White House Going To Launch Its Own Innovation Prize Platform

from the x-prize-me dept

Last year, I spoke with some folks in the White House about various systems for doing "innovation prizes." I got the impression that the whole concept of innovation prizes and encouraging and rewarding breakthrough ideas in more creative means was a top priority among the team I spoke with. So, it's no surprise to hear that the White House is now planning to launch its very own innovation prize platform. The government, of course, is no stranger to innovation prizes, with things like the DARPA's "grand challenge" for autonomous vehicles, and its more recent "find the red balloons" challenge. However, I believe the goal here is to make it easier for other parts of the federal government to quickly offer up compelling innovation challenges. While I'll be curious to see the platform in action -- and how it's used -- it's really great to see the federal government seriously exploring and enabling new and different ways to encourage innovation. It's especially encouraging to see them not just fall back on thinking that "patents" are the only lever they can pull.

Filed Under: darpa, innovation, prize, white house

Former Sun CEO: Tech Companies Suing Over Patents Is An Act Of Desperation

from the indeed dept

Three years ago, after one of Microsoft's regular bursts of FUD claiming that Linux violated all sorts of Microsoft patents, then CEO of Sun, Jonathan Schwartz, wrote up a brilliant post knocking Microsoft down a peg by repeating a line we've said many times here ourselves: real innovative companies innovate, not litigate. Now, following Apple's patent offensive against HTC, and unencumbered of corporate responsibilities, Schwartz is sharing a bit more detail on his views over patents (found via Mathew Ingram).

Schwartz tells the story of Steve Jobs calling him and threatening Sun with a patent infringement lawsuit, to which Schwartz quickly warned Jobs that going down that path would lead to a patent nuclear war, as he pointed out how recent Apple products likely infringed on Sun patents. He then tells another story about a visit from Bill Gates, with a similar threat over patents -- and a similar response, pointing out that Microsoft clearly copied certain Sun technology. In both cases, the counterweight made the threats go away. This is the whole "nuclear stockpiling" scenario -- and, as such, it creates a ton of waste. You have to keep building up those stockpiles just to make sure the other side is too scared to sue you.

But the key point is made after this, where Schwartz again makes a statement quite similar to ones we've made when a big tech company suddenly goes on the patent offensive. It's a canary-in-the-coalmine sign that something is wrong:

For a technology company, going on offense with software patents seems like an act of desperation, relying on the courts instead of the marketplace.

He also highlights how these lawsuits can backfire in a big, big way:

Having watched this movie play out many times, suing a competitor typically makes them more relevant, not less. Developers I know aren't getting less interested in Google's Android platform, they're getting more interested -- Apple's actions are enhancing that interest.

Indeed. It's a point that still seems missed by so many when discussing these patent lawsuits.

Filed Under: jonathan schwartz
Companies: apple, microsoft, sun

The Godfather of Energy Efficiency Gets A Unit

from the offers-you-can't-refuse dept

A group of scientists has proposed a unit of energy efficiency to be named after Arthur Rosenfeld. The Rosenfeld unit will be defined as the electricity savings of 3 billion kilowatt-hours per year -- an amount that is roughly equal to the annual production of a 500 megawatt coal-fired power plant. This efficiency unit is dedicated to Rosenfeld's long career in energy analysis and energy standards which have resulted in billions of dollars in energy savings. While this somewhat esoteric unit may not catch on as a mainstream concept, Rosenfeld also has the Rosenfeld Effect named after him (which describes the phenomenon that California's electricity use per capita has remained relatively constant from 1973 to 2006 while that figure has risen more than 50% for the rest of the US). Additionally, there's also Rosenfeld's Law which states that the energy required to generate one dollar of GDP has decreased by about one percent per year since 1845.

From 1845 to the present, the amount of energy required to produce the same amount of gross national product has steadily decreased at the rate of about 1 percent per year. This is not quite as spectacular as Moore's Law of integrated circuits, but it has been tested over a longer period of time. One percent per year yields a factor of 2.7 when compounded over 100 years. It took 56 BTUs (59,000 joules) of energy consumption to produce one (1992) dollar of GDP in 1845. By 1998, the same dollar required only 12.5 BTUs (13,200 joules).

Rosenfeld conducted much of his research work in California, and U.S. Secretary of Energy Stephen Chu has called Rosenfeld a hero of his. Now, if only the beneficial effects of energy efficiency could spread beyond California and reduce the Rosenfeld Effect.

Computer Science Degrees Still In Demand

from the good-to-hear dept

While there's been some talk that students aren't as interested in computer science these days -- and that getting a job with a CS degree is getting increasingly difficult -- CIO magazine seems to be reporting exactly the opposite. More and more students are getting CS degrees, and one of the reasons is that it's one degree that's still in high demand from employers. The article looks at a variety of CS programs and finds strong demand from both students and recruiters, when it comes to CS degrees -- and that there appear to be lots of jobs available -- despite the doom and gloom of folks who claim that computer science and IT jobs are disappearing.

Filed Under: computer science, cs, degrees, education

Larry Ellison: What Am I Missing About Cloud?

from the partly-cloudy dept

Things have been settling down a bit since the rush of information from Oracle after their takeover of Sun Microsystems was approved. The webcast was a marathon event, and it contained a lot of useful (and re-assuring) information for those of us who deal with Sun on a daily basis.

As usual, Larry Ellison's keynote speech was the highlight of the event. Mad Larry is one of the few CEOs left in IT who feels powerful or comfortable enough in his position to speak his mind -- and then to push hard to deliver his vision. Scott McNealy was another -- it's a mindset that's usually seen in small, feisty startups, and rarely survives that initial growth spurt into global IT giant.

Apart from accusing the SAP guys of being mentally unstable, a key point from the speech was Larry's attitude to Cloud Computing. Oracle has famously rubbished the concept for a number of years, and Larry took the opportunity to once again speak out.

At its core, Cloud Computing is essentially outsourcing -- the old cost cutting measure we've known and loved for over 20 years. Facilities Management, Outsourcing, Cloud Computing -- it's essentially the same thing. All that is old shall become new again.

Larry quite rightly pointed out that Oracle has been doing this sort of thing for years. In fact everyone has -- it's not a new concept at all. The hype machine has been in full swing, though, at a level that's not been seen since 2001 and the first dot-com boom. What irks Larry most, I suspect, is that he didn't coin the phrase, and this is a hype train that he's not driving. With Oracle's history with networking computing and databases as appliances, it's clear where his annoyance comes from -- he's been beating this particular drum for over a decade, and now Oracle are getting stick from analysts for not having their own Cloud Computing strategy.

In part, the hype around Cloud Computing is understandable -- infrastructure can be expensive, good technologists definitely are very expensive, and bandwidth and x86 servers are not only cheap, they're now commodities. It's easy to sign up, slap some code out there, and see if it grows. We'll worry about revenue later -- let's get something live.

Cloud Computing has the danger of becoming a heady mix of both dot-com bubble thinking, and poorly thought-out outsourcing deals. The ease with which you can get infrastructure provisioned is overshadowing the very real dangers that come with weak SLAs, poor performance profiling, non-existent data migration, lack of management oversight -- all the problems that have plagued every outsourcing deal since the first CFO looked at IT and proclaimed: "These guys are a cost centre."

When Larry asked 'What am I missing about Cloud?' -- he was referring to the escalating hype about an old technology. The question, though, is a serious one that everyone should be asking before they get too seduced by the low costs and easy provisioning of Cloud Computing.

Schwarzenegger Calls For Data Centers To Reduce Energy Usage 30% By 2012

from the come-with-me-if-you-want-to-live dept

The Californian government is trying to halve IT expenses, from $3 billion to $1.5 billion. To do so, Governor Arnold Schwarzenegger has issued an executive order to cut energy usage from IT operations by 30% by July 2012 and to consolidate data center space by 50% a year earlier than that. The order also sets goals to increase transparency in IT spending, to standardize IT policies and to develop shared services that are currently spread out over several state agencies.

A few years ago, Schwarzenegger created a cabinet-level agency with authority over statewide IT policy, naming Teri Takai as state CIO. So with this mandate to drastically cut IT expenses, Takai will be shifting all mission-critical and public-facing applications to a new Tier III data center before this September. The state will also close all server rooms housing non-network equipment by June 2013. All existing network services plan to be migrated to a unified California Government Network by July 2010 as well. And in April, state agencies will submit quarterly reports to Takai and start documenting progress towards these goals on a public website.

For all the IT gurus out there, do these measures sound feasible? Are there opportunities for California to become even more efficient with its IT spending? Will the IT spending transparency really mean anything? Submit your comments, we'll keep track of these developments in the future.

Will Bloom Energy Live Up To The Hype?

from the off-the-grid dept

Last Sunday, Bloom Energy was covered by 60 Minutes for developing a fuel cell technology that can produce cleaner energy more efficiently from a variety of fuels. The Bloom Box promises to deliver reliable electrical power to data centers as well as homes, without transmitting power over long distances, since its "energy servers" can be located where the power is needed. After 8 years of development, Bloom Energy has emerged from its startup stealth-mode with 100kW generators that are already operating at Google and eBay campuses, providing electricity at prices in the $0.08/kWh to $0.10/kWh range (average retail electricity costs about $0.11/kWh). According to the press releases, five "parking spot"-sized Bloom Boxes power about 15% of one of eBay's campuses, and Google's Larry Page says that he looks forward to being able to power a whole data center.. someday.

The 60 Minutes piece hinted that Bloom Energy could be a flop like the Segway -- since both Segway and Bloom Energy share Kleiner Perkins as a backer. Beyond that, though, the amount of skeptical analysis for Bloom Energy seems a bit lacking. The story of a secret lab creating a solution to the world's energy problems is a great fiction. But the reality is likely far less inspiring. Plenty of others point out the reliability and cost issues for using a technology that hasn't yet been around for a decade and takes about 5 years (give or take a couple years) to pay for itself from savings in energy efficiency. A 100kW system costs about $750,000 -- so it's a sizable upfront investment for a company to adopt. Additionally, while the system can run on a variety of fuels, Bloom Boxes are currently using natural gas, which is still a fossil fuel with all the associated drawbacks -- even if the power is generated more efficiently.

Ultimately, more competition for generating clean power benefits all energy-intensive businesses. And as some observers note, these Bloom Boxes may help augment other energy technologies -- such as wind or solar -- for more consistent and reliable alternative energy. But there might need to be a much clearer advantage to installing Bloom Energy's off-the-grid generators. Certainly, Bloom Energy has done a great job of getting lots of attention for its technology, but it hasn't proven that fuel cells will revolutionize the economics of energy production. In fact, more efficient use of fossil fuels may actually delay our move away from non-renewable fuels, meaning Bloom would fall short of the hype in more than one way. Instead of a cost-effective alternative to fossil fuel-based energy, it has so far only delivered a somewhat expensive new way to continue using natural gas.

In the end, though, this demonstrates how true innovation is almost always an ongoing process rather than a "flash of genius." Time and time again we hear about amazing breakthroughs coming out of some secret proprietary lab -- but when they're actually revealed, the reality is just another marginal improvement. It's what happens next that's really important. Bloom is getting all sorts of hype for doing something revolutionary, but the result appears just kind of ordinary, at this point. The real question is: can it actually continue the process of innovation to become something that lives up to the hype? It's that process that's really important, not the initial concept.

Filed Under: energy, innovation
Companies: bloom energy

Tape Storage Gets A Density Boost

from the it's-not-dead-yet dept

Researchers in Switzerland and Japan have created an improved magnetic tape material for higher density data storage along with a technique for retrieving the data from these new magnetic tapes. Their method, which uses barium ferrite, can store nearly 30 billion bits per square inch, providing almost 40 times greater performance than regular magnetic tapes. However, such a large increase in data-density creates some problems in the data retrieval process. Electromagnetic interference and residual magnetization of heads can cause data loss. To overcome this, the researchers have developed algorithms to predict (and negate) these effects of interference and magnetization.

Whereas solid state drives (SSDs) cost a few dollars per gigabyte of storage, and hard drives cost a few pennies per gigabyte -- tape storage can cost under a penny per gigabyte. Given the low cost of tape drives, it shouldn't be too surprising that the technology is still around. In fact, in the third quarter of 2009, the overall market for magnetic data tape was valued at over a half a billion dollars. So significant improvements in tape storage could delay the development and arrival of newer storage technologies. According to some tape advocates, tape is a superior technology over other types of data storage when the quantity of data is large enough, and particularly when this data is in "cold storage" -- that is, not accessed frequently. Still, this new tape technology is estimated be 5 years away from hitting the market, and other storage alternatives are not standing still.

Google Awards $1M Research Grant To Reduce Energy Use In Data Centers

from the one-miiiillion-dollars dept

Google has awarded a $1 million dollar research grant to a team of researchers from the University of California at Santa Barbara (UCSB), Rutgers University, the University of Michigan and the University of Virginia in order to develop methods to cut energy consumption in large data centers. This 2-year grant is part of a $5.7 million effort on behalf of Google to provide University funds for researching areas of interest to the company. The research grant will be further divided up to fund a bunch of approaches, and it could be renewed for a third year based on the progress of the teams.

One of the project leaders, Ricardo Bianchini, a computer science professor at Rutgers, claims that many large data centers are only using 20-50% of their capacity, yet they "consume about the same amount of energy whether their workload is low or high." His team is closely looking at ways to put systems to sleep at times of low activity, yet keep their data instantly accessible. Because current computers require the central processing to be powered in order for memory to be accessed, these scientists are proposing a separate power feed to the memory controller, allowing the rest of the CPU to stay asleep, conserving 40-50% of the power consumed by current servers. The research team members will conduct the experiments at the Greenscale Experimental Datacenter, where changes can be made that would not be possible in normal production data centers.