source: http://perspectives.mvdirona.com/2010/09/overall-data-center-costs/ (2010)
By now you have heard about Mega Data Centers being built at record pace for mega bucks.
It’s less likely you have heard of or even considered the environmental cost of these data centers, unless of course you fall for the green-washing promoted by the typical large players like Google, Microsoft, Oracle, EMC who claim they are working on it in an impactful way (say 10% of their data center footprint running on renewable energy whether directly or indirectly).
Let’s face it, it’s nice to be green, but can you do it profitably? Today, Monthly Power Bills are anywhere from 18-20% of the cost of a Mega Data Center’s monthly operating charge. Competitive reality says all of the big Data Center Players use OPEX financing models, in order to favorably attract shareholders looking for a dividend, so this means going to debt markets to finance such a distributed hosting service delivery strategy couple with off-grid power strategy for Cloud Services to the B2B and B2C market segments, which means the returns have to be there year 1.
To get started, so the move to micro data center based Cloud Service delivery is profitable in an impactful way, a market player in this space needs to commit to a small 5% foot print conversion to renewable power, and that means a 1 year investment of at least US $1B even for even a small player like Telus in Canada, or Swisscom in Switzerland. Certainly not a “chump change” decision, and what does it really buy them in terms of L2R “Likely to Refer” score improvements with their ever fickle commercial or residential service subscribers, where L2R is THE key investor indicator of market success in the Telecoms subscriber retention game. Amortized over a three year facilities/equipment lease, it means annual revenue must be maintained above $390M/yr for this 1% of the micro DC foot print and edge delivery of cloud Services move to make sense, where the equipment deployed has a useful 5 year life, generally, with the exception being Solid State Drives (the key to fast Cloud Service Delivery and optimal VM density), which on average see these SSDs decommissioned every 4 months (50/50 R/W use).
Reality tells us the efforts of these “World Order” giants, National Leviathans and local state or country Wannabe Telecoms players to reduce power consumption and convert such power consumption to renewable energy sources in order to improve L2R scores, grow stock value and beef up operating margins, is only now becoming a main operating expense line item to be scrutinized intensely as part of the overall strategic plan. Many of these wannabe cloud services players include “slow to move” country and state level former Telecoms Monopoly “spawn” trying to “pull on their big boy pants” and change up their “same old, same old” game in order to survive, let alone thrive in this ever growing and changing market of mega data centers. So far, none have been successful in implementing the above suggested strategy.
So “What’s a poor ‘Telecoms centric’, Mega Data Center wannabe player to do?” Grow bigger with bigger Mega Data Center footprint in order to compete with the likes of world players Google, Microsoft, Oracle, Equinix, and big national players like Rackspace, etc? Well they can try but I say good luck with that.
Hmm? Maybe there is another way.
How about go against the flow, play to their small market strengths and go distributed, with localized modular micro data centers hosting delivery of Cloud Services with their own off-grid generation and storage, using the grid as backup of last resort to improve Cloud Service delivery Access times and raise L2R scores with subscribers, oh yeah, and improve operating margins?
Is it even possible? What new market dynamics need to be harnesses to make it possible?
Will localized Cloud Service delivery from local micro data centers really boost L2R scores, meaningfully impact stock value and improve operating margins?
The answer to the first question is a “qualified” Yes as it is for the last question, provided the progressive CTO gets the backing he needs from the CEO to really scale up micro data center powered Cloud Services, locally hosted and accessed by the B2C and B2B customers found at the edge of their existing service network which connect to smaller secondary markets. Such a deployment of Cloud Services can deliver 2X to 4X better response times and boost L2R scores at much lower cost, due to long term cost savings in power consumption. (In fact excess power can be sold back to the grid to further improve monthly operating margins, where the co-lo operator has enough roof space for large solar and even micro wind power generation).
The 2nd, third and fourth tier players such as Telus, Swisscom et al, looking to carve out decent earnings for their shareholders in the Data Center delivered Cloud Services markets can find success in harnessing the market dynamics of renewable energy, to actually boost higher L2R scores, retain and grow subscribers and, in the process, improve overall operating margins, however, the responsibility for such distributed micro data center success largely rests in the hands of the smaller player CTO and relies on their ability to chart a strategic course of action to CEO backed lower power costs per month. The CTO needs to tailor the deployment and management of Cloud Services in an operationally efficient, cost effective manner which actually plays to their local market strengths and loyalties , by leveraging centrally managed, yet localized execution of lower Cost of Energy “CoE”, which manifests itself as off grid powered or edge grid connected (for backup only), which are primarily battery powered micro data centers capable of hosting dynamically deploy-able, containerized, scalable cloud services as needed, to satisfy subscriber real-time demand, all the while managing the power consumption dynamically and efficiently. When the micro data center is not busy in slow periods or in the middle of the night any excess power accumulated is then sold off to the public grid at a time of day which is most profitable. This means installing potentially more storage that what is required by the micro data center for normal operation to take advantage of selling power at higher prices.
Quick Deploy Containerized Data Center Solutions:
Also the really good news is there are more and more “canned” and even containerized solutions emerging every day, which make this “go against the flow” strategy of localized micro data centers not only a possibility, but a competitive reality which cannot be ignored. Of course none of these solutions are created equal when considering power consumption, so one has to be a careful shopper of Cost of Energy “CoE” when evaluating these canned or containerized micro data center solutions.
The Micro Data Center Secret Sauce:
DevOps Driven Deployment “D3” via Cost Effective Orchestration
Always Software combined with New Deployment Operational Mastery is the key these days where container management within a VM Virtual Machine distributed framework is the order of the day when moving software servers , components and even applications around, and those that master the current tools of the DevOps trade will win, provide they can cost effectively and efficiently orchestrate a collage of vendor equipped deployment capabilities as IaaS, Infrastructure as a Service to squeeze our those better subscriber experiences used to keep stock prices buoyant and operating margins plump.
If you want to cut through the vendor spin of DevOps Orchestration, here are some useful reviews from the DEvOps faithful.
In the End its all about the Power Baby to go against the flow
For those DevOps faithful, power today in the Mega Data Center is taken for granted, “the cost is a given” or, is it?
In fact someone else tends to provide current DC power under sub contract, providing all management and servicing of the power source under the 3 or five year DC co-lo agreement, which by the way, is tough and costly to break, which means there goes your market agility as your competitor’s out grab you in market share, improve their L2R and attract investors their way, because they went against the flow to lower power costs with distributed co-lo-based Cloud Services operating from micro data centers powered by offgrid renewable generation and new low cost energy storage with predictable flat power charges for the next 15 to 20 years.
The fact is, how you, as a Tier 2, 3 or 4 Cloud Services player, provide Power to your Data Center, big or small, is without question, a strategic decision, moving forward. Such a move can become part of boosting L2R scores across 50% of the subscriber base long term, which is done so by simply and effiicently shifting part of your data center build and deployment of co-lo software servers and components back to old local co-lo points and into new local co-lo points of access as containerized Cloud Services operating in remote VMs to service these remote fixed and mobile subscribers.
One can simply strategically commit 1-3% of your renewed or new growth DC deployment per year to become off-grid and grid edge micro data center builds, supplied by 80-100% by renewable energy generation. These new deployments will be powered up with cutting edge Lithium Polymer and Zinc based large format and containerized battery systems, fed by those same renewable generation systems, and backed up by the less than reliable, aging and efficient, legacy power grid, which by the way will become more and more expensive as centralized power generation costs continue to rapidly rise over the next five years.
The off-grid powered Micro Data Center hosted Cloud Services Bottom Line…
L2R goes up, Operating Margins go up, Power Costs Flatten…
Within three to five years, 5% and as much as 15% of your DC foot print will be transformed into micro data center hosted Cloud Services with higher L2R scores and fatter Cloud Service margins based on lower power consumption costs. No guarantees here, but there is also a good likelihood your company’s stock price will float to the top of the competitive heap…
Can you afford to wait? Off-Grid Power Stations for Micro Data Centers:
Profitable Cloud Service Delivery at the Service Network Edge
One can always wait.., for impending doom, or go for it and not only survive but thrive, with the right strategy, energy, and application of capital in the right market location when it comes to micro data center powered Cloud Services delivered at the edge of your service network, leveraging localized off-grid or grid edge renewable power generation and new cutting edge battery storage for power supply.
The existing market dynamic now in full force is the confluence of new price/performance synergies between new Solar and Wind Power Generation technology advances and price reductions and the new low cost Lithium Polymer as well as Zinc Air and Zinc Bromine Energy Storage which now makes micro data center deployments of Cloud Services a reality, so there is no need to wait.
Low cost, Large format Battery Storage has finally arrived, as has cheap small scale Solar and cheap small Wind powered electricity generation, and more and more there are canned renewable power stations employing these advances which can be deployed rapidly to power your localized ” canned” or containerized” micro data centers serving up your most popular Cloud Services.
To find out more about of off-grid and near grid clean power solutions capable of powering your micro data center driven Cloud Services, check out our partner’s site at www.mobismart.ca.
If you need some help designing, sourcing and supporting the deployment of off-grid clean power stations in order to realize your edge deployment of Cloud Services in off-grid powered and/or grid edge connected and powered micro data centers, then send mobismart.ca an email or give them a call, and we would be glad to provide you with a plan and way forward.