processing
Latest
NVIDIA's AI will help USPS handle packages 10 times faster
The US Postal Service (USPS) delivers an estimated 146 billion pieces of mail annually, including 6 billion packages. In an attempt to process package data more efficiently, USPS is experimenting with AI. Today, NVIDIA announced that it will provide USPS with its AI tech. NVIDIA claims its system will process package data 10-times faster and with higher accuracy.
Google backs its Bristlecone chip to crack quantum computing
Like every other major tech company, Google has designs on being the first to achieve quantum supremacy -- the point where a quantum computer could run particular algorithms faster than a classical computer. Today it's announced that it believes its latest research, Bristlecone, is going to be the processor to help it achieve that. According to the Google Quantum AI Lab, it could provide "a compelling proof-of-principle for building larger scale quantum computers."
Clever photog builds a fully-automatic photo processor
German photographer Lukas Fritz has spent the last year developing an easier means of processing his film negatives at home, without the need for a dedicated dark room. He recently published a video of his completed work, the Filmomat, and it is nothing short of genius.
Future PS4 games will have more power to play with
It seems the PlayStation 4 is now just a little more powerful than before. At least for game developers, anyway. None of the console's components have changed -- instead it's being reported that a seventh core has been "unlocked" in the CPU. Until now, six of the PlayStation 4's eight-core CPU have been dedicated to games, while the remaining two handle the operating system. In all likelihood, this was a conservative setup to ensure consoles ran smoothly at launch. It also gave Sony some wiggle-room if they needed to make any system changes or optimizations. Now, however, it seems the company is happy with the console's performance and willing to give developers a tad extra power.
Etch-a-Sketch 3.0 hands-on (video)
The Etch-a-Sketch. A standard bearer for childhood, and one that most of us never really mastered. While Yelizaveta Lokshina can't help you create awe-inspiring portraits from aluminum powder, she has managed to update the toy for the digital age. Using an Arduino, a few buttons and a pressure sensor crammed inside a hollowed-out Etch-a-Sketch, the 3.0 version of the doodler is able to draw in old school gray, as well as vibrant colors created by blending an RGB palette. While holding down the red, green or blue button you squeeze the pressure sensor to add more or less of individual hues. The same sensor is used to change brush width when you hold down the black button. There's even a secret mode that automatically cycles through colors and thicknesses for creating vibrant, almost hallucinatory patterns. At the moment, the dual doodle knobs need to be physically connected to a computer so that a Processing script can work its magic and render the virtual Etch-a-Sketch. But, future versions may include wireless for sketching out images from the comfort of a couch and an accelerometer for the replicating the satisfying sensation of shaking the red fram to erase your creation. Basically, it's still a work in progress. Drawing with the Etch-a-Sketch 3.0 is just as satisfying, in a tactile sense, as the original, though we struggled slightly to get the hang of the pressure sensitive selector. One thing's for sure, though, the kids love it even more than the 1960 creation. Check out the video after the break to see it in action on the floor of the ITP Spring Show.
Daily Mac App: NameChanger
Batch-processing files is a great way to save time. Automator scripts work will for simple jobs, but if you're trying to do something complex like remove or alter the filename of a great many files, a free tool like NameChanger could be a useful addition to your armory. NameChanger takes on file renaming with aplomb. It'll allow you to replace parts of a file name, append or prepend characters to existing file names, as well as add dates, sequences or any other regular expression. It's a little more complex than the likes of Quick File Renamer; but once you get to grips with it, it's a rapid way to process files. One thing that's not that obvious when you first start out, is that to get NameChanger to ignore file extensions in it's processing of your file name, you have to reach into the Options menu and tick the "Hide Extensions" box. That way you can maintain the original file extensions of the files you're processing. Once you've got that down pat, it's just a case dragging and dropping the files you want modified onto the app. You use the drop-down box to select the type of processing you want, type in the characters or select the sequence or date string and hit "Rename." Alternatively, if you want to do it all within the program you can use the "Add," "Remove" and "Clear" buttons on the toolbar. While it's true you can perform all these functions with Automator's file actions, NameChanger brings swift flexibility to your batch-renaming tasks. It's free, so if you're after a file renamer, it's well worth checking out. Thanks for the suggestion Brandon.
Fujitsu's 10.51 petaflop K supercomputer is fastest in the world
Fujitsu, sweaty palmed from the construction of the K supercomputer, swore the colossal machine would manage a speed of ten petaflops by the year 2012. The effort has paid off handsomely: the hangar-sized machine has a top speed of 10.51. If you wanted to duplicate that setup at home, you'd only need 864 racks, 88,128 processors and enough cash in your back pocket to front an annual electricity bill of $10 million a year. The company will be spilling all at November's SC11 conference in Seattle, where the K will cement its position as the fastest computer in the world. One last thing -- we have a hunch that this machine can just about play Crysis.
NVIDIA's Jen-Hsun Huang: Windows on ARM should hit tablets first, battling Intel is a bad idea, would love his chips in iPad
NVIDIA's founder and president Jen-Hsun Huang has never been one to dodge a question, and that made for an excellent closing interview here at AsiaD. Outside of (re)confirming what lies ahead for Tegra, he also spoke quite openly about his feeling towards Windows on ARM in response to a question from Joanna Stern. Here's the bulk of his reply: "It's important for [Microsoft] not to position these as PCs. From a finesse perspective -- I can't speak on their behalf -- but I would come out with tablets first with Windows on ARM. It helps to establish that this isn't a PC. Will yesterday's Office run on tomorrow's Windows on ARM PC? Will a new version of Office run on tomorrow's Windows on ARM tablets? Both questions are about legacy, and both are about Office. The actual implementation of it is radically different. I see no reason to make Office 95 to run on Windows on ARM. I think it would be wonderful, absolutely wonderful -- I'd say, as someone who uses Windows -- it would be almost a requirement to me that [the ARM] device runs Windows interoperably. If Office runs on Windows on ARM -- it's the killer app. Everything else is on the web." He elaborated to say that he would hope Office for Windows on ARM would support the same files that today's Office does, much the same way that Office for Mac eventually synced up with its Windows-based sibling. For more from Huang's interview, hop on past the break!
NVIDIA CEO confirms Tegra roadmap, building all now: Kal-El, Wayne, Logan, Stark
NVIDIA's historically outspoken CEO, Mr. Jen-Hsun Huang, just took the stage here at AsiaD, and among other things, he confirmed to Walt that the Tegra roadmap is well established, and in fact, the entire next-gen range is being produced (internally, of course) right now. That's Kal-El, Wayne, Logan and Stark, all codenamed after superheroes -- Superman, Batman, Wolverine and Ironman, in order of mention. In response to a question of if ASUS' Transformer Prime would be "the first Tegra 3-based product," Huang simply answered "probably." He continued by explaining that it generally takes around three years to build a new generation of Tegra: "We'd like to have a processor every year, and so we're building three in a row." Tegra 3 will end up being the world's first quad-core ARM processor (much like the Tegra 2 was the first dual-core), and he confirmed that NVIDIA has invested some $2 billion in Tegra alone. Finally, he confirmed that the inner workings we've heard about in Project Denver will first be present in the Tegra line with the introduction of Stark -- a long ways out, but at least you've got something (else) to look forward to.
Benchmarks clock iPhone 4S' A5 CPU at 800MHz, show major GPU upgrade over iPhone 4
Pre-orders for the iPhone 4S only began shipping this week, but a handful of early owners have already taken Apple's first A5-based smartphone for a test run, and they've got the benchmarks to prove it. The results, obtained by AnandTech, are hardly what we'd call shocking. In terms of Javascript performance (pictured above), the 4S measures up rather nicely against the Tegra 2-based Honeycomb competition, while out-dueling the iPhone 4 in overall CPU muscle. Geekbench results, meanwhile, clock the 4S at around 800MHz, with a score of 623. That's about 25 percent lower than the A5-based iPad 2, but notably higher than the iPhone 4 (see graphic, after the break). When it comes to GPU performance, GLBenchmark 2.1 tests in 1280 x 720, off-screen render mode place Apple's new handset well above the Galaxy S II, with scores of 122.7 and 67.1, respectively. It still trails the iPad 2, not surprisingly, but the 4S' scores show a major advantage over the iPhone 4, which registered a score of 15.3. For more statistics and graphics, check out the source link below.
Square already processing $2 billion in payments
The iOS-based payment system, Square, is blowing up lately, just like we figured it would way back when. The company already handles up to US$2 billion in payments per year, according to the AP, and is used by 750,000 merchants. That "per year" is a little weird to say, especially considering the company has only been around for a little over a year. In short, Square's experiencing phenomenal growth, and merchants everywhere are already depending on the company to see through billions of dollars in revenue. That's pretty unprecedented, even in this crazy fast-moving world of iOS apps. And that's not all. Square is also lowering the barrier of entry to the program, allowing even new merchants to have their money deposited in the bank as quickly as possible. Previously, merchants had a limit of $1000 that could be processed per day, but those limits are apparently now gone for all merchants, so everyone who sends a payment through Square will get it as soon as the company can send it out. That means we'll see even more money going through the system. I've already used Square a few times with various merchants I've dealt with -- early on, those were the folks at Macworld who were more than ready to depend on an iPhone-based processing system, but since then, I've seen it running everywhere from a local theater to a flea market merchant processing credit card payments. Square's growth is really impressive, and if all the indications are true, it's only beginning.
Arduino-powered lighting system infuses your LED with some Ambilight-like pizazz (video)
Why spend your hard-earned money on one of Philips' Ambilight displays when you can just make your own, using some Arduino-based wizardry and a little bit of elbow grease? Fortunately, Minty Boost creator ladyada is here to help. All you'll need is a strand of digital RGB LED pixels, a five-volt DC power supply (along with a female power adapter), any USB-equipped Arduino micro-controller and, of course, the appropriate Processing programming environment. You can find the full how-to at the source link below, but the results are pretty impressive -- a capture-based sketch system that's compatible with just about any media player. See it for yourself in the video after the break. [Thanks, Phil]
Qualcomm unveils Snapdragon roadmap, 2.5GHz CPUs coming early next year
We've been hearing about Qualcomm's next generation of 2.5GHz processors for a few months now, but the company's quad-core future has now become a little bit clearer. Speaking at the Innovation Qualcomm event in Istanbul yesterday, Senior Vice President of Product Management Cristiano Amon confirmed that the chipmaker's S4 line of silicon will be shipped to manufacturers by the end of this year and should appear in consumer products by the beginning of 2012. Available in single-, dual- or quad-core models, the new, 28nm additions to the Snapdragon family will also support Adreno graphics, 3D and 1080p HD, in addition to 3G and LTE connectivity. If all goes according to schedule, then, we could see a slate of S4-equipped handsets at next year's Mobile World Congress in February, though we'll try to contain our excitement until we get a more specific launch date.
IBM developing largest data drive ever, with 120 petabytes of bliss
So, this is pretty... big. At this very moment, researchers at IBM are building the largest data drive ever -- a 120 petabyte beast comprised of some 200,000 normal HDDs working in concert. To put that into perspective, 120 petabytes is the equivalent of 120 million gigabytes, (or enough space to hold about 24 billion, average-sized MP3's), and significantly more spacious than the 15 petabyte capacity found in the biggest arrays currently in use. To achieve this, IBM aligned individual drives in horizontal drawers, as in most data centers, but made these spaces even wider, in order to accommodate more disks within smaller confines. Engineers also implemented a new data backup mechanism, whereby information from dying disks is slowly reproduced on a replacement drive, allowing the system to continue running without any slowdown. A system called GPFS, meanwhile, spreads stored files over multiple disks, allowing the machine to read or write different parts of a given file at once, while indexing its entire collection at breakneck speeds. The company developed this particular system for an unnamed client looking to conduct complex simulations, but Bruce Hillsberg, IBM's director of storage research, says it may be only a matter of time before all cloud computing systems sport similar architectures. For the moment, however, he admits that his creation is still "on the lunatic fringe."
ITC patent ruling against Apple will stand, Kodak nods approvingly
July is shaping up to be a pretty good month for Kodak. Just a few weeks after granting the camera-maker a second wind in its ongoing patent battle against Apple and RIM, the ITC has issued yet another decision in its favor, determining that a May ruling against Cupertino will stand. At issue is an Apple complaint, filed in April 2010, charging Kodak with infringement of two patents on image processing and power management. On May 12, ITC Judge Robert Rogers shot down Apple's attack, ruling that the patents were not infringed and that one of them was invalid. The full Court had been scheduled to review Rogers' decision later this year, but that won't be happening, now that the ITC has decided to close the investigation (see the PDF, below). Kodak was understandably pleased with the result, though its focus will now turn to August 30th, when an administrative law judge is expected to weigh in on the company's patent offensive against both RIM and Apple.
Fujitsu K supercomputer now ranked fastest in the world, dethrones China's Tianhe-1A
Remember the K -- the Fujitsu supercomputer that promised to do a whopping ten petaflops by the year 2012? Well, it hasn't reached that threshold just yet, but according to the latest Top 500 supercomputer list, it's still faster than any other machine on Earth. In fact, the top-ranked beast is more powerful than the next five supercomputers combined, consumes enough electricity to power about 10,000 homes for a full year, and is capable of churning out about 8.2 quadrillion calculations per second -- three times as many as what runner-up (and former number-one) Tianhe-1A can process. Today's announcement marks the first time since 2004 that a Japanese creation sits atop Top500.org's rankings, but Fujitsu isn't exactly resting on its laurels. Before deploying it next year, engineers at the Riken Advanced Institute for Computational Science are aiming to add about 100,000 cores to the K's collection of 548,352, which would provide it with even more computational muscle, and likely spell doom for all of humanity. Find out more in the PR after the break.
AMD announces new, more energy efficient Embedded G-Series APUs
AMD has just rolled out two new additions to its line of Embedded G-Series APUs, combining Fusion-based processing with reduced power consumption. Both the T40E and T40R rock the same 64-bit x86 Bobcat CPU cores and DirectX 11-capable GPUs we've already seen in previous G-Series incarnations, but AMD says the pair can operate using thirty-nine percent less power than its cousins. The single core T40R boasts a thermal design power (TDP) rating of just 5.5 watts, while the dual core T40E offers a TDP of 6.4 watts. The two applications are designed for compact fanless systems, including kiosks and mobile industrial devices, though Axiomtek is apparently planning on incorporating the new APUs in a new Pico-ITX consumer PC, as well. Full PR after the break.
Our annual data consumption estimated at 9.57 zettabytes or 9,570,000,000,000,000,000,000 bytes
The internet is a mighty big place that's only growing larger each day. That makes it a perfectly unwieldy thing to measure, but the traffic it generates has nonetheless been subjected to a rigorous estimation project by a group of UC San Diego academics. Their findings, published online this month, reveal that in 2008 some 9.57 zettabytes made their way in and out of servers across the globe. Some data bits, such as an email passing through multiple servers, might be counted more than once in their accounting, but the overall result is still considered an under-estimation because it doesn't address privately built servers, such as those Google, Microsoft and others run in their backyards. On a per-worker basis (using a 3.18 billion human workforce number), all this data consumption amounts to 12GB daily or around 3TB per year. So it seems that while we might not have yet reached the bliss of the paperless office, we're guzzling down data as if we were. Check out the report below for fuller details on the study and its methodology.
Kinect hack turns Arduino-controlled delta robot into aggressive claw crane (video)
Candy crane, teddy picker, claw machine, whatever you call it, this arcade mainstay was robbing children of their golden tokens long before we slid into our first pair of Hammer pants, but despite the changing face of the plush offerings within, the crane game's remained mostly the same. Now a team of students at the Bartlett School of Architecture have produced a Kinect hack that could change the way you drop that claw. The rather temperamental delta robot enlists the ever-hackable peripheral in combination with Processing and Arduino to mimic the movements of a user's arm. As you can tell by the video below, the delta hasn't quite figured out the subtleties of human gestures, but the robot's creators say they intend to implement "several autonomous behaviors" once all the kinks are worked out. Frankly, we'd pay our weight in tokens to see the crane game bite back at an unsuspecting whippersnapper. Video after the break.
Apple patent suggests Magic Mouse with display panel
A newly revealed Apple patent has an interesting take on the Magic Mouse: the patent shows a display embedded in its touch surface, creating sort of a mini iPod touch-style interface. It's not clear just how much processing power the display-enabled mouse itself would have, but uses mentioned in the patent seem to involve the display working in conjunction with actual Mac apps, like showing a calculator when using the Numbers app on your Mac, or a magnification of text when using Pages. There is an actual processor listed in the patent, so the Mouse would have some capability of its own to run integrated apps. You may be able to set up parts of the touch display surface to work as buttons, maybe to handle special functions in games or apps. And the patent also says that this display technology could be used in the Magic Trackpad or even in the MacBook trackpad itself, providing an extra space for displaying graphics wherever the user has a touch interface to play with. It's an interesting idea. As with most of these patents, it sounds like this is simply something Apple is researching rather than actual technology going into an approaching Magic Mouse refresh -- there are a lot of elements that have to be worked out (like why would you put a display on a device usually covered by your hand?) before this kind of thing goes on sale. But it is an interesting idea, and we may one day see Apple use it in a real product.
