The PC industry's two largest graphics companies released new top-of-the-line models this week. The new graphics processors will bring not just better videogame performance, but will also turn ordinary desktop PCs into the equivalent of supercomputers -- if programmers can figure out how to take advantage of the chips' massively parallel architectures.
"We're talking about every man, woman and child basically having a supercomputer on their desk," says Jon Peddie, a graphics-industry veteran and president of Jon Peddie Research.
AMD, which acquired graphics maker ATI in 2006, released two new chips, the Radeon HD 4850 and the Radeon HD 4870. Nvidia, the other dominant player in the space, unveiled its new GeForce GTX 260 and GeForce GTX 280 processors.
According to both companies, the new series of chips feature performance measured in teraflops (that's a trillion floating point operations per second), billions of transistors, hundreds of cores and new architectures that, according to industry analysts, could have a staggering effect on not only Crysis frame rates, but also how and what we use our computers for.
Indeed, cheap access to such formidable computing power could mean that, over the next few years, we will see an explosion of new independent research along with profound new discoveries, analysts say. Additionally, new consumer applications will be able to draw on the graphics processing unit (GPU) for even more eye-watering special effects and even occasionally useful visual information.
"We'll start to get things like real-time mapping from Google that incorporates all manner of real world information," says Bob O'Donnell, an analyst at IDC. "All of this is going to bubble up more and more."
As Peddie observes, it was only 11 years ago that the U.S. government spent approximately $33 million to build ASCI Red, one of the first supercomputers to achieve 1 teraflop. The new graphics chips offer similar power to the 1997-era supercomputer for a fraction of the cost.
"Now we can go down to Fry's or Best Buy and buy a graphics board that has 1 teraflop of processing power for $600 or less," says Peddie.
Getting that processing power to work for the average computer user, however, remains a challenge.
With the exception of a few games, most applications still aren't made to take advantage of the GPU's power. That's because GPUs are made for parallel processing (crunching lots of bits of data at the same time, then assembling the results all at once), whereas most current software programs are written to be executed serially (operating on one piece of data at a time, then proceeding to the next step).
That is starting change, albeit slowly, thanks to new initiatives designed to spur parallel processing.
Just last week, Khronos, the industry consortium behind the OpenGL standard, announced what it calls Open Computing Language, or OpenCL. With this new heterogeneous computing initiative, the group hopes to come up with a standardized (and universal) way of programming parallel computing tasks.
In many ways, it's the Holy Grail developers have been waiting for: a hardware-agnostic standard that unleashes the power of multi-core CPUs and GPUs using a familiar language.
Apple is throwing its weight behind parallel processing too, and last week committed to using the OpenCL specification as part of its next operating system release, Snow Leopard.
Other companies, including AMD, Nvidia, ARM, Freescale, IBM, Imagination, Nokia, Motorola, Qualcomm, Samsung and Texas Instruments have joined the OpenCL working group.
If initiatives like OpenCL gain momentum, the days of researchers applying for grants and traveling across the country to use a given university or research facility's super computer may well be at an end. Similarly, distributed computing projects like Folding@Home and Seti@Home may see an huge boost in performance by using hundreds of thousand of computers equipped with these new powerful processors.
Of course, if curing cancer or looking for aliens isn't your thing, we can also be fairly certain that Crysis will really scream on any system equipped with these new GPUs.
The scary part about taking your sex-tech project to the mainstream is that on the long, hard journey from quirky to safe, you risk wrecking the very thing that made you special.
Then, when the Bowdlerized version doesn't do well, the backlash affects everyone in the sex-tech space, not just the particular application or product. "See?" say the analysts and the venture capitalists and the advertisers. "That's why we don't back sex things."
Then when there's a new sex thing they cautiously express interest about, the developers bend over backwards to show how nonthreatening and comfortable it really is.
Ah, the cycle of romantic startups.
Two years ago, I gave "relationships-management software" Girlfriend X a cautious thumbs-up. Now I'm bummed I can't do the same for its latest incarnation, or for the beta version of its sister site, Boyfriend X.
The genius of the original Girlfriend X and its companion PDA app ("GFX Wingman") was that it indulged in so-impolitic-it-must-be-true irreverence. Marketed as a dating solution for men, it took all those things that women naturally do in our heads and turned it into a database-driven toolset for players -- or those who wanted to play at playering.
For example, its Yield Generator module plotted how much money you spent against how much sex you got, presenting you with a nice cost-per-hookup graph. Other modules sent automated love notes to the right woman at the right time, tracked anniversaries and other milestones for ongoing arrangements, and suggested hundreds of (terrible) pickup lines whose very awfulness could serve to break the ice with new prospects.
After that column came out, Sex Drive readers inundated the developers with e-mail -- at least half a dozen requests, says founder Rick Pierce -- for a similar application for women.
Because yeah, we're good at keeping this stuff in our heads, but parallel dating -- sleeping with more than one person on a regular basis, without those people knowing any details other than, "I'm seeing other people" -- can get complex to keep up with if you do it for more than a month or two.
Unfortunately, Boyfriend X lacks the rueful humor of the early Girlfriend X, managing to be bland and insulting at the same time. (The first question in the Boyfriend X FAQ is "What if I'm not smart enough to get this all working?") And Girlfriend X was neutered on its way from stand-alone software to web-based portal.
Pierce says they wanted to get more serious and to make a sort of "one-stop shop" for relationship management, blending a niche contact manager with a content-driven site.
In other words, there's no Put Out Calculator for Boyfriend X that compares how many bases you've let him touch to how much money he's spent on you, then recommends what you could hold out for next.
The watered-down Girlfriend X is still about creating bad boys out of nice guys, in their minds if not in their actions. But then Boyfriend X warns women to stay away from bad boys and find nice guys.
It may be what the moneymen believe the masses want (without the taint of associating with porn), but it also turns off the very people who might have used the racier version.
No wonder the sex-positive movement despairs of the mainstream.
Girlfriend X and Boyfriend X do have their redeeming qualities. It is handy to have access to dozens of dating and networking sites from one page, and it's great to be able to search all of the reviewed profiles at once regardless of their site of origin.
Both services encourage members to rate people's online dating and social networking profiles (he's short! she's fat! he's married! she's psycho!) and to post positive testimonials for friends and dates. Both provide detailed tracking mechanisms where you can store details about each person and every interaction you've had.
The mashups Pierce plans in the coming months are definitely cool. One compares a person's interests with a local events calendar and the weather forecast, then generates a list of targeted date ideas. Another sifts through the headlines to keep you apprised of current affairs relating to your prospect's interests, so you always have something to talk about.
If they manage to automate more of the data entry -- perhaps an import function so you can bring profile information into your contact manager with a single click -- then X will mark the spot indeed, and the convenience will more than make up for the loss of personality.
I guess I'll have to track my own blow-job-to-oil-change ratio from now on.
See you in a fortnight,
Regina Lynn
Regina Lynn is the author of Sexier Sex: Lessons from the Brave New Sexual Frontier.
1840: Samuel F.B. Morse receives a U.S. patent for his dot-dash telegraphy signals, known to the world as Morse code.
The code Morse devised in partnership with Alfred Vail uses a system of dots and dashes to represent letters and numbers. It went into practical use in 1844, after he and Vail produced a working electromagnetic telegraph transmitter. Vail worked on various refinements to the transmitter before leaving the business altogether in 1848, feeling that he was being low-balled on his salary.
Some scholars argue that it was Vail, not Morse, who actually came up with the dot-dash system. He did hold a small piece of Morse's patent but didn't get rich from it.
Regardless of who devised it, the original code was a little different than the one in use today. What we recognize as Morse code is actually an international variation of the original, or "American," code. The American code contained not only dots and dashes, but also spaces in five letters: C, O, R, Y and Z. (C, for example, was rendered like this: . . .) The numbers 0-9 were also different.
The international version, known as Modern International Morse Code, was introduced at a conference in Berlin in 1851. The American code remained in widespread use until the 1920s, when everyone finally lined up behind the international version.
1840 was a busy year for Morse. An accomplished, respected painter trained in photography, he opened a portrait studio in New York. Morse had met Louis Daguerre in Paris the previous year, and in New York he taught the daguerreotype process to several photographers -- including Mathew Brady, who put it to pretty good use during the American Civil War.
Following a failed run for mayor of New York, Morse turned his attention in earnest to telegraphy. With Vail, he finished up work on the first telegraph transmitter. He spent several years trying to drum up interest in his telegraph, which was met with initial skepticism, both official and unofficial.
When he finally received a patent for the telegraph itself, it came first from the Ottoman Sultan Abdulmecid in Constantinople (now Istanbul), who personally tested it and gave it his blessing. Others, notably Englishmen Charles Wheatstone and William Cooke, had patents on similar (and some say, superior) hardware, but Morse eventually triumphed in the legal battle. His adept promotion, one-wire transmission system and simple software -- the Morse code -- won the day.
Morse code has now been in use for more than 160 years. It still has practical applications in the modern world because almost anything can be used, from telegraph key to flashlight to pencil to fingertip, to tap out or flash a message. Severely disabled people even use Morse to communicate, sending out the code by eye movement or puffing and blowing.
Source: Various
: You don't have to trek out to the dusty hell of Burning Man in order to see inspired feats of mechanical art and engineering. In fact, the back rooms and museums of your hometown may conceal feats of industrial genius that would put any steampunk artist to shame.
Take San Francisco's Fisherman's Wharf. Tourists know it for picturesque views of the bay, vendors selling clam chowder in bread bowls and bad street-corner buskers. But tucked into the corners of the San Francisco waterfront are such marvels as the most advanced mechanical computer ever made, prototypes of a gigantic clock intended to run for 10,000 years and a working steam engine three stories tall.
"It just occurred to me -- the most mechanical geek I know -- that if I didn't put these things together, the rest of San Francisco didn't either," says Alexander Rose, one of the organizers of a one-day, self-guided tour of San Francisco's mechanical marvels.
The tour, dubbed Mechanicrawl and sponsored by the Long Now Foundation, where Rose works, will take place on July 12. Wired.com got an early preview of some of the day's attractions, which include special access to exhibits at the Exploratorium, the Long Now Foundation's offices, a World War II submarine and Liberty ship, and the Musée Mécanique.
Left: One of the stops on the tour is San Francisco's hands-on museum, the Exploratorium. During the tour, volunteer docents will point out exhibits that are particularly interesting to the mechanically minded. Here, kids pedal to generate electrical power in an exhibit built by museum founder Frank Oppenheimer. The generator is mounted on an early 20th-century cast iron lathe.
Photo: Jim Merithew/Wired.com
: Caution: 2,000-Degree Sparks
The Exploratorium's "Catch a Falling Spark" exhibit gives visitors a chance to turn a hand-cranked grinding wheel, spinning it against a thick piece of twisted steel cable to generate white-hot sparks and a distinct odor of burnt clutch. Although the sparks are 2000 degrees Fahrenheit, they're so small that it's safe to let them bounce off your bare hand.
Photo: Jim Merithew/Wired.com
: Loop Dreams
The Exploratorium's "Rope Squirter" is a simple powered flywheel that throws a loop of rope into the air, forming an appealing curve of string that you can play with.
The museum's "head explainer" Ken Finn says he took this exhibit to a meeting of the American Geophysical Union, where it sparked a controversy about whether the shape of the rope's arc is parabolic or not. Of course, the exhibit is equally appealing to children, making it a good vehicle for stimulating mechanical imagination in young and old alike.
"My six-year-old can enjoy it and I can watch geophysicists argue about it," says Finn.
Photo: Jim Merithew/Wired.com
: Visible Sound
The Exploratorium's Kenn Finn shows how the museum's "Oscylinderscope" works: An oversized "guitar" with extra-long nylon strings is set up in front of a spinning drum that has alternating bands of black and white. As the drum spins, you can actually see the vibrating strings' waveforms against the moving stripes. Pluck the strings closer to their middles and you get nice, round sine waves; pluck them closer to the guitar's bridge and you get sharper saw-tooth waves that correspond to the harsher sound.
Photo: Jim Merithew/Wired.com
: Never Needs Winding
At the offices of the Long Now Foundation, visitors check out some of the foundation's recent work. The foundation is designing and building a clock intended to run for 10,000 years -- an engineering challenge that requires designers to anticipate problems like the accumulation of dust and the fact that ball bearings will freeze up if they sit for long periods without moving.
In the middle of this picture is a mechanical orrery -- a kind of planetarium -- designed to show the current positions of the six planets visible to the naked eye.
Photo: Jim Merithew/Wired.com
: Mechanical Binary Computing
The mechanism of the Long Now Foundation's orrery lies underneath the model planets. It consists of a stacked set of geared wheels. The rotation of each wheel corresponds to the rotation of one of the planets in the orrery above.
The orrery's mechanism is a binary mechanical computer with 28 digits of precision for calculating each planetary period. The wheels and levers of each layer comprise a mechanical code for calculating the rotational speed of each planet (for example, 224.68 Earth days for Venus, 11.862 Earth years for Jupiter).
The gear-and-lever design of the orrery resembles that of Charles Babbage's Difference Engine No. 2, although the Difference Engine operates on decimal (base 10) numbers instead of binary (base 2).
Photo: Jim Merithew/Wired.com
: Equation of Time
Some early clockmakers used a kidney-shaped cam to convert between clock time and solar time. That's because the Earth does not revolve around the sun at a constant speed, so solar noon (when the sun is at its highest point) varies from clock noon by as much as several minutes, depending on what time of year it is. The shape of these cams was governed by something known as the equation of time.
The Long Now Foundation's clock uses an equation of time, too, because it resets itself daily based on local solar noon, ensuring continued accuracy over the millennia it will be working. However, the Earth's orbit varies slightly from year to year. For a clock that's expected to run for 10,000 years, those differences mean that a single cam would be reasonably accurate only for a relatively short time (just a few hundred years at most).
To overcome that problem, the clock's designers came up with a three-dimensional cam, whose cross-section gradually changes shape along its vertical axis. This complex, compact shape enables the clock to compute the difference between solar time and clock time for every day over a period of 12 millennia (there's a grace period of 1,000 years on either end of the cam's expected useful life). The numbers along the cam correspond to years (02000, at the bottom, is the year 2000).
The Long Now's Equation of Time cam is available in the foundation's gift shop for $500.
Photo: Jim Merithew/Wired.com
: Torpedo Targeting
Volunteer docent Richard Pekelney shows off the torpedo data computer (TDC) aboard the U.S.S. Pampanito, a World War II-era diesel submarine docked on San Francisco's waterfront.
The TDC was built in 1943. It was -- and may still be -- the most-sophisticated mechanical computer ever made. It used a combination of clockworks, electric motors, dials and levers to compute the angles at which torpedoes should be launched in order to hit their targets.
Torpedo targeting wasn't the only computation-intensive problem at the time. High-powered naval guns developed in the early 20th century proved difficult to aim, because of their long trajectories, the effects of wind and even the Earth's rotation. As a result, research into mechanical and electronic computing proceeded hand-in-hand with weapons research throughout the 1930s and 1940s.
"Most of what we consider early computing was driven by the need to aim these long guns," says Pekelney.
Photo: Jim Merithew/Wired.com
: Stay on Target
The Pampanito's torpedo computer was hand-built in the 1940s in New York, primarily by Jewish émigrés from Germany, says Pekelney.
"What you've got here is the precision of a fine Swiss watch," says Pekelney.
In order to perform its calculations, the TDC incorporated data about the sub's location, bearing and speed as well as those of the target ship. The computation involved multiple differential equations, integrations and mathematical operations.
The TDC resides in the Pamapanito's conning tower, an area of the sub usually off-limits to visitors. However, it will be open to Mechanicrawl visitors on July 12.
For people interested in how the targeting computer worked, the complete TDC manual is available online. Archivists have also digitized rare audio recordings of a successful torpedo attack utilizing a TDC.
Photo: Jim Merithew/Wired.com
: Torpedo Tube
World War II-era torpedoes could make a single turn, shortly after being fired from the sub, so the TDC computed the radius of that turn, then transmitted the setting to the torpedo by means of a remote servo before the torpedo launched. The servo controlled a small rod, which extended into the torpedo tube and connected with a mechanical linkage on the torpedo itself.
This image shows a close-up of the hatch on the back of a torpedo tube. The painted-on flag represents a Japanese ship sunk by a torpedo fired from that tube.
According to Pekelney, submarines were among the most dangerous places to work during World War II, but also were one of the war's most effective weapons. Submariners represented less than two percent of the fleet's personnel, but they were responsible for more than half of enemy ships sunk by the Navy.
Photo: Jim Merithew/Wired.com
: Three-Cylinder Steam Engine
In the berth next to the Pamapanito floats the S. S. Jeremiah O'Brien, a World War II Liberty ship. This vast cargo ship has been restored to working order and the O'Brien now makes occasional fundraising cruises in the San Francisco Bay.
The O'Brien, like other Liberty ships, is powered by an enormous three-cylinder steam engine. It was designed to be very simple to build and very reliable.
This photo shows the engine's three cylinder heads. High-pressure, superheated steam enters the smallest cylinder on the right, then passes to a larger, lower-pressure cylinder in the middle, and finally goes to the largest, lowest-pressure cylinder on the left. This design, known as a triple expansion steam engine design, enables the engine to capture as much of the steam's energy as possible.
At cruising speed, the engine spins at just 76 RPM, pushing the metal hulk through the water at 7 knots. Although the ship will remain docked, the engine will be running during the Mechanicrawl event July 12.
Photo: Jim Merithew/Wired.com
: Wrench Collection
At least 2,715 Liberty ships were built; only a few survive. The Jeremiah O'Brien was restored in the 1970s but, says the Long Now Foundation's Alexander Rose, many of the people who restored the ship are no longer living.
Rose hopes that the Mechanicrawl will inspire a new generation to begin restoring and caring for mechanical treasures like the O'Brien.
"The steampunk crowd, they go all the way to the point of dressing up in period clothing and restoring old steam engines. It would be really awesome if they'd help the Jeremiah O'Brien maintain its steam engine," says Rose.
Plus, then they'd get to play with cool tools, like these enormous wrenches in the Jeremiah O'Brien's engine room.
Photo: Jim Merithew/Wired.com