Intel: How To Build a Better Cloud

At an Intel cloud conference here in Portland, Ore., the chip giant has just described how and why cloud service providers (CSPs) need to make their data centers far more efficient.

Intel General Manager of High Density Servers Jason Waxman kicked off the day, outlining basics of the Intel Cloud Builders Program. But first, he laid out the reasons for its inception. The cloud’s acceleration is increasing to the point where we’ll need 45 new dedicated power plants to handle the growth over the next five years, he said.

Needless to say, that’s horribly inefficient, bordering on outright wastefulness. There has to be a better way to handle all the new users, new devices, and new data that’s being brought to the cloud every day. Enter Intel: Waxman says that together with partners, Intel is working on ways to make a smarter, more secure, more automated cloud that saves money even as it helps save the planet.

After all, Waxman says that a full 75% of the TCO of most infrastructure-as-a-service deployments comes from power and delivery. If Intel can cut that down, they can help lower the cost of cloud services and reap a portion of a data center market that they see growing at a compound rate of 20% every year even as they enable it. Even the smallest things, like matching a cloud service provider’s workloads to the right kind of silicon, can produce a net efficiency gain that can’t be ignored.

But it gets really interesting when you start talking about the relationship between cloud services and the data centers that run them. Waxman mentioned the Intel-facilitated Open Data Center Alliance (ODCA), which is pushing for four key IT challenges in IaaS: Security and compliance; efficiency; lock-in; and manageability.

None of these are news to TalkinCloud readers, but the ODCA is pushing to figure out how to make them things of the past by providing the customer base with a single soapbox.

Intel Cloud Builders Program

Finally, we get back to the Intel Cloud Builders Program. Intel, in conjunction with many major cloud industry players, from Microsoft to Parallels to NetApp to Canonical, to put together reference architectures, essentially giving guidelines on how to deploy technologies from different vendors to create cloud solutions. Some cloud service providers specialize in storage, some in compute, some in POS, and so on.

But Intel has around 50 reference architectures to help them get there. While Waxman warns that the architectures are by their nature not customized to the specific needs of every provider and “won’t solve world hunger,” Intel Chief Cloud Engineer Billy Cox says that they’re based on the company’s own experiences with testing and deploying cloud partner solutions.

On a final note, Cox joked that the cloud has forced Intel to think in terms of services, not technology — not something the chip manufacturer is used to. But all the same, Intel stands to make a lot of money in the data center if they can help their technology partners figure out better, faster, cheaper, and interoperable cloud services. And smaller VARs and MSPs will reap the benefits.

Follow Talkin’ Cloud via RSS, Facebook and Twitter. Sign up for Talkin’ Cloud’s Weekly Newsletter, Webcasts and Resource Center. Read our editorial disclosures here.