Though it wasn't specifically on the agenda, the topic emerged in almost every discussion about cooling and was even included in the talking points handed out at the beginning of the conference.
"Though water can conduct 3,467 times as much heat as the same volume of air and requires an order of magnitude less energy to move a given volume, the outdated preconception of water cooling methods has limited its implementation," reads a statement compiled by the conference organizers.
The buzz comes after Hewlett Packard this week joined IBM in the water cooled rack business. HP rolled out its product Monday, the HP Modular Cooling System. IBM launched its Cool Blue technology late last year.
Sun, the lone holdout, is doing what many expected -- taking IBM and HP to task for not solving the problem at the server level. "Sun solves the problem at its roots with breakthrough innovation at the microprocessor level, unlike HP and IBM that take a band-aid self-serving approach with more expensive equipment and manpower that still does not solve the customer problem on the long run," claimed Sun in a statement, promoting its new CoolThreads and UltraSPARC T1 technology.
No one wanted to be the first company to jump into the proverbial pool for that very reason, but now that the flood gates are open, people are taking it seriously.
"Water is coming back. From an efficiency standpoint, the closer your cooling gets to the heat load, the more efficiency you're going to get," said Bob Sullivan, senior consultant and cooling expert from the Uptime Institute. According to Uptime estimates, only 28% of air conditioning is actually going where it's needed while 72% of air is just mixing with return air, essentially being wasted.
The issue had gotten so much attention during the meeting that attendee Don Beaty, chair of ASHRAE's technical committee on data center facilities, announced that the organization is currently in the process of writing a book about liquid cooling.
"There is significant interest and passion for liquid cooling," Beaty said. "We need to determine the limits of air cooling and decide what needs to happen."
Beaty said vendors can't afford to fund liquid cooling and air-based cooling research and production at the same time. For development of liquid cooling technologies IT pros are going to have to convince companies that they're willing to buy it.
Users are interested but wary.
"Our facilities guys are talking to engineers. We're checking the pulse to see where water cooling is headed. It's a tactical approach right now," said Dragan Jankovic, vice president of technology at financial giant, Goldman Sachs and Co. "Blasting air from under the floor is very inefficient. Taking cooling to where it's needed is interesting. But it's young. We're waiting to see where it goes."
Ted Hight, senior technical architect at Minneapolis-based Target took it one step further. For 2007, the retail juggernaut is building a brand new, 45,000 square foot data center -- and Hight said the new building will have the capability to tap into chilled water for cooling the racks. "But it wouldn't break my heart if it didn't come to anything," Hight said.
How about opening up data centers to make them less dense?
"That goes back to the CFO walking through the data center and seeing the empty racks and empty floor space and asking why we need another data center. It's a Catch-22," Hight said. "The amount of air cooling you need today is amazing. I think water cooling could reduce energy costs."
The bottom line for many companies is going to be proof of a return on investment. IT shops want data that shows water cooling is actually more efficient. And they want the manufacturers to get past finger pointing and get down to the business of sorting out how this is going to work in a heterogeneous environment.
Let us know what you think about the story; e-mail: Matt Stansberry, News Editor