizon’s FiOS has been out there for 10 years now – long enough that the data rates which were once astonishingly high are now normal, and will start looking too slow in the not-too-distant future. Digging up cable to replace it is expensive, time consuming work, however, which is why Verizon is testing an alternate approach to upgrading their fiber network.
Shaun Nichols shares their work with Adtran on technology which will upgrade existing cable to 40 Gbps, simply by replacing the (much more accessible) hardware at either end of the connection. Add in SDN orchestration, and broadband networks are set to become much more flexible and robust in the next few years, with relatively minor tweaks to the infrastructure rather than a wholesale rip and replace.
InformationWeek | Andrew Froehlich describes seven innovations being developed for both existing data centers and those currently under construction, which will make them more efficient and capable than anyone thought possible just a few years ago. Click through to read his take on: Google’s DeepMind AI improving cooling efficiency; Microsoft’s Project Natick (the underwater data center prototype); cloud service providers’ use of SDN; Facebook’s “free cooling” techniques; the growing popularity of edge micro data centers; increasing use of close-coupled cooling; and the ongoing development of directly modulated lasers on silicon (aka DIMENSION), which takes the phrase “onboard optics” to a whole new level.
What do you think? I’d love to hear your examples of using innovation to surpass the former limitations of existing equipment and infrastructure! Leave a comment below, or use the big blue button to send me an email!