Feeds

Intel sees future in Mega Data Center

Hitches self to cloudwagon

Providing a secure and efficient Helpdesk

Intel says that by 2012, Mega Data Centers run by the likes of Google, Microsoft, Amazon, and Facebook will account for 20 to 25 per cent of its server chip sales. Unless the economy is still melting.

If the economy is still melting, Facebook will be dead, and even Google will be hard-pressed to prop up that Intel bottom line.

According to Jason Waxman, the man who oversees Intel's new-found obsession with high-density computing, about 10 per cent of the company's server chips currently find their way into the Mega Data Center. And as the world continues to embrace distributed grid cloud architectures from the net's biggest names, he says, that percentage should increase at least two-fold over the next three years.

"This includes the Googles, the Amazons, the Microsofts, but also telcos doing hosting like AT&T and Verizon and smaller Web 2.0 companies like Facebook," Waxman explained during a mini-press-briefing Tuesday morning in downtown San Francisco. "These are companies that are purchasing thousands of machines a month and putting them into data centers in the mega-watt range."

To satisfy these data-center-happy outfits, Waxman and his Intel cohorts are honing their cloud fluffing skills. In Chipzilla's world, that means optimizing server hardware and software for use in any "shared, dynamically-scalable resource pool".

These pooled server resources might be offered up to the world at large, Amazon Web Services-style ("Cloud architecture," in Intel speak). They might be use to serve up a public web application like Google Gmail ("Cloud services"). Or they might used for a company's internal applications ("Private clouds").

From where The Reg is sitting, "private cloud" sounds a lot like an oxymoron. But marketing types will tell you otherwise.

So, just to sum up, Intel is working hard to optimize the performance and lower the cost of pooled compute resources - whether they're used by the Googles of the world or the customers of the Googles of the world. And because of this, the company says, it deserves some buzzword love.

"We're probably the least well known cloud computing company out there, who has been involved with all these different deployments," Waxman said. "We've learned a lot over the last year and now we want to use what we've learned with technologies and services to help companies lower their total cost of ownership (TCO)."

According to Intel's calculations, if you're running a data center as some sort of dynamically scalable resource pool, nearly 75 per cent of your cash is spent on server hardware and power. "If you're a small IT shop, there's a lot of labor relative to the total cost of the data center," Waxman explained.

"But with a cloud data center, you get so much scale that the cost of labor becomes [much smaller]. And what you're left with is that about 50 per cent of the total cost of ownership...is the cost of the compute infrastructure - server and storage - and about another 25 per cent is the cost of power."

And, naturally, Waxman says Intel can help reduce those costs. The company has even reorganized itself to create a single organization dedicated to cloud fluffing. Its new high-density compute organization is building lower-power, high-density servers. It's fashioning containerized data centers that run at higher temperatures and consume far less power. It's optimizing code.

In other words, Intel is passing on many of the valuable lessons it has learned from Google over the years. Though Waxman doesn't see it that way.

In any event, Intel says it can save the Mega Data Center some serious cash. In a hypothetical data center spanning 50,000 homogeneous servers dynamically allocating resources, Waxman says, the company's cloud fluffing can save $35m over three years: $6m with hardware optimization, $8 with power management, $20m with better optimized code, and $1m with a better data center design.

So, Waxman says, even if the economy is melting in 2012, Intel is still on solid ground. "In tough economic times, companies want to cut costs by leveraging technologies and services."

Ah, but even cheap data centers are ridiculously expensive. And at the moment, the likes of Google, Yahoo!, and Microsoft have significantly slowed their cloud building operations.

In some cases, a buzzword takes you only so far. ®

Providing a secure and efficient Helpdesk

Whitepapers

Secure remote control for conventional and virtual desktops
Balancing user privacy and privileged access, in accordance with compliance frameworks and legislation. Evaluating any potential remote control choice.
Saudi Petroleum chooses Tegile storage solution
A storage solution that addresses company growth and performance for business-critical applications of caseware archive and search along with other key operational systems.
High Performance for All
While HPC is not new, it has traditionally been seen as a specialist area – is it now geared up to meet more mainstream requirements?
Security for virtualized datacentres
Legacy security solutions are inefficient due to the architectural differences between physical and virtual environments.
Providing a secure and efficient Helpdesk
A single remote control platform for user support is be key to providing an efficient helpdesk. Retain full control over the way in which screen and keystroke data is transmitted.