With its huge expansion in Las Vegas, Switch is investing in its belief that the infrastructure for cloud computing will be concentrated in huge, multi-building campuses built atop abundant supplies of fiber and power.
The company’s planned 2 million square foot campus in Las Vegas is notable for its scale - it's easily the largest project announced yet. But it reflects a growing school of thought that these "ultra-scale" projects are the best way for data center developers to manage the costs of building Internet infrastructure.
Why are companies building bigger campuses? There are several reasons. One is the expectation that the shift to a digital economy will continue to drive demand for larger and larger amounts of data center space. Internet delivery of media content has transformed the economics of the news and music industries, and similar shifts are underway in consumption of books, TV and movies.
And then there's cloud computing. While debates continue about the pace of adoption for cloud services, there's growing acceptance that cloud services will eventually shift a large volume of IT workloads from in-house server rooms to third-party facilities operated by data center specialists.
Then there's data center economics. Building at scale big allows data center operators to realize economies of scale that can dramatically slash the cost of construction. The key financial metric for data center construction is now "cost per megawatt," reflecting the primacy of power in facility capacity and the structure of leases.
Finally, there's data center design, where the wide-open "barn" layouts of the dot-com boom have yielded to smaller pod architectures and modular data centers. The focus on smaller spaces provides greater flexibility, but also allows data center builders to standardize many elements of the process, enabling an “industrialization” of data center design. conserve capital by building large footprints in phases
Digital Realty can then lower costs in its supply chain through volume orders of these components, and also have a “buffer inventory” of critical items with long delivery timeline
Background, existing campuses built over time by Equinix, DuPont Fabros and Digital Realty Trust. Example of Digital’s Dallas campus as largest campus yet.
The campus is really a superior model. You get a lot of advantages. You get a dedicated substation, you have the anchor tenant concept, and construction costs are spread out. It delivers a new level of flexibility and scale, I think our Santa Clara campus has proven the concept.
Yahoo - the campus approach addresses the difficulty of projecting demand beyond an 18 month time frame.
i/o Data Centers -
With the expansion to 2,000,000+ square feet, the SuperNAP-West campus will deliver an unrivaled:
- 500 MVA of power capacity
- 567 MVA of generator capacity
- 294 MVA of UPS (Uninterruptible Power Supply)
- 202,000 tons of cooling
- 22,000,0000 CFM (cubic feet per minute) of airflow
- Over 31,000 cabinets
- 24/7/365 on-site security, safety, operations, and critical infrastructure personnel
- 500,000 square feet of additional on-net office space
- How much power capacity is currently available at the SuperNAP? Are you building additional substations/power infrastructure to support the expansion, or is the power capacity in place? The capacity of the current SuperNAP is 100MW. We have a total of 250MW currently, with an additional 250MW to be constructed.
Switch is offering a key differentiator for the future of the data center industry: the campus ecosystem. SuperNAP-West is also home to a massive fiber gateway that includes North America’s largest telecom providers.
The increasing demands of the world’s most advanced companies requires the ability to grow without relocating or travelling to a different area of the country. This expansion sets Switch apart as the leading provider of the world’s most powerful data center environments.