That, said Intel CEO Paul Otellini, is one of three myths he aimed to bust today in a speech at a Credit Suisse technology conference in Phoenix.

First off, he pointed to the emerging markets. Intel is seeing significant growth in countries like Argentina, where the market for PCs grew 38 percent in 2011; Venezuela, where it grew 34 percent; and Russia, which grew 26 percent. “This emerging market trend is real,” Otellini said. And it’s not likely to end anytime soon: In 2010, the top five PC markets by country were the U.S., China, Germany, Japan and Brazil. In 2015, Intel forecasts suggest, the top five will be China, the U.S., Brazil, Russia and Germany.

With all the attention on phones and tablets, especially Apple’s iPad and various Android devices, the PC has gotten a little stale, Otellini conceded, so it’s time to make it exciting again. Intel’s answer is the Ultrabook. Thinner, sleeker notebooks that boot up fast and have touch-enabled screens and long battery lives will get consumers excited again, he said.

And don’t forget the part that Intel plays in the cloud: Most of the servers running cloud services have Intel chips inside them. And Intel’s data center business has doubled over five years and will double again in the next five. “This is where we see more and more of our customers buying from us direct. They’re building custom boards to run the data centers at Facebook, at Amazon, at Google, at Baidu and Alibaba,” he said.

Myth No. 2: Intel chips are too power-hungry for mobile devices.

Moore’s Law is still alive and well, Otellini said. In 1997, Intel built a supercomputer called ASCI Red that could compute one teraflop. It required 2,500 square feet of space and 9,298 chips to get the number crunching done. Earlier this month, Intel announced a chip codenamed Knight’s Corner that can do a teraflop by itself. In the mainstream marketplace, today’s notebooks are 300 times more powerful than notebooks built in 1995.

Intel, Otellini says, has built its own demonstration Android smartphone to show off the upcoming Medfield generation of its Atom processor, due in 2012. When its power consumption during basic phone functions like things like standby, audio and HD video playback is measured, Intel isn’t the best, but it’s not the worst, either. It usually comes in second or third place when compared against smartphones already in the market, but ahead of others, though Otellini didn’t say which phones it beat and which ones it didn’t.

And on three computing performance benchmarks it beats the others hands down: When using a browser on a phone, the Intel chip smokes the others. It also wins on GLBench, a graphics metric, and SunSpider, a Java test.

Myth No. 3: Intel can’t compete with ARM.

Yes, it’s true that ARM-based chips are everywhere that Intel would like its Atom chips to be and are showing no signs of giving ground.

But, noted Otellini, despite the growth of newer platforms, Intel and its x86 instruction set still command the largest army of software developers — north of 14 million — and the largest body of software created so far — more than 6 million applications. “As we come into these markets we’re bringing an incredible legacy of people that know Intel and know the Intel architecture.”

Then there’s Windows 8. Otellini called it “one of the best things that has ever happened to our company.” It will allow tablets to gain mainstream acceptance, especially in the enterprise that they don’t have today. “A lot of IT managers are worried about security and about porting their legacy applications to an Android tablet or an iPad. What Microsoft is doing is making that seamless for them.”

Finally, he said, Intel has a history of entering markets dominated by someone else and dominating them over time. In the early 1980s as the first Intel chips went into PCs, the dominant machines on the market were the VIC-20 and the Apple II. In the early 1990s, when Intel first went after servers, the dominant chips came from Sun Microsystems and IBM. Now Intel rules both markets. And the same thing has happened in supercomputing. (See the slide from Otellini’s presentation below, though someone needs to check the first field at left because the VIC-20 appears twice.)

Did his mythbusting work? Intel shares rose a bit today, closing up 12 cents to $23.58, and the shares are up 12 percent so far this year. We’ll see how right he was in 2012.

Update: Intel has corrected Otellini’s slide. No longer does it show the VIC-20 occupying two market segments. Well, it was kinda popular.

And while it was Fujitsu, using SPARC chips, that made the top of the list, you couldn’t help noticing how many machines used chips from Intel. Of the 500 supercomputers on the list, 384 of them use chips from the semiconductor giant.

At the SC11 Supercomputing conference in Seattle today, Intel is making some important disclosures about what it is doing to maintain its role as the chip vendor of choice, and also offering its competitive response to a potential threat from the graphics chip specialist Nvidia.

As I’ve explained a few times before, the graphics chips, or GPUs, that Nvidia makes are starting to make some inroads into supercomputing and high-performance computing environments, thanks to their ability to handle floating point computations at a high rate of speed. Sometime next year, at the Oak Ridge National Laboratory in Tennessee, a machine called Titan, using a combination of chips from Advanced Micro Devices and Nvidia, is expected to break the 20 petaflop barrier when it begins operation.

The narrative that has emerged recently is that GPUs are generally better at the floating point operations that are increasingly used in supercomputing — better in many cases than traditional x86 chips from Intel and AMD. Even so, if you add up the number of systems on the Top 500 list using Intel and AMD chips, you’d hit a percentage that’s just shy of 90.

In a presentation today (on what just happens to be the 40th birthday of the Intel microprocessor — hence the two people I saw today outside the “Today” show at Rockefeller Center on my way to work), Rajeeb Hazra, Intel’s general manager of Technical Computing, detailed Intel’s response. First off, Intel is supporting a new technology, called PCI Express 3.0, that will speed up the ability of chips inside a supercomputer to share data. In systems this big, and working on such large amounts of data at once, the processors spend a lot of time tapping their feet and waiting for data to work on. Engineers call this latency, and the point of the new interconnect technology is to cut latency by doubling the bandwidth available. The result is an improvement in the raw FLOPS (floating point operations) available by 2.1 times in lab tests, and a 70 percent improvement in real-world workload tests. In supercomputing terms, that’s real progress, and it effectively means getting answers to big questions faster.

Another advance that Intel talked about today is a chip bearing the codename “Knight’s Corner.” It’s a coprocessor, meaning it’s an additional chip that would be added to a computer to boost its performance. Intel says it can do a full teraflop — a trillion floating point operations a second — and that’s just the result of demonstrations from the first silicon. When in full production, it will probably do even better.

And not only will it do a teraflop on a single chip, it will perform those calculations to what engineers call “double precision,” which is a fancy way of saying the result of each operation will be accurate to a higher level of granularity. As John Hengeveld, Intel’s director of technical computer marketing, told me last week, the rule of thumb in these matters says that moving from single to double precision boosts the amount of time you have to wait by four times.

Why is that important, when an off-the-shelf GPU from Nvidia can do 2 teraflops — though only at the single-point precision? Programming. If you’re a scientist who 10 years ago wrote a program to simulate weather patterns or nuclear explosions or some other classic supercomputing problem to run on systems running Intel chips, there’s nothing new to learn in terms of programming. While the GPUs are great, there are new programming rules to learn.

Finally, Intel is reiterating its plan to keep working on the exascale problem, which is the next great summit in supercomputing. Right now the world’s top supercomputer maxes out at 10.51 petaflops, and a candidate to top the list next year will go north of 20 petaflops, or quadrillions of floating point operations. Sometime this decade — say, about 2018 or so — the hope is that supercomputers will break the exaflop barrier, where machines will run quintillions of FLOPs.

The fundamental problem there isn’t the computing so much as it is power, as in electrical power. Already some of these machines consume as much power as a small city. Getting to exascale will require chips and other components that can run full out at speeds we can as yet only imagine, but doing it consuming a lot less power than they would otherwise be expected to. Think in terms of a Prius that could win the Indy 500 — and not just by a hair, but by a long mile — and do it day after day without really using much more gas than the other cars. It’s kind of like that.

Anyhow, Intel has said that it plans to enable exascale supercomputing that will require only a doubling of the power needed, rather than, say, 10 times as much. To that end, it said today it will open its fourth research lab in Europe. This one is in Barcelona and joins one in Paris; another in Juelich, Germany; and a third in Lueven, Belgium. They’ll all have a lot of work to do between now and 2018.

If you’re a regular AllThingsD reader, you’ve already been introduced to the world’s most power supercomputer: It is the Fujitsu K Computer, which the Japanese computing concern disclosed earlier this month, and it runs in Japan’s quasi-public research institution RIKEN. That’s it in the picture above.

It’s capable of performance as high as 10.51 petaflops, or 10.51 quadrillion floating point operations per second. The same machine had been rated in the top spot on the list before, but was less powerful then, because it was still being assembled, and then capable of only 8.16 petaflops.

The machine is based on SPARC chips — the chips for which Sun Microsystems, now part of Oracle, gained such renown. Fujitsu has been building SPARC chips under license and using them in its own servers and supercomputers for years. In this case, there are 705,024 SPARC64 processing cores in action. And if my memory is correct, the chips in question each have four cores on board, meaning there are 176,256 individual processing chips in the machine.

It’s the first machine on the Top 500 list to venture past the 10-petaflop milestone; however, work is underway in the U.S. on a machine known as Titan, which will supposedly break the 20-petaflop mark sometime next year.

In the meantime, the second most powerful machine in the world is in China. The Tianhe-1A system took the top spot on the list a year ago — and in the process, caused President Obama such consternation about the state of American leadership in innovation that he mentioned it in his State of the Union address to Congress. Its performance reaches 2.57 petaflops and it’s powered by a combination of Intel-made Xeon processors and Nvidia graphical processing units.

In fact, the supercomputers in the top 10 spots on the list are otherwise unchanged from the list released in June.

At No. 3 is Jaguar, the system at the Oak Ridge National Laboratory that is being rebuilt into the machine called Titan, which I mentioned before. It’s a system built by Cray primarily around Nvidia GPUs and Opteron processors from Advanced Micro Devices. Its current performance is just shy of 1.8 petaflops.

The No. 4 system is in China. It’s called Nebulae and is at the National Supercomputing Centre in Shenzen. Its performance is just short of the 1.3-petaflop mark. No. 5 is called Tsubame 2.0, and is at the Tokyo Institute of Technology in Japan.

Chip companies in particular like to crow about the use of their products in the systems that wind up on the list. That makes this a banner day for Intel. Of the 500 systems on the list, 384 of them — 77 percent — use Intel chips. Chips from AMD, Intel’s main rival, are in 63 systems.

It’s a banner day for Nvidia, too. Its GPU chips can be found in 35 systems, more than double the number from the previous list. GPUs were invented to make the graphics in computer games more stunning and realistic; as such, it meant they were, from the beginning, pretty good at performing a certain type of math problem known as a floating point operation. It turns out that the people who run supercomputers do a lot of floating point operations — or FLOPs — too. So as GPUs have gotten more powerful, they’re finding their way into an ever-larger number of the world’s top supercomputers. Two supercomputers on the list use GPU chips from AMD’s graphics chip unit, ATI. Two more use IBM’s PowerCell architecture, which is a sibling of the Cell processor chip found in the Sony PlayStation 3.

President Obama shouldn’t feel so bad about the U.S. not being in the top spot. For one thing, practically all of the systems on the list are built on American-made technology. And among the systems that can reach 1 petaflop in performance or more, the U.S. has five, more than any other country. China and Japan have two each, and France has one. And the U.S. has more supercomputers on the list than any other country: 263. European countries have a combined 127; China has 75 and Japan has 30.

Intel may furnish more chips to the Top 500 list than anyone, but the king of the systems vendors on the list is unquestionably IBM, followed by Hewlett-Packard. IBM built 223, or more than 44 percent, of the machines on the list; HP built 140 of them. IBM also led the performance pack: Its machines are responsible for more than 27 percent of the total. Fujitsu, which made the list-topping K Computer, was in second place, with 14.7 percent. Cray and HP were in a statistical dead heat, with about 14 percent each.

Here’s a link to the full list, and a bunch of other things related to supercomputing.

Today it was Fujitsu’s turn. The Japanese computing giant teamed up with RIKEN, the quasi-public Japanese research institution, to announce that they had built a machine they call the K Computer, which can perform 10.51 petaflops, or 10.51 quadrillion floating point operations per second.

And while all that may sound very impressive, it’s not quite as muscular as the Titan machine being assembled in the U.S. at the Oak Ridge National Labs, which can — or will — perform 20 petaflops.

The machine (pictured) is made up of 864 racks with 88,128 interconnected CPU chips, all of them based on the SPARC architecture for which Sun Microsystems, and therefore Oracle, are best known, though Fujitsu has long been a SPARC licensee. The new K Computer is basically an improvement and extension to the same K computer that took the top spot on the last Top 500 list in June, supplanting in the process a Chinese machine that had taken the crown last November.

Never mind that it contained all U.S.-made chips, the Chinese feat caused the leader of the free world to kvetch about the apparent sorry state of U.S. supercomputing, thus prompting, perhaps indirectly, the Titan machine at Oak Ridge.

It’s not as though China hasn’t been heard from on the supercomputing front recently. Last week its Sunway BlueLight MPP raised eyebrows not for its performance — a relatively pokey 795 teraflops — but rather for the fact that it’s built using all Chinese-made components.

So what will it be used for? Weather simulations, research into drugs and solar cells, and simulating earthquakes and tsunamis.

Here are the more formal descriptions from the announcement:

–Analyzing the behavior of nanomaterials through simulations and contributing to the early development of such next-generation semiconductor materials, particularly nanowires and carbon nanotubes, that are expected to lead to future fast-response, low-power devices.

–Predicting which compounds, from among a massive number of drug candidate molecules, will prevent illnesses by binding with active regions on the proteins that cause illnesses, as a way to reduce drug development times and costs (pharmaceutical applications).

–Simulating the actions of atoms and electrons in dye-sensitized solar cells to contribute to the development of solar cells with higher energy-conversion efficiency.

–Simulating seismic wave propagation, strong motion, and tsunamis to predict the effects they will have on human-made structures; predicting the extent of earthquake-impact zones for disaster prevention purposes; and contributing to the design of quake-resistant structures.

–Conducting high-resolution (400-m) simulations of atmospheric circulation models to provide detailed predictions of weather phenomena that elucidate localized effects, such as cloudbursts.

So what’s a petaflop anyway? A FLOP is a floating point operation. Its a type of mathematical function that involves decimal points. Adding 5.6 and 11.21 is a floating point operation and is therefore slightly more complicated from a computing standpoint than adding 11 and 5. But in computing, even day-to-day computing, it’s massively more complicated than all that.

A top-of-the-line NVidia GeForce GTX 590 graphics card, which specializes in floating point operations, can run about 2,400 gigaflops. Since a gigaflop is a billion flops, I guess that technically puts the GeForce GTX 590 into the teraflop, or trillion-flop range.

Petaflops are then in the quadrillion-flop territory, which as I noted before makes them fun because they’re among those rare numbers that are larger than the U.S. national debt. So 10.51 quadrillion flops gets written like so: 10,510,000,000,000,000. Didn’t I say this was fun?

All this is leading up to a big supercomputing conference starting in 10 days in Seattle. So expect lots more supercomputing news in the coming days!

Krishna Yeshwant, a Google Ventures partner, and Geoff Duyk, a TPG partner, have also joined DNAnexus’s board of directors.

Andreas Sundquist, the CEO of DNAnexus, told me that one of the big advantages of having an investment from Google is access to its computing infrastructure and some of the 20 percent time from Google employees.

Google, he said, will collaborate with DNAnexus to provide access to a huge archive of publicly available DNA information. The archive will take over where the federal government’s National Center for Biotechnology Information (NCBI) is leaving off, after being shut down because of budget cuts.

DNAnexus and Google have teamed up to take over that database and will continue to provide access — for free — to medical researchers. It will now live in Google’s cloud, and researchers will now have a new, easy-to-use interface for accessing it. It represents the largest single dataset ever put on Google’s infrastructure by a third party.

Don’t mourn the government effort. DNA databases are probably better handled by the private sector, Sundquist says, mainly because sequencing a genome, which used to require NASA-sized multibillion-dollar budgets that only big governments can sustain, is no longer so complicated or expensive.

“The reason we started the company is that we started to see that DNA sequencing was getting about 10 time cheaper every 18 months,” he told me. “Ten years ago it cost about $3 billion to sequence a human genome. Now you can do it for about $4,000. It’s like Moore’s Law on crack. In a few years it will be less than $1,000.”

That kind of cost reduction means there’s likely going to be an explosion in the amount of DNA information collected, the kind of surge that Google is uniquely capable of scaling up to manage. “We’re moving from a world where practically no one has their DNA sequences to a world where nearly everyone does, and it just becomes a part of your medical record,” Sundquist says. “The question is, how do you manage all that. It’s one of the biggest and most complex sets of data in the world.”

Answer: The cloud. Think of DNAnexus as sort of a Salesforce.com for people who need access to DNA information. The data will live in the cloud, and researchers will have access to it through a software-as-service model. “DNAnexus is really a DNA data management and analysis platform in the cloud,” Sundquist said. “We’re trying to build database technologies that unlock the possibilities of DNA-based medicine in the cloud.”

Who would pay for it? Anyone who needs DNA sequencing work done: Medical researchers, drug companies, medical doctors. DNAnexus will do the heavy lifting associated with getting the sequencing done. Beyond that, it will manage the ever-growing trove of DNA data and provide all the computing tools that those customers need in the course of doing their work, via a SaaS platform. It already has customers in academia, at places like Stanford University and Harvard University; at pharma companies; and even practicing medical pros in their day-to-day practices, using DNA information to improve their health care and diagnosis problems.

Oak Ridge National Lab's "Jaguar" computer

Despite the fact that the Chinese system was built largely with American-made or American-designed components, the news came as a bit of a blow to American pride, and even caught the attention of President Obama, who kvetched about it in January’s State of the Union address.

By June (the list is updated twice a year) the Chinese machine had fallen to second place, its crown seized by a supercomputer in Japan, relegating the top supercomputer in the U.S. to third place.

Today, the Oak Ridge National Lab in Tennessee, part of the U.S. Department of Energy, will announce plans to build a system that has a good shot at reclaiming the top spot. The machine will be named “Titan,” and its primary computing engine will be the Tesla chip from Nvidia, the company best known for turning out chips that enhance the graphics of games on personal computers.

Nvidia has been making inroads in high-performance computing for some time. Earlier this year I wrote about how the Tesla chips were helping Lucasfilm make movies faster.

I talked with Steve Scott, the CTO of Nvidia’s Tesla business unit, who told me that the Titan machine will be 10 times more powerful than the current Jaguar machine, and that 85 percent of its computing power will come from Nvidia chips, while the remaining portion will come from conventional CPU chips from Advanced Micro Devices.

Why GPUs and not CPUs? It turns out that graphics chips are really good at doing a certain kind of math known as a floating point operation, much faster than a typical CPU chip from Intel or AMD found inside a PC or server.

It’s also an issue of power. For years, as chips and the transistors on them have shrunk, the amount of power required to send pulsing through them has dropped as well. Scott says that is no longer the case. “We’ve reached the point where processors have become power constrained. If you pack all the transistors that you can onto a chip and run it as fast as you can, the chip will melt. We’ve entered a time where performance is constrained by power, and its only going to get worse, so you need processors that are power efficient,” he says. “It’s a fundamental sea change in the underlying technology of high performance computing.”

GPUs, originally designed for gaming and professional graphics applications like editing movies and visualizing complex problems for engineers and scientists, are inherently designed to perform several repetitive tasks at once. In explaining this, I always think back to the old saying “many hands make light work,” though here it’s applied to computing. Two people who divide up the task of folding a pile of laundry get it done faster than one. And four people will get it done faster than two.

Basically, a GPU chip is designed to render what happens to every pixel of a computer screen 50 times a second or even faster. Essentially, lots of small computational jobs are carried out at once. It’s called parallel computing, and, fundamentally, CPUs chips aren’t as good at it as GPU chips. CPUs are better at doing one job at a time, getting it done really fast, and then moving on to the next one. Generally speaking, Scott says, GPUs are about eight times faster at floating point operations than CPUs.

For Nvidia it will be a return trip to the top spot. China’s supercomputing champ, the Tianhe-1A at National Supercomputing Center in Tianjin, which is now ranked No. 2 in the world, uses Nvidia GPUs. This certainly got the world’s attention concerning the potential for GPUs in high performance computing.

The plan at Oak Ridge calls for Titan to have 18,000 nodes, each with an AMD CPU chip coupled with an Nvidia Tesla GPU. Most of the heavy lifting will be done by the GPUs, Scott says. Its total computing capacity will top out at 20 petaflops. FLOPS are floating point operations per second. “Peta” refers to how many the system can do every second: In this case, the answer is 20 quadrillion. Just because I can — and because it’s one of the rare cases where I get to use a number that’s larger than the national debt — I’m going to write that number out: 20,000,000,000,000,000.

And what will it be used for? While many of the Department of Energy’s computers are used to simulate nuclear explosions that are no longer allowed thanks to the Test Ban Treaty, this one won’t be. The mission at Oak Ridge, Scott says, is to advance the boundaries of science. Scientists will use it to model climate change, and to predict the results of different methods of mitigating it. They’ll also use it to design engines, study biology and genetics, and explore the possibilities of using nuclear fusion for energy. If you have interesting scientific work to do that requires this kind of computing oomph, you can even write a proposal explaining how you’d use it.

In the first phase of Titan’s deployment, which is already under way, Oak Ridge will upgrade its existing Jaguar supercomputer with 960 new Tesla chips. In a second phase, expected to start next year, Oak Ridge plans to deploy the 18,000-node Tesla-based system.

Down the road, the hope within supercomputing circles is that performance improves to the point where we’re no longer talking petaflops, but exaflops, or quintillions of floating point operations every second. The government is already working on that, and earlier this year President Obama asked Congress for $126 million in the federal budget to begin research to work on ways to get there by 2018. The biggest problem: How to supply enough electrical power while delivering the computing muscle. Today’s announcement by Oak Ridge is a big step in that direction, but there are still 981 more petaflops to conquer.

Not bad, but I’m certain I’m not the only one hoping that somehow “Saturday Night Live” gets in on the action by reviving its old “Celebrity Jeopardy” sketch from a few seasons back. Perhaps Sean Connery could hack Watson’s programming? You know what would happen.

PREVIOUSLY:

• Done With Silly Game Shows, IBM’s Watson Finds a Job
• All Humans Bow Before the Mighty Watson, Master of “Jeopardy”
• IBM “Jeopardy” Challenge Day 2: Very Different From Day One
• IBM “Jeopardy” Challenge Day One Ends in a Tie
• That Human Vs. Machine Practice Round of “Jeopardy” Didn’t End the Way You Heard It Did
• “Final Jeopardy” Question: Would You Buy an E-Book Without an Ending?
• This Supercomputer Defeated Human Champions of a TV Game Show in 2011
• I’ll Take Computer Company PR Stunts for $1,000,000

Day two was very different. Watson dominated, winning nearly every buzzer and answering nearly every “Jeopardy” clue put to it, correctly. The first segment was all Watson, and it would be like that all night. The computer jumped out to an enormous lead, quickly breaking last night’s tie, soon running up a score of $21,035 to $5,000 for Brad Rutter and $2,000 for Ken Jennings, the scores they had at the close of play last night.

One question about French art stumped all three players, including the computer, and so all three lost the same amount of money on the board.

And at one point, Watson drew an early Daily Double about the designer of Emanuel and Pembroke Colleges. The answer was “Sir Christopher Wren.” Watson wagered $6,435, an oddly precise amount that drew some laughter from the crowd. Watson answered correctly, and the room erupted in applause, and a shot of the crowd showed IBM researcher David Ferrucci, widely seen as the public face of the research team that built and worked on Watson, looking something like a proud father. IBM CEO Sam Palmisano was also visible in audience shots.

At another point, when asked a question about items stolen from a museum in a certain city in 2003, Watson had only 32 percent confidence in what it thought was the best answer, which was Baghdad. It said, “I’m going to guess,” before giving the right answer.

By the end of the second segment, Watson’s lead was bordering on the ridiculous–$36,881, to $5,400 for Rutter and $2,400 for Jennings. One interesting moment occurred when all three players passed on a question about a painting stolen in Argentina in 1987. It all came down to Final Jeopardy. The clue was in the category of U.S. Cities: This city’s largest airport is named for a World War II hero, and its second largest airport is named for a World War II battle.

Here the story takes a surprising turn. And once again we’re fortunate to the have the color commentary of Stephen Baker, my old Businessweek colleague and author of “Final Jeopardy,” a book on the inside story of the IBM Jeopardy Challenge. He witnessed the match in person and spent months reporting on the run up to this event. I recorded our Skype call.

In the audio clip below, Baker starts out describing the strange turn the game took at the end during Final Jeopardy, where Watson displays at once how it can be both stupid and smart at the same time.

Steve Baker Talks About The Day 2 Of the IBM Jeopardy Challenge by ahess247

PREVIOUSLY:

• IBM “Jeopardy” Challenge Day One Ends in a Tie
• That Human Vs. Machine Practice Round of “Jeopardy” Didn’t End the Way You Heard It Did
• “Final Jeopardy” Question: Would You Buy an E-Book Without an Ending?
• This Supercomputer Defeated Human Champions of a TV Game Show in 2011
• I’ll Take Computer Company PR Stunts for $1,000,000

The IBM supercomputer and human player, Brad Rutter, each had $5,000 on the scoreboard, while Ken Jennings, who had bested Watson in the much-publicized practice match, ended with $2,000.

Watson missed some questions and in interesting ways. At one point Watson repeated a wrong answer, the “1920s,” which Jennings had just said. Host Alex Trebeck referred to these as “weird little moments.”

Watson sprang to a huge lead early. By the first commercial break, Watson had $5,200 to Rutter’s $1,000, and $200 for Jennings. It began a serious run interestingly enough after hitting the Daily Double and making a bet. This is interesting in that Watson, in the practice match which it ultimately lost, showed a weakness in situations where betting was called for. This was a weakness that Jennings exploited to his benefit. This made it a surprise when Watson threw down and bet $1,000, more than it had on the board at the time.

I asked Stephen Baker (pictured), author of the forthcoming book on the match, “Final Jeopardy,” to call me after the episode aired for a little color commentary from the point of view of someone who was in the studio to witness it. Our conversation, which I recorded on Google Voice, is below.

Steve Baker Talks about The IBM Jeopardy Challenge, Day 1 by ahess247

Worry no more, Mr. President. Your government is on the case. The U.S. Department of Energy announced today that it has cut a deal with IBM to bring a 10-petaflop supercomputer, named “Mira,” to the Argonne National Lab in Illinois.

Mira is a Blue Gene/Q and it will be up and running in 2012. It’s 20 times faster than the current system in use at Argonne, named Intrepid, which can do 557 teraflops–or 557 trillion calculations–a second, and as recently as 2008 ranked as the third most powerful computer in the world.

Meanwhile, another even more powerful computer, also an IBM Blue Gene Q, is going to Lawrence Livermore Labs next year. This one will be a 20-petaflop monster named “Sequoia.” And there’s more where that came from. These “petascale” computers are helping scientists get their heads around the idea of “exascale” computers that would be faster yet by a factor of a thousand, performing quintillions of calculations per second. (I think a quintillion is 1 followed by 18 zeroes.)

What can you do with 10 or 20 petaflops? Meteorologists could predict local weather down to the 100-meter range with a 20-petaflop system. And running a simulation of how a beating human heart reacts to new medicine, which takes two years of computing time today, will get done in two days on a 10-petaflop system.

Take that, China.

Having bested humanity in chess 13 years ago, the supercomputing scientists at IBM say they’re finally ready to let their latest machine, named Watson, take on the two highest-earning champions of the game show “Jeopardy” in televised games that will air in February of next year.

The game has been in the planning stages for years. The New York Times covered the brewing matchup in a big takeout last year.

Why teach a computer to play “Jeopardy”? The company said it’s all about understanding natural language and detecting the subtle cues of human speech. “Jeopardy” questions can involve clever turns of phrases, riddles and other tricks of speech that can have multiple interpretations. While a computer can make fast work figuring out the best series of moves on the chess board, it’s a much taller order for a computer to answer these kinds of questions.

Take this example, which I borrowed from the J-Archive, a “Jeopardy” fan site: “This city didn’t exist at the time of the Trojan War, so Paris couldn’t have abducted Helen from there.”

A human will know that “Paris” here refers to the son of the king of Troy, not to the capital of France. That’s because we probably know a little about Greek mythology from junior high school, and even if we don’t we pick up a lot from the clue “Trojan War.”

The answer: “What is Sparta?” So, even if you don’t know the precise answer, chances are you can make a good guess by first mentally eliminating answers referring to the capital of France. The trick is in teaching the computer to go through the same process of elimination.

Earlier this year, IBM played a series of 50 “sparring games” against former “Jeopardy” champs. The two human players in the televised games are Ken Jennings (who set the record for winning 74 “Jeopardy” games in a row during the 2004-2005 season, winning $2.5 million) and Brad Rutter (who won $3.6 million–the most by a single “Jeopardy” player ever).

The grand prize for this challenge is $1 million, with $300,000 for second place and $200,000 for third. Rutter and Jennings will donate half their winnings to charity, while IBM will donate all of its winnings.

Playing “Jeopardy” is a good way to push the boundaries on a computer’s ability to answer questions posed in natural language, which IBM says will one day give computers the ability to help diagnose patients in health care settings, improve help desk calls and help tourists find their way around cities.

The machine playing is an IBM Power7 server that the company has optimized with numerous proprietary technologies to analyze spoken questions and then sift through the possibilities that might constitute the correct answer–and do it all within the stiff time limits the game requires.

If nothing else, this will get mainstream television audiences acquainted with the power of supercomputing from the comfort of their own living rooms.

While there was a lot of press coverage of the Garry Kasparov-Deep Blue chess matches in 1996 and 1997, for all the vaunted “man vs. machine” importance attached to it, I don’t recall it penetrating popular culture.

This just might. IBM, of course, hopes so.

Until the match, here is the company’s video on the pending matchup: