Home Is Where the Data Is

For those with large troves, the cloud may not be ideal.

The cloud era is still booming, as is clear from the market’s double-digit growth rate. Yet a backlash is already emerging: Companies are showing renewed appreciation for what’s known as on-premises computing. The recognition that it just doesn’t make sense to hand over certain tasks or data to Microsoft or Amazon.com is creating opportunities for companies that promise to deliver the efficiencies of the cloud from the comfort of your very own data center. “We’re getting to a phase in the hype cycle where there’s a lot more rationality setting into the market,” says Ed Anderson, cloud analyst at Gartner. “You can and should ask hard questions about if I move that workload to a cloud, will I get a meaningful benefit from that? And in some cases, the answer is probably not.”

The Allen Institute for Brain Science in Seattle generates petabytes (1 petabyte equals a million gigabytes) of data through its work creating 3D maps of brains—it started with mice and has made its way up to humans. Last year the research operation costed out a move to the cloud with Amazon Web Services. It turned out the switch would have been five times more costly than keeping the data stored in-house because of the sheer amount of information involved, the difficulty of transferring it, and the speed lost in having the data and related processing located off-site, says Chinh Dang, the institute’s chief administrative officer.