Mark Rose, general manager and CEO of Bluebonnet Electric Cooperative, was perhaps the most vocal on the subject, saying that he believes a utility’s role was to provide a service, and an open platform upon which consumers and businesses can track their energy from end to end in whatever way they want. But when it came to providing new applications and services that could be built on top of that service, he questioned if the utility was the right entity to provide that.

However, others on the panel clearly disagreed with that vision. Carl Richie, vice president of government affairs at TXU Energy, said that applications and services built on top of the smart grid will be what differentiates providers in a competitive market. The Texas market is a little different than other states, however, and there is some level of competition in the Texas power provider market. TXU Energy has based its business model on reselling energy and offering customers innovative and interesting products, like some of its broadband-based tools.

As a wireless reporter for GigaOM, the utility market reminded me of the telecom world some three years ago, which required developers to go through the phone companies in order to deliver an application on a cell phone to a consumer. That’s since changed, and open operating systems like Google’s Android have made it possible for an application developer to build to a specification and let consumers pick it up if they want, without the phone companies’ involvement.

In the earlier days of telecom, carriers hoped to make money and differentiate themselves by acting as a gatekeeper to products and services that the consumer would get on their network. But let’s face it: The phone companies were never that good at developing consumer-focused value-added services that run on top of the network. Thus cell phone-based data applications never really hit the big time until Apple forced the market open with the introduction of the iPhone. Now, applications are a big business, and carriers are running to embrace them — all while still trying to escape being a service provider that offers an open voice and data platforms.

So if the telecom world is any indicator of how the utility sector will pan out, it appears that if utilities try to tightly manage consumer-facing smart grid applications (like home energy dashboards and online services), they could run the risk of stifling innovation and stunting that market. While that might mean they have to be the dumb pipe of energy, they need to create an open platform so that these tools and technologies can flourish.

There’s another similarity between the two industries: their business models are under threat. Wireless companies, for example, are seeing a huge influx in data use that requires expensive network buildouts, but are still figuring out how to get people to pay more per megabyte to support those network buildouts and existing profit margins.

The utilities on the panel were worried that the creation of a smart grid, which encourages folks to conserve energy and contribute their own renewable power to the grid, would result in utilities selling less power, yielding lower sales and profits. As a result, some of the utilities said they were looking at charging differently for their products and unbundling some of the services they offer.

Something’s gotta change for the utility industry. Both John Baker, the chief strategy officer at Austin Energy, and Steve Hauser, with the National Renewable Energy Lab, talked about there being different types of power (presumably based on the time of day and the power source) in the future, with each type having different pricing. Baker also noted the idea of unbundling electricity service, which would add more risk to a customer’s bill, but would help utilities recoup costs. Consider how wireless carriers charge different rates for texts vs. data plans on a per-megabyte basis, and clearly the two industries have a lot to learn from one another.

Paul Holland, general partner with Foundation Capital, also warned that the stimulus package won’t have a lot of money for startups because it’s designed to generate jobs today rather than in the distant future. Steve Westly, managing partner at The Westly Group and an investor in Tesla, which is seeking government subsidies, argued that the government has a role to play in cleantech, not just for subsidies but also for regulatory mandates and standards.

Westly also pitched Tesla, saying the company had revenue of $14 million last year and was on track to bring in $140 million in the coming year. Ironically, Westly also said he was unwilling to make investments in cleantech startups that will require a lot of capital and take a lot of time to reach an exit. Tesla seems to contradict this, but Westly’s rationale that the returns on cleantech deals will be lower than the thousandfold returns on web and IT-related deals of the last boom makes sense.

Westly was the optimist of the panel, predicting that the market for initial public offerings will open up in the first quarter of 2010 and “perhaps sooner.” Mark Zanoli, managing director at J.P. Morgan, was less optimistic. He pointed out that the reason there are no solar IPO deals today is mostly because the startups need to have manufacturing at scale and gross margins that investors will like. Most simply aren’t big enough. Added to this, the customers of solar panels need money to buy the solar panels that they don’t have today. Zanoli anticipates that in the coming months these solar startups will face a race between when the solar companies get their market and when their money runs out.

He outlined how Microsoft is empowering its employees to think about environmental impact: An employee-suggested switch from Styrofoam cups to compostable dishware cut Microsoft’s annual waste stream by 50 percent. On the IT side, Bernard said Microsoft is focusing on the 2 percent of the world’s energy consumed by data centers and bringing Moore’s Law to them. Utilization of servers is one aspect of that, as well as offering IT professionals metrics to measure their progress with energy consumption. Bernard said only 15 percent of IT staffers have even seen their utility bills.

Taking the role of IT beyond data centers and into building management will also improve efficiency of buildings by about 30 percent Bernard said. About 37 percent of greenhouse gases come from buildings — something the Obama administration is hoping to address. In addition to bigger role for IT, Bernard talked about Microsoft’s efforts to help scientists gather data from a wide variety of sources and mash that up into a usable set of numbers so scientists can study larger issues, such as the widescale affects of climate change on water systems.

Albert Esser, V-P at Dell, trumpeted the computer giant’s ability to become carbon-neutral and explained that it’s difficult to become green, but it will improve your bottom line by saving on costs. Esser’s focus was on increasing utilization to lower electricity demand. He said newer facilities are 40 percent more efficient but if a company focuses on its utilization of servers, they can increase efficiency by more than 200 percent. Rich Lechner, V-P of Energy and Environment at IBM, echoed the focus on utilization, both internally and also with IBM customers, for which IBM builds data centers. He also took it a step further and encouraged  virtualizaiton of networking, storage and desktops.

Better utilization of IT assets and end devices are also an important component of energy consumption that IT can manage, stressed the panelists. Robert Aldrich, senior manager and principle of energy-efficient solutions at Cisco, said the networking company spent $150 million last year on power and hopes to cut that by 20 percent in part through better management of end devices.

Aside from utilization, panelists debated the role that renewables should play in a data center. Christina Page, director climate and energy strategy at Yahoo, touted free air cooling as a way to reduce costs associated with data centers and said companies should focus more on reduction in energy demand rather than using clean power. Esser disagreed, pointing out that as a consumer of several megawatts of power, data center operators can galvanize a local utility to switch to renewables by acting as a guaranteed customer for it.

Tang brought up the difficulty utilities have when it comes to managing intermittent renewable power, and increasing resource loads from devices such as electric cars. Those problems and others will drive the need for a smart grid. The panelists agreed that when it comes to communications on the smart grid, there are some huge problems to surmount. Utilities cover a huge area, need communications that are more reliable than a cellular network, and have to do it cheaply. Whatever communications are used on a smart grid also need to also be open and standards-based, secure and future proofed — a utility cannot switch out its meters each time the technology changes. Eric Miller, chief solutions officer of Trilliant, claims the grid will require multiple communications technologies, and will require a mesh networking structure.

For startups concerned about finding their place in the smart grid, the panel didn’t offer much hope. Tang said the road to a contract will likely run through one of the existing large businesses who supply the power companies. Sunil Sharan, director, Smart Grid Initiative, GE Energy, said GE is working with startups to bring their technology before the utilities, but Eric Dresselhuys, VP and co-founder, Silver Spring Networks, says the current grid is less secure than the future smart grid will be. Tang said the current grid’s protection is “security by obscurity,” and he credited the mix of proprietary standards and equipment for the current protection.

At the close of the panel, each participant was asked to name three innovative startups. Sharan named Silver Springs Networks; eMeter, an electronic metering startup; and Google (GOOG). Miller named Sustainable Spaces, which does home retrofitting; Arcadian Networks, which provides wireless networking to utilities; and SunPower, which makes photovoltaic cells. Karl Lewis, chief strategy officer of GridPoint, named Control4, a home automation company; a variety of grid storage companies; and GM. Dresselhuys also named Control4; Serious Materials, a sustainable materials company; and Naverus, which offers sustainable flight plans.

Koomey’s big story isn’t about electricity use in the data center but how IT affects efficiency in the broader society. One effect is that we’re becoming smarter about how we use resources; the other is dematerialization. “Moving electrons is less environmentally damaging than moving atoms,” Koomey says. Essentially the idea is that it is more efficient to send a PDF rather than a piece of paper. As part of getting smarter, he calls for the electric grid to be modeled on the dynamic data center management software that can track and monitor electricity delivery in real time.

Don’t let things harden into categories: In the early days of the Internet, the consumer devices (phones) and the network were controlled by the same company.  There were divisions in regulatory environments with computers controlled by the Department of Justice and communications regulated by the Federal Communications Commission and divisions between video, voice and data that are now meaningless. So when it comes to energy we need to avoid the same hard categories, such as thinking our focus on corn-based ethanol won’t disrupt our food production.

We used the Internet to build the Internet, and we can use the Internet to build the Enernet: The web can reduce energy consumption in a variety of ways, partially by serving as a proxy for  for travel, as well a backbone of the smart grid. Taking advantage of Silicon Valley for company creation is also important. “It’s easy to teach energy to the entrepreneurs than to teach entrepreneurship to the energy industry,” Metcalfe said.

It is not our goal to darken the Earth; it is our goal to lighten it. We now use more bandwidth than ever; we should turn the idea of conservation on its head and create “a squanderable abundance of cheap and clean energy.” Metcalfe proposes space travel, clean water and removing C02 from the atmosphere as some of the future drivers of energy consumption calling them the “YouTubes of the Enernet,” after the bandwidth sucking online video application that comprises a ton of web traffic today.

Learn from the mistakes that were made: The mistakes of the Internet were the lack of security, guaranteed quality of service and an economic model, which has meant that the web now runs on advertising. Metcalfe calls on the creators of the Enernet to make different mistakes.

There will be bubbles: The Internet bubble predicated the growth of the web and many services today. The global warming bubble that is inflating an energy bubble today should actually be separated from the energy crisis. Solving global warming will not change the need for clean and cheap energy in the future.

He ended by calling on Washington to fund research into clean energy while we create an Enernet that mimics the layered approach and distributed network of the Internet.

The solution in Stark’s talk begins with the understanding of each person’s energy use, the creation of what he calls an “energy identity.” To create such an identity there are open standards that should be created (AMEE is trying) as well as privacy concerns that must be addressed, given the granularity of the information that will be shared. Such information could include your purchases, your energy consumption habits such as when you watch TV and what time you turn out the lights and go to bed. Securing such data will be key as well.

“Addressing some of the real privacy issues up front is absolutely critical,” Starks said, drawing a parallel to the issues surrounding the privacy issues with our digital selves on the web. The goal of such surveillance will be to then offset and reduce such energy use while getting people to change their lifestyle from a consumption-based one to a services oriented one that uses fewer resources.

Finally, we can’t just focus on carbon, Starks warned; we have to consider that we’ve hit the peak of much of our resource consumption.

So if this is indeed true, what do we need to do?

Move from a product society to a service society, Starks said — redefine our lives based on value.

When asked for details, a spokesman for Applied said, “While we haven’t officially announced any products, our EES (Energy and Environment Solutions) group is looking at various ways we might participate in the solid state lighting area. Nothing to talk about really yet, other than it shares the same dynamic as many an emerging markets: huge potential, currently too costly for consumers and looking for a technology solution that involves Applied’s thin-films expertise.”

Applied, which spent more than $1 billion buying solar equipment companies over the last few years, has bet big on the space, and is the largest supplier of thin-film and PV manufacturing equipment in the world. However, the solar section isn’t yet profitable, in part because of charges related to acquisition costs. Adding lighting to its portfolio might be risky up front, but the opportunity for organic light emitting diodes could be large if production costs go down.

Last year, Applied Ventures, participated in a funding round for Group IV Semiconductor, which makes solid state lighting on silicon — the same material most semiconductors are made on. In addition to the funding, Applied has offered to help Group IV develop a process for placing its light-emitting films on silicon wafers, which plays to Applied’s thin firm and chip equipment expertise.

Applied is also participating in the lighting industry through the European Organic Phosphorescent lights for Applications in the Lighting market 2008 Program. The organization aims to develop an OLED production technology capable of achieving the cost target of a few euro cents per square centimeter for a white OLED device. Members of the consortia include Aixtron, Osram, Philips and BASF.

Brand spent all but two minutes of the 30-minute speech highlighting how technology — from supercomputers to biotech — will aid in the capture and refinement of fossil fuels at a commercially acceptable price. Of course it’s no surprise that Brand would focus on how to bring down the cost of using fossil fuels for power — ConocoPhillips is the second-largest oil and natural gas refiner in the U.S. and the fifth-largest non-governmental controlled refiner in the world.

Faster supercomputers make it easier to analyze seismic data quickly to “see” where to dig, he said, and biotechnology could create bacteria to rid fuels of contaminants prior to extraction. Nanotechnology was also given a lot of play for its theoretical value, but other than saying the oil giant plans to buy an IBM Blue Gene supercomputer, Brand didn’t talk about specific efforts at the company. He did, however, note that ConocoPhillips has two ethanol refineries and said it’s interested in later generations of the technology rather than corn ethanol.

When asked about algae and geothermal, Brand said ConocoPhillips was interested in algae and revisiting geothermal programs the company had started in the ’70s. Beyond the role technology could play, Brand said the government should get involved. He called for comprehensive federal legislation that would tackle both climate change and create a national energy policy. As for the legislation:

• It should broaden the diversity of supply, bringing in solar, biofuels, etc.
• It should call for greater energy efficiency
• It should include environmental stewardship, notably a focus on carbon emissions
• It should encourage innovation, particularly on the technological and research and development side

So this is all pretty standard stuff from a company that makes its money from the discovery, extraction and refining of fossil fuels. Other than recognizing that a greater diversity of energy sources means oil companies will find ways to differentiate themselves from one another, it sounds like it’s business as usual for ConocoPhillips.

Microstaq is a manufacturer of microelectromechanical machines (MEMS), silicon-based chips that combine that digital information with the analog world (such as the microphone inside your cell phone that translate your voice into digital signals for wireless networks). The Microstaq MEM is a tiny, computer-controlled valve that aims to replace the larger and more expensive expansion valves used in air conditioning and refrigeration (shown at left). It’s more accurate than non-electrical valves and can keep the temperature better regulated, leading to the efficiencies.

Microstaq has attached the chip to a larger casing, so the MEM is easy to drop into place without re-engineering the entire air conditioning unit (shown at bottom). The 8-year-old company, which to date has raised $12.5 million, says it has signed deals with three of the top five U.S. air conditioning makers but cannot name them. It also plans to go after the industrial refrigeration market. Sandeep Kumar, Microstaq’s CEO, says the cost of the chip will be equivalent to the cost of the valves already in use today, which are about $15.

He doesn’t expect to see these valves in residential refrigeration systems anytime soon, because they are currently cooled with capillary tubes that cost about $3 — making the chip-based solution way more expensive — unless a government mandate changes things.

In the meantime, the firm is also working to put its valves inside automobile transmissions in a bid to improve fuel efficiency and lower the costs of providing a smooth-shifting manual transmission. The medical diagnostics market can also use the devices to reduce the size and cost of machines that test blood. Other uses, such as those that involve insertion in the body, are possible as well, but would require FDA approval.

Hopper’s vision combines cloud computing and renewable energy: He wants to take electrical transmissions costs out of the equation by placing a data center directly at the site of a renewable energy source and use fiber optic cable to link it to the entity that uses it. Hopper is also the co-founder of Level 5 Networks, which was bought by 10 Gigabit chip maker SolarFlare.

Virtualization and fast Ethernet, which enable services such as Amazon’s EC2, will make Hopper’s idea feasible. The ability to separate the hardware from software through virtualization is what enables computing clouds to exist. Those clouds allow companies, developers or anyone with the ability to tap into that resource, to ship its computing jobs over to Amazon’s servers, no matter where they are located in the world.

The challenge is figuring out how to build software that can monitor electrical generation, prioritize compute jobs and then figure out when and where to send those jobs based on whether the wind is blowing or the sun is shining. Hopper believes it would make sense to attach data centers, possibly in a container (as shown in the image), directly to a renewable energy source. The source could be located in the middle of a desert, on a platform attached to an ocean wind turbine, or anywhere else where power could be easily generated.

One of the issues with renewable sources of energy, is that the places where it’s most abundant, such as winds blowing across the ocean or solar power in the desert, are inaccessible and thus, expensive to attach to the electrical grid. An example is Texas’ $4.9 billion plan to bring wind energy generated in the barren, western part of the state to the more populous center. Bringing the data center to the power solves that problem as long as the area can be reached via a cheaper fiber optic cable. There are still issues of servicing such remote data centers, but the plan to have multiple ones around the world offers redundancy.

But there are still computing tasks that need to be worked out if this vision is to materialize in the next decade or so. “If it turns out you’re chasing the energy and copying a lot of data, then that’s less attractive,” says Hopper. “But with good caching, and if you’re only moving the data once or twice it, might work. You could design software similar to old-fashioned job scheduling on a mainframe. Back then the scarcity was the computing and today it’s energy.”

As data centers take up more and more energy, Hopper’s ideas may help the computing industry solve one of its fastest-growing problems.

image of a theoretical modular data center installation courtesy of Sun Microsystems

This dollar figure represents a slight increase from the first quarter, when $870.9 million went into 60 deals, and a whopping 62 percent rise over the $544.4 million dollars that went into 53 companies during the second quarter of 2007. And the top two deals involved clean technology companies: OptiSolar, which captured $132 million (and far more than the $38 million we had noted back in April), and BrightSource, which raised $115 million.

Notably, Thomson Reuters changed its definition of cleantech this quarter to include more companies; it also changed the historical data to reflect the category change.

Cleantech companies are faring better than the overall venture climate, which stayed flat from the previous quarter and flat compared with the second quarter in 2007. Venture capitalists invested $7.39 billion in 990 deals in the second quarter of 2008, while in the first quarter, $7.5 billion was invested in 977 deals. For the second quarter of 2007, VCs placed $7.37 billion into 1,033 deals.

Cleantech is a growth area for venture firms; it comprised half of the industrial and energy segment deals in the latest three-month period. Trevor Loy, a managing partner with Flywheel Ventures, a seed-stage firm focused on clean tech and hardware deals, said the investments are now moving beyond renewable energy into categories such as clean building trends, petrochemicals and devices that use less energy. Water is also a big interest for Flywheel as the firm believes it will grow in importance as a resource over the next 10-20 years.

Or at it will least next year, says Steve Cummings, director with Hewlett-Packard, which today launches its data center in a container called HP’s Performance Optimized Data Center or, POD. From a power efficiency standpoint, these PODs are 38 percent more efficient than the average data center, partially because these containers can run hotter. Some customers will use these PODs as a stopgap for gaining compute power while the company builds out new data centers, while others will likely buy these on an as-needed basis.

HP’s POD or Sun’s Blackbox data centers literally come in self-contained shipping containers (HP’s are of the 20-foot or 40-foot variety). HP’s largest version can carry as much as 12 terabytes petabytes of storage and contains 3,500 compute nodes. That’s using HP’s gear, which you don’t have to do if you don’t want to. HP says it is using a standard chassis and layout to make deploying this thing as customer-friendly as possible.

The PODs also not to be confused with some of many data-center-in-a-box products, which are essentially pre-loaded and configured racks of servers that are shipped out ready to install in a data center.

Chip makers are always striving to cram more transistors onto a chip in order to increase the processing abilities of a semiconductor. But at the same time, chip makers are also trying to put more chips on a wafer, so there’s a need to make the chips themselves smaller. It’s like putting more and more penguins on a shrinking iceberg.

The field is known as nanolithography. Lithography, where lines are etched or imprinted on chips, is a difficult problem in the chipmaking world as chips shrink and more transistors are expected to fit on the semiconductor. Currently Intel is making some of its high-end laptop chips at 45 nanometers and expects to go to 32 nanometers by 2011. The MIT research demonstrates the ability to etch such lines across larger surface areas — which would be good for PV cell manufacturers — and without some of the chemical treatments currently being used by others in nanolithography.

The MIT technique uses sound waves to control how a laser moves across the surface of a circuit while the laser is etching the lines. Other methods of lithography use equipment from startup Molecular Imprints, which imprints lines as small as 20 nanometers in a manner similar to a waffle iron or slower, electron-beam lithography efforts.

Image courtesy of MIT.

ABI Research VP Stuart Carlaw said in a release touting the report:

“Although reducing power consumption provides good ecological credentials for carriers and vendors alike, the real driver for improving power consumption is financial.”

The report suggests carriers should work on reducing power consumption by the base stations inside each cell tower that transmit and receive wireless signals from mobile phones. Such efforts could save carriers a sizable 43 percent on power costs from 2007 to 2013 and will start to make a dent in spiraling operations costs by 2012.

Other greener options include using alternative energy, such as solar or wind, to power cell tower sites rather than conventional electricity or diesel generators.