Sign up for our Technology Newsletter 
Will Facebook Go Open-Source
December 29, 2015 by admin
Filed under Around The Net
Comments Off on Will Facebook Go Open-Source
Facebook has unveiled its next-generation GPU-based systems for training neural networks, Open Rack-compatible hardware code-named “Big Sur” which it plans to open source.
The social media giant’s latest machine learning system has been designed for artificial intelligence (AI) computing at a large scale, and in most part has been crafted with Nvidia hardware.
Big Sur comprises eight high-performance GPUs of up to 300 watts each, with the flexibility to configure between multiple PCI-e topologies. It makes use of Nvidia’s Tesla Accelerated Computing Platform, and as a result is twice as fast as Facebook’s previous generation rack.
“This means we can train twice as fast and explore networks twice as large,” said the firm in its engineering blog. “And distributing training across eight GPUs allows us to scale the size and speed of our networks by another factor of two.”
Facebook claims that as well as better performance, Big Sur is also far more versatile and efficient than the off-the-shelf solutions in its previous generation.
“While many high-performance computing systems require special cooling and other unique infrastructure to operate, we have optimised these new servers for thermal and power efficiency, allowing us to operate them even in our own free-air cooled, Open Compute standard data centres,” explained the company.
We spoke to Nvidia’s senior product manager for GPU Computing, Will Ramey, ahead of the launch, who has been working on the Big Sur project alongside Facebook for some time.
“The project is the first time that a complete computing system that is designed for machine learning and AI will be released as an open source solution,” said Ramey. “By taking the purpose-built design spec that Facebook has designed for their own machine learning apps and open sourcing them, people will benefit from and contribute to the project so it can move the entire industry forward.”
While Big Sur was built with Nvidia’s new Tesla M40 hyperscale accelerator in mind, it can actually support a wide range of PCI-e cards in what Facebook believes could make for better efficiencies in production and manufacturing to get more computational power for every penny that it invests.
“Servers can also require maintenance and hefty operational resources, so, like the other hardware in our data centres, Big Sur was designed around operational efficiency and serviceability,” Facebook said. “We’ve removed the components that don’t get used very much, and components that fail relatively frequently – such as hard drives and DIMMs – can now be removed and replaced in a few seconds.”
Perhaps the most interesting aspect of the Big Sur announcement is Facebook’s plans to open-source it and submit the design materials to the Open Compute Project. This is a bid to make it easier for AI researchers to share techniques and technologies.
“As with all hardware systems that are released into the open, it’s our hope that others will be able to work with us to improve it,” Facebook said, adding that it believes open collaboration will help foster innovation for future designs, and put us closer to building complex AI systems that will probably take over the world and kill us all.
Nvidia released its end-to-end hyperscale data centre platform last month claiming that it will let web services companies accelerate their machine learning workloads and power advanced artificial intelligence applications.
Consisting of two accelerators, Nvidia’s latest hyperscale line aims to let researchers design new deep neural networks more quickly for the increasing number of applications they want to power with AI. It also is designed to deploy these networks across the data centre. The line also includes a suite of GPU-accelerated libraries.
Courtesy-TheInq
Is Canon Betting Its Future On IoT?
Canon has announced that it is joining the raft of technology companies attempting to take on the Internet of Things (IoT) through what it is calling the ‘Imaging of Things’.
Speaking at the firm’s EXPO 2015 event in Paris on Tuesday, Canon CEO Fujio Mitarai talked up the firm’s global vision for the future as the IoT becomes more pervasive.
“Canon is showing how the world of imaging is expanding rapidly in the age of the IoT,” said Mitarai.
“In the future nearly everything will be connected through smart devices. These rely on built-in cameras or sensors and the data they generate. As a result, Canon predicts that the IoT will largely depend on the ‘Imaging of Things’.”
To take on this future, Mitarai plans to overhaul Canon’s business structure to build a network of smaller Canon companies and thus create an “ecosystem of innovation”.
The CEO said that these companies have been designed to “harness innovation and creative talents from across the regions”, and will include more investment in what Canon does but on a more local level in different regions across the world, as opposed to all of the innovation being created in Tokyo, as it is at the moment.
This will allow “regional independence and international collaboration [to be] put into practice”, Mitarai said.
In this new “network of companies”, Mitarai explained that each regional headquarters will manage local R&D and manufacturing, as well as service and support customised to its market.
In Europe, the smaller Canon companies will focus on printing and network video surveillance, and the firm has already brought in specialists in these business areas such as Océ, Axis and Milestone Systems.
Mitarai said that, along with its global reputation for cameras, this will make Canon the largest printing and network video surveillance company in the world.
On a B2B level, the move is also about helping other firms build new competitive advantages and improve services for their own customers.
“We are changing our own operation model and go to market structure to build more expertise in these areas and connect with our customers,” said Jeppe Frandsen, head of the Production Printing Group at Canon Europe.
“Our customers are changing so we are now looking at a way customers are changing to what their customers want – new ways to do business together.”
Canon’s EXPO 2015 event was also an opportunity for the company to show off many of the latest projects from its R&D centre in Tokyo for the first time in Europe.
These tie in with the firm’s new focus as it launches smaller companies in more regional areas, and include a range of innovative practices such as responding to society’s monitoring needs, 3D printing as part of a partnership with 3D Systems in Europe, and graphic arts via investment in digital print technologies.
Source-http://www.thegurureview.net/technology-2/is-canon-betting-its-future-on-iot.html
Is Russia Behind Recent US Malware Attacks?
Comments Off on Is Russia Behind Recent US Malware Attacks?
It would appear that while the US has been blaming China for all its cyber break-ins it appears to be ignoring Tsar Putin’s elite hacking team for the last seven years.
For the past seven years, a cyberespionage group operating out of Russia on the orders of Tsar Putin have been conducting a series of malware campaigns targeting governments, political think tanks and other organizations.
Researchers at F-Secure have been looking into the antics of an outfit called “the Dukes” which has been active since at least 2008. The group has evolved into a methodical developer of “zero-day” attacks, pulling together their own research with the published work of other security firms to provide a more detailed picture of the people behind a long-running family of malware.
The Dukes specialize in “smash and grab” attacks on networks, but have also used subtle, long-term intrusions that harvested massive amounts of data from their targets.
The group’s targets do include criminal organisations operating in the Russian Federation, which suggest there is some form of policing element to it. But they are mostly interested in Western governments and related organisations, such as government ministries and agencies, political think tanks and governmental subcontractors.
F-Secure team wrote. “Their targets have also included the governments of members of the Commonwealth of Independent States; Asian, African, and Middle Eastern governments; organisations associated with Chechen terrorism; and Russian speakers engaged in the illicit trade of controlled substances and drugs.”
The group was named after its earliest-detected malware, known as PinchDuke. Its targets were associated with the Chechen separatist movement. Later that year they were going after Western governments and organisations in search of information about the diplomatic activities of the United States and the NATO.
Most of the attacks used spear phishing emails as the means of injecting malware onto targeted systems, one of their attacks have spread malware through a malicious Tor exit node in Russia, targeting users of the anonymising network with malware injections into their downloads.
The targets have always followed Russian government interests. There are a number of Russian-language artifacts in some of the malware, including an error message in PinchDuke. GeminiDuke also used timestamps that were adjusted to match Moscow Standard time.
Before the beginning of the Ukraine crisis, the group began using a number of decoy documents in spear phishing attacks that were related to Ukraine. They included a letter undersigned by the First Deputy Minister for Foreign Affairs of Ukraine.
However, after the crisis happened the attacks dropped off suggesting that it was an intelligence gathering operation. It is also a big operation, which, if operating in Russia would most likely require state acknowledgement, if not outright support.
Source-http://www.thegurureview.net/computing-category/is-russia-behind-us-malware-attacks.html
Drones To Have Intel Inside
Intel is taking its competitive game up a notch by investing in its own drones.
Intel has written a check for more than US$60 million to Yuneec International, a Chinese aviation company and drone maker.
This is not the first time that the Chipmaker has invested in drones. It has written smaller amounts for the drone makers Airware and PrecisionHawk. The Yuneec deal is its largest investment in a drone company yet.
Apparently Intel thinks that drones are potential computing platforms for its processors.
Intel CEO Brian Krzanich said he believed in a smart and connected world. And one of the best ways to bring that smart and connected world to everyone and everywhere has been drones.
Amazon and Google are developing drones as they seek new ways to deliver items to consumers, Intel just wants to make sure that its chips are delivering the payload. There is no indication that it is building a secret airforce which it will use to take down competition – that would be silly.
Yuneec makes a range of drones built for aerial photography and imaging. Its technology also powers manned electric aircraft.
Source-http://www.thegurureview.net/computing-category/drones-to-have-intel-inside.html
Will Qualcomm Unveil The Snapdragon 820 SoC
Comments Off on Will Qualcomm Unveil The Snapdragon 820 SoC
Qualcomm is set to unveil its new Snapdragon 820 SoC on August 11 in LA and more details are being leaked than you would see at a Welsh leak recipe contest.
It appears that the new Snapdragon 820 will have the catchy title MSM8996 and it promises some significant performance improvements in key areas. We already know that it will not be catching fire, but it also has a 40 per cent GPU performance increase with its A530 GPU.
The device is also claimed to have a 30 per cent power improvement with 64b of shared virtual memory with the CPU.
Another big area of improvement is the Hydra CPU, which claims a 35 percent improvement compared to the Snapdragon 810.
The Snapdragon 820 will support 4k60 entertainment and high-speed data connectivity.
There are rumours that there will be a QFE3100 Envelope Tracking system this will not speed up mail in the criminally slow Italian Post Office, but should create a lower power and a thermal footprint. A dedicated low power sensor is integrated for always on use.
Another major upgrade compared to the older SoC is a switch from 20nm to 14nm FinFET manufacturing process. We are still expecting the Xiaomi Mi5 to be the first one to use it.
Qualcomm Gives Snapdragon More Umph
Qualcomm has released a new Trepn Profiler app for Android which will profile Snapdragon processors and tinker with them.
The Trepn Profiler app identifies apps that overwork the CPU or are eating too much data. The app will pinpoint which of the apps drain the battery faster.
All data that will be obtained by this app can provide information you need to know which program is slowing down your phone.
Most Android phone users will not give a damn, but developers will find it useful. Those who are interested in testing roms, custom kernels, and their own apps can use the data gathered by the Trepn Profiler.
Developers can measure optimisation and performance on Snapdragon-powered mobile devices. Data are real-time include network usage, battery power, GPU frequency load, and CPU cores’ load. Key features also include six fast-loading profiling presets, and an advanced mode to manually select data points and save for analysis.
The Advanced Mode allows profiling a single app or device, offline data analysis, and increasing of data collection interval. This special mode also allows longer profiling sessions, displaying two data point in one overlay, and viewing of profile data.
All up this should enable developers to come up with more Snapdragon friendly apps.
Qualcomm Goes Ultrasonic
Qualcomm has unveiled what it claims is the world’s first ‘ultrasonic’ fingerprint scanner, in a bid to improve mobile security and further boost Android’s chances in the enterprise space.
The Qualcomm Snapdragon Sense ID 3D Fingerprint technology debuted during the chipmaker’s Mobile World Congress (MWC) press conference on Monday.
The firm claimed that the new feature will outperform the fingerprint scanners found on smartphones such as the iPhone 6 and Galaxy S6.
Qualcomm also claimed that, as well as “better protecting user data”, the 3D ultrasonic imaging technology is much more accurate than capacitive solutions currently available, and is not hindered by greasy or sweaty fingers.
Sense ID offers a more “innovative and elegant” design for manufacturers, the firm said, owing to its ability to scan fingerprints through any material, be it glass, metal or sapphire.
This means, in theory, that future fingerprint sensors could be included directly into a smartphone’s display.
Derek Aberle, Qualcomm president, said: “This is another industry first for Qualcomm and has the potential to revolutionise mobile security.
“It’s also another step towards the end of the password, and could mean that you’ll never have to type in a password on your smartphone again.”
No specific details or partners have yet been announced, but Qualcomm said that the Sense ID technology will arrive in devices in the second half of 2015, when the firm’s next-generation Snapdragon 820 processor is also tipped to debut.
The firm didn’t reveal many details about this chip, except that it will feature Kryo 64-bit CPU tech and a new machine learning feature dubbed Zeroth.
Qualcomm also revealed more details about LTE-U during Monday’s press conference, confirming plans to extend LTE to unused spectrum using technology integrated in its latest small-cell solutions and RF transceivers for mobile devices.
“We face many challenges as demand for data constantly grows, and we think the best way to fix this is by taking advantage of unused spectrum,” said Aberle.
Finally, the chipmaker released details about a new a partnership with Cyanogen, the open-source outfit responsible for the CyanogenMod operating system.
Qualcomm said that it will provide support for the best features and UI enhancements of CyanogenMod on Snapdragon processors, which will be available for the release of Qualcomm Reference Design in April.
The MWC announcements follow the launch of the ARM Cortex-based Snapdragon 620 and 618 chips last month, which promise to improve connectivity and user experience on high-end smartphones and tablets.
Aberle said that these chips will begin to show up in devices in mid to late 2015.
Sony Exits PC Business
Sony will unload its struggling PC business to a Japanese investment firm, the company said Thursday, raising the possibility that the “Vaio” brand could all but disappear from markets outside Japan.
Tokyo-based investment fund Japan Industrial Partners (JIP) will operate the Vaio PC brand under a newly established firm and initially sell PCs in Japan only.
In another reform aimed at bolstering its restructuring efforts, Sony also said it would turn its beleaguered TV business into a subsidiary.
The moves come as Sony said it now expects a net loss of $1.1 billion for the year to the end of March, a reversal of its October profit forecast.
Vaio, which Sony introduced in 1996, looks set to vanish from most markets, at least for short term, as the new company will initially concentrate on selling consumer and corporate PCs in Japan. Whether or not Sony will continue to produce products under the Vaio brand remains to be seen, Sony said.
Although Sony is selling its PC business, it will continue to produce tablet computers, part of its renewed focus on mobile devices including smartphones.
Sony did not put a price on the sale. Sony will take a 5% stake in the new firm, it said.
Sony will stop making and selling PCs after its 2014 Spring lineup launch, but about 250 to 300 Sony staff, including some from a subsidiary that produces TV sets, cameras and computers at factories in Japan, will be hired by the new company, which is to be based at the hub of Sony’s current PC business in Japan’s Nagano Prefecture.
Meanwhile, Sony said it will turn its TV business, which has faced a decade of losses, into a wholly owned subsidiary by July 2014.
Sharp Shows MEMs
October 9, 2013 by admin
Filed under Around The Net
Comments Off on Sharp Shows MEMs
Sharp on Monday unveiled its latest prototypes of a new kind of display screen that it says brings several advantages over today’s liquid crystal display (LCD) screens.
The screens, called microelectromechanical systems (MEMS) displays for the tiny moving parts they contain, are being developed by the Japanese company in partnership with Qualcomm and were on show at the Ceatec electronics show just outside of Tokyo.
Behind each pixel in a MEMS display is a backlight that flashes red, green and blue in fast succession, and in front of it is a tiny shutter can be opened to let light through.
Synchronized to the backlight, the shutter can control the amount of each color of light allowed through. The eye perceives these flashes as the desired hue.
In contrast, today’s LCD screens create colored pixels using three filters. The filters swallow about two thirds of the brightness of the backlight before it leaves the display, said Akira Imai, deputy general manager of Sharp’s new business development center.
The MEMS display can allow all the light through, so the intensity of the backlight can be reduced using less power for the display, said Imai.
In a portable gadget, the screen often consumes more power than any other component, so reducing its demands can have a big impact on battery life.
The screens on show at Ceatec were 7-inch models with 800 pixel by 1,280 pixel resolution. The colors were bright and the screen image was sharp, although people viewing the screens did tend to see a brief flash of red, green and blue pixel each time they turned their eyes away from the display. That’s something Sharp is working on, said Imai.
Sharp also showed a version of the screen working in several low power modes.
The development work with Qualcomm began earlier this year when the U.S. company said it would invest $120 million into Sharp. The money, which was invested in two parts, was accompanied by Qualcomm’s MEMS expertise. Sharp has a long history in flat-panel display technology, especially LCD, and has recently been working on a new type of display called IGZO, on which the MEMS display is partly based.
Toshiba To Offer A 20-megapixel Image Chip
Comments Off on Toshiba To Offer A 20-megapixel Image Chip
Toshiba is gearing up for to offer a 20-megapixel image sensor for digital cameras that it says will be the highest resolution of its kind.
The Tokyo-based firm said the new chips will be able to support capturing 30 frames per second at full resolution. They will also be able to shoot video at 60 frames per second at 1080P or 100 frames at 720P.
Toshiba said it will begin shipping samples of the new CMOS chips in January, with mass production to begin in August of 300,000 units monthly. Toshiba is best known in components for its NAND flash memory, which it develops with partner SanDisk, but is also a major manufacturer of LSI and other semiconductors.
Digital point-and-shoot cameras are steadily falling in price, squeezed between brutal competition among manufacturers and the increasing threat of smartphones and mobile devices. While the number of pixels a camera can capture is not always a direct measure of the overall quality of its images, it is a key selling point to consumers.
The image resolution of top-end smartphones now often meets or exceed that of digital cameras. The Nokia 808 PureView launched earlier this year has a 41-megapixel image sensor.
The Japanese manufacturer said it has increased the amount of information pixels in the new chip can store compared to its previous generation of CMOS, producing better overall images. It has also reduces the size of pixels – the new 20-megapixel version has individual pixels that measure 1.2 micrometers, down from 1.34 micrometers in its 16-megapixel product.