Subscribe to:

Subscribe to :: TheGuruReview.net ::

Linux Appears To Be The King In The Supercomputing Space

November 21, 2017 by  
Filed under Computing

Looking at the November 2017 TOP500 Supercomputer list one thing is particularly clear – the open saucy Linux is king.

In 1998, Linux first appeared on the TOP500 Supercomputer list and it was regarded as unusual – indeed many just said it was because it was really Unix in drag. But the November list showed that all 500 of the world’s fastest supercomputers are running Linux.

There had only been two non-Linux systems left on the list but the pair of Chinese IBM POWER computers running AIX were too slow to rate a mention any more.

Before Linux took the lead, Unix was everywhere but slowly, since 2003, the Linux was the TOP500 main OS of choice. By 2004, Linux had taken the lead for good.

The reason Linux did well in this arena and not the desktop is that most of the world’s top supercomputers are research machines built for specialised tasks, each machine is a standalone project with unique characteristics and optimisation requirements.

Linux means that research teams can easily modify and optimize open-source code to their one-off designs.

Courtesy-Fud

OnePlus Phones Have Dangerous Hacking Backdoor

November 17, 2017 by  
Filed under Mobile

Hackers who obtained OnePlus phones can obtain virtually unlimited access to files and software through use of a testing tool called EngineerMode that the company evidently left on the devices.

Robert Baptiste, a freelance security researcher who goes by the name Elliot Alderson on Twitter after the “Mr. Robot” TV show character, found the tool on a OnePlus phone and tweeted his findings Monday. Researchers at security firm SecureNow helped figure out the tool’s password, a step that means hackers can get unrestricted privileges on the phone as long as they have the device in their possession.

The EngineeerMode software functions as a backdoor, granting access to someone other than an authorized user. Escalating those privileges to full do-anything “root” access required a few lines of code, Baptiste said.

“It’s quite severe,” Baptiste said via a Twitter direct message.

OnePlus disagreed, though it said it’s decided to modify EngineerTool.

“EngineerMode is a diagnostic tool mainly used for factory production line functionality testing and after sales support,” the company said in a statement. Root access “is only accessible if USB debugging, which is off by default, is turned on, and any sort of root access would still require physical access to your device. While we don’t see this as a major security issue, we understand that users may still have concerns and therefore we will remove the adb [Android Debug Bridge command-line tool] root function from EngineerMode in an upcoming OTA.”

SecureNow found the tool on the OnePlus 3 and OnePlus 5. Android Police reported it’s also on the OnePlus 3T. And Baptiste said it’s also on the new OnePlus 5T.

Baptiste had spotted evidence that EngineerMode was written by mobile chipmaker Qualcomm. But Qualcomm said Wednesday that’s not the case.

“After an in-depth investigation, we have determined that the EngineerMode app in question was not authored by Qualcomm,” the company said in a statement. “Although remnants of some Qualcomm source code is evident, we believe that others built upon a past, similarly named Qualcomm testing app that was limited to displaying device information. EngineerMode no longer resembles the original code we provided.”

Facebook Has A Unique Way To Fight Revenge Porn

November 10, 2017 by  
Filed under Around The Net

Facebook is prompting people to share their nude photos. But this isn’t what it sounds like.

The goal of the social network’s plan is to make sure people’s nude photos aren’t used for revenge porn by a disgruntled ex-boyfriend or girlfriend, according to the Australian Broadcasting Corp.

The way it’ll work is people will share their photos with Facebook via its Messenger app and the company will then “hash” the images, which is a process that converts the photos into a unique digital code. Once Facebook has that code, it can block the images from ever being uploaded to its site. The company will store the images for a short time and then delete them.

The company is piloting the technology in Australia with a small government agency headed by e-Safety Commissioner Julie Inman Grant.

We see many scenarios where maybe photos or videos were taken consensually at one point, but there was not any sort of consent to send the images or videos more broadly,” Inman Grant told the ABC.

Other tech companies have used similar types of hashing technology in efforts to rid the internet of child pornography. Google, Microsoft, and Twitter have used unique digital codes to detect exploitative images, some of which have led to the arrests of people distributing the photographs on the web.

Facebook didn’t immediately respond to request for comment.

Is The Linux Desktop Gaining In Popularity

November 9, 2017 by  
Filed under Computing

Linux fanboys were dancing in the streets when NetMarketShare’s desktop operating system analysis, which showed Linux leaping from 2.5 percent in July, to almost five percent in September.

The news was seen as living proof that 2017 was the year of the Linux desktop and that things would only get better for next year.

Sadly, though, it was all a terrible mistake Vince Vizzaccaro, NetMarketShare’s executive marketing share of marketing has said that the reported Linux share was incorrect.

“We are aware of the issue and are currently looking into it”, he said.

According to the US federal government’s Digital Analytics Program (DAP) that shows desktop Linux, as usual, hanging out in “other” at 1.5 percent.

Windows, as always, is on top with 45.9 percent, followed by Apple iOS, at 25.5 percent, Android at 18.6 percent, and macOS at 8.5 percent.

The figures are US-based so they over emphasis Apple. Linux tends to do better in poorer parts of the world where people must justify spending money on more important functions such as power, stability rather than just the logo on the case.

The article does, however, acknowledge that Linux’s real market share is probably a little higher this month.

Courtesy-Fud

Singapore Entice Seniors To Code With New Program

October 31, 2017 by  
Filed under Uncategorized

While some elderly people around the world are still busy figuring out how to use the internet, Singapore’s senior citizens are getting started on coding.

Over the weekend, the libraries in Singapore played host to a special version of the global Hour of Code movement. Seniors aged 50 years and older learned how to code in Swift alongside student volunteer partners. The event was part of an initiative driven by the Singapore government to get its senior citizens up to speed with technology.

The Hour of Code is designed to help students to pick up coding and computer science. It can be organized by anyone, including those who don’t know how to code. The movement claims to have already reached out to 100 million students in 180 countries.

Using Apple’s iPads and the Swift Playgrounds app, older participants spent an hour so learning the basic of programing, first getting the avatar moving around and collecting gems before moving on to loop integrations, though this depends on how quickly they pick up the language.

 For former school principal Foo Chee Meng, 71, the session offered him the opportunity to revisit coding. He learned Fortran and COBOL decades ago when he was still at studying at a university.

“Swift is very different from what I learned previously, they’ve made it so that even ordinary people can just pick it up quite simply, not so for Fortran COBOL, you’ll really need some knowledge for those,” said Foo.

While the participants are unlikely to end up with killer coding skills or walk in with experience with coding like Foo, the one-hour session could help spark interest in a different hobby to keep their minds active.

For 15-year-old volunteer Juraais Bin Hasbullah, his main concern was being able to communicate with older participants who don’t speak English as he doesn’t speak the Chinese dialects. In Singapore, English adoption among the elderly isn’t as widespread as the younger generation, though they can speak a smattering of it.

“It’s not going to be easy for me as I’m afraid of the language barrier. But in terms of teaching them on how it’s being done, I think it’s easy as it’s a very intuitive app,” said Juraais. “And if there are more sessions, I’ll definitely volunteer again.”

Google’s Brass Fixates On Cheeseburger Emoji

October 31, 2017 by  
Filed under Around The Net

Google’s technical achievements are well known : Google Search, Google Maps, the Pixel 2 XL. And yet, it doesn’t seem to know the simple formula for assembling a cheeseburger.

The company’s problematic approach to making up everyone’s favorite meaty sandwich was pointed out on Twitter by writer Thomas Baekdal, who noted that in Google’s version of the cheeseburger emoji the cheese sits underneath the patty. Quelle horreur!

Fortunately, Google immediately put its top man on the job. Company CEO Sundar Pichai, retweeted the image posted by Baekdal on Sunday, promising: “will drop everything else we are doing and address on Monday.”

Great, you might think, that’s that taken care of. But not quite. Pichai will make the fix on one condition: “if folks can agree on the correct way to do this.” This suggests that he himself is not quite sure where the cheese should go in the order and seems to think it’s open to debate, which it obviously is not.

Let us help you out here, Sundar. The cheese goes on top of the meat. On the top.

Microsoft’s Edge Browsers Appears To Be The Best At Thwarting Malware

October 19, 2017 by  
Filed under Around The Net

Microsoft’s Edge easily beat rival browsers from Google and Mozilla in third party tests of the behind the scenes services which power anti-malware warnings and malicious website-blocking.

NSS Labs said Windows 10’s default browser is better at blocking phishing and socially-engineered malware attacks than Google Chrome or Mozilla Firefox.

The outfit said Edge automatically blocked 92 percent of all in-browser credential phishing attempts and stymied all socially-engineered malware (SEM) attacks.

The latter encompassed a wide range of attacks, but their common characteristic was that they tried to trick users into downloading malicious code.

The tactics that SEM attackers deploy include links from social media, such as Facebook and Twitter, and bogus in-browser notifications of computer infections or other problems.

Edge decisively bested Chrome and Firefox by decisive margins. Chrome blocked 74 percent of all phishing attacks, and 88 percent of SEM attacks.

Meanwhile, Firefox came in third in both tests, stopping just 61 percent of the phishing attacks and 70 percent of all SEM attempts.

Chrome and Mozilla’s Firefox rely on the Safe Browsing API, but historically Mozilla’s implementation has performed poorly compared to Google’s.

Edge also took top prize in blocking attacks from the get-go. In NSS’s SEM attack testing the Voleware stopped every attempt from the first moments a new attack was detected. Chrome halted 75 percent and Firefox halted 54 percent of the brand new attacks

The researchers spent three weeks continuously monitoring the browsers on Windows 10 computers.

Courtesy-Fud

FireFox Quantum Browser Coming In November

October 6, 2017 by  
Filed under Around The Net

After being stuck in the slow lane for ages, Mozilla’s new Quantum browser is starting to look like it might be faster than Chrome.

A beta version of Firefox Quatum lets you test whether Mozilla’s newly named web browser, replete with changes built over more than a year, is a match for Google. We had a quick look and it managed to make Fudzilla’s esoteric CMS machine go like the clappers. Opera on the other hand keeps on insisting that it needs a password for every screen.

Mozilla CEO Chris Beard claims that the new browser is a “big bang” although we suggest that probably means he needs to get out more. Company executives have acknowledged they let Firefox languish but now it is ready to do better with its life.

Firefox 57 is faster at starting up and loading web pages, judged on page-load speed, “Firefox Quantum is often perceivably faster” while using 30 percent less memory, Nguyen said in a blog post Tuesday. And it’s twice as fast as Firefox a year ago.

The new Firefox revamp includes Quantum Flow, which stamps out dozens of performance bugs, and Quantum CSS, aka Stylo, which speeds up website formatting. Photon that kills Firefox’s rounded tabs and adds a “page action” menu into the address bar. It also builds in the Pocket bookmarking service Mozilla acquired and uses it to recommend sites.

All up, it does not appear too bad. The phrase “at bloody last” crosses my mind. It still needs its acid test – whether or not it can handle Mrs Farrell’s shopping, which for some reason requires 105 open tabs which must never be closed unless you want to be divorced.

Firefox Quantum will arrive in its final form on November the 14th.

Courtesy-Fud

Is The Market For A.I. Chips Getting Hot

September 13, 2017 by  
Filed under Computing

It is starting to look like the AI chipmarket is going to be a battle ground for chip development.

The AI market will be worth an estimated $59.8 billion by 2025 — up from just $1.4 billion last year. Nearly all major players in the IT industry have joined the race to develop AI chips and applications with the aim of establishing an early presence.

Currently mainstream products in the AI chip markets are those applied to machine learning and deep neural networks, including ASICs, GPU, FPGA and CPU chipsets. These have involved Nvidia, Intel, Qualcomm, Google, Apple, Microsoft, Amazon, Facebook, and IBM, Samsung, Huawei, Baidu and Tencent.

Nvidia is doing rather well with its GPU chipset series and has its foot on the development of a larger slice of development architectures. This includes Tesla Accelerator applied to cloud and supercomputers, Jetson fitted in robots and drones, and Drive PX adopted in automobiles. All of them share the same architecture to enable algorithm acceleration for deep learning.

Nvidia CEO Jensen Huang said that his company has kept improving the design, system architecture, compiler and algorithm of its GPU solutions, and the deep neural network velocity has been already boosted by 50 fold within three years as a result, much faster than the time frame set by the Moore’s Law.

Intel is stepping up its AI blueprints ranging from network edges to datacenters, with AI chip platforms including Xeon, Xeon Phi processors and EPGA accelerators supporting the optimization of specific data load. The company has completed tests of its Lake Crest AI chipset designed to perfect neural network operation to improve deep learning efficiency.

Intel has its Myriad X, which can be applied to drones, smart cameras or AR (augmented reality) devices to sense and understand fast-changing external environments and facilitate interactions and learning.

The other force is Qualcomm with its newly released its Neural Processing Engine (NPE), with its software development kit (SDK) able to help developers optimise the AI performance on the firm’s Snapdragon 600, 800 series processors and support such AI architectures as Tensoflow, Caffe and Caffe 2.

NPE can manage image recognition, scene detection, camera filtration, and photo retouching. Facebook now uses it to accelerate the execution of the AR function of photos and real-time films through smartphone apps.

Google has its Tensor Process Unit (TPU) chipset for deep learning and algorithm, targeting AI developers and open for commercial and academic users through cloud services. Huawei has recently unveiled its Kirin 970 SoC, which is claimed to be the world’s first AI mobile chip fitted with neural processing unit (NPU).

The firm’s new-generation flagship smartphone models Mate 10 and Mate 10 Pro, set to hit the market in October, will carry Kirin 970 chipset, marking Huawei’s official entry into the AI arena. Baidu and Tencent are also actively developing customised AI chips.

AMD meanwhile has Vega to run cutting-edge artificial intelligence and machine learning tasks—the kinds that fuel Siri and Alexa and are used by corporate giants like GE to analyze “big data” streams.

While investors are certainly optimistic about AMD, there’s little proof that the company can overtake NVIDIA in the AI at the moment. NVIDIA is already a self-driving car tech leader, is the go-to GPU maker for many AI servers, and has a clear understanding of its market potential in the AI space.

Courtesy-Fud

Young Star Helps Astronomers Solve Stellar Mystery

June 27, 2017 by  
Filed under Around The Net

Astronomers using the powerful Atacama Large Millimeter/submillimeter Array (ALMA) in Chile have precisely measured the rotating fountains of gas flowing out from a massive newborn star, revealing the complex interplay between the star’s magnetism and centrifugal forces.

Astronomers are still puzzled by the way massive stars form in interstellar space, the new study’s researchers said in a statement. When a massive rotating cloud of gas collapses under gravity, stellar fusion becomes possible, and a baby star is born. As angular momentum is conserved while the cloud shrinks, the resulting baby star should be spinning very fast, according to the laws of physics. 

To get a better idea of the conservation of angular (or rotational) momentum, imagine a spinning ice-skater. As ice-skaters spin with their arms outstretched, they spin slowly; when they bring their arms close to their bodies, they spin faster. Physics dictates that this concept should hold true for a shrinking cloud of star-forming gas: As it shrinks, it should spin faster.

But astronomers have found that stars in our galaxy spin much more slowly than the laws of physics predict they should. Therefore, there must be some mechanism that’s dissipating angular momentum from stars soon after they are born, the researchers said.

In the new work, published online June 12 in the journal Nature Astronomy, astronomers observed a massive newborn star called Orion KL Source I in the Orion Nebula and used ALMA to reveal the rotation of its powerful stellar winds. 

“We have clearly imaged the rotation of the outflow,” Tomoya Hirota, an assistant professor at the National Astronomical Observatory of Japan (NAOJ) and SOKENDAI (the Graduate University for Advanced Studies) and lead author on the paper, said in the statement. “In addition, the result gives us important insight into the launching mechanism of the outflow.”

Hirota’s team noticed that the outflow of stellar gases is rotating in the same direction as the star and that it emanates from Source I’s hot gas disk, and not from the star itself. This finding agrees with a theoretical “magnetocentrifugal disk wind model,” the researchers said.

In this model, gas is ejected from the rotating disk and is forced to move outward. Like a spinning lawn sprinkler, propelled by centrifugal forces, the water spirals outward, away from the sprinkler head, siphoning some of the star’s angular momentum. But in the case of this star, the spinning gases leaving the disk are also directed up and down along magnetic-field lines to create the spinning outflows that ALMA has detected. And the researchers believe that these flows are dissipating rotational energy from the baby star, slowing down its rotation, and therefore possibly explaining why stars in our galaxy rotate more slowly than expected.

“In addition to high sensitivity and fidelity, high resolution submillimeter-wave observation is essential to our study, which ALMA made possible for the first time,” Hirota said. “Submillimeter waves are a unique diagnostic tool for the dense innermost region of the outflow, and at that exact place, we detected the rotation.

“ALMA’s resolution will become even higher in the future,” Hirota added. “We would like to observe other objects, to improve our understanding of the launching mechanism of outflows and the formation scenario of massive stars with the assistance of theoretical research.”

Courtesy-Space

Mozilla Rolls Out Improved Version Of Firefox

June 19, 2017 by  
Filed under Around The Net

Mozilla continued its years-long campaign to make Firefox more technologically competitive with the competition, Google’s Chrome in particular, by boosting performance, increasing stability and reining in memory consumption.

The open-source developer also patched 31 security vulnerabilities, three of them rated “Critical,” the firm’s most serious ranking.

Firefox 54, released June 13, expanded on Mozilla’s multi-process project, code-named “Electrolysis” (shortened to “e10s”), that since 2009 has tried to mimic Chrome, and separate the browser’s operation into more than one CPU process. Previously, Firefox split its user interface (UI) and all content into separate processes — running all tabs in one of those processes — to prevent the browser from completely crashing when a website or web app failed. Firefox 54 uses up to four processes to run the browser’s tabs, assigning each to one of the CPU buckets.

“By separating the tabs into separate processes, we make better use of the hardware on your computer, so Firefox can deliver you more of the web you love, with less waiting,” assured Nick Nguyen, the product lead for Firefox, in a post to a company blog. In the same piece, Nguyen bragged that version 54 was “the best release of Firefox ever.”

Because operating multiple processes bloats a browser’s memory consumption, and also because Firefox pre-e10s was extensively criticized as a RAM pig, Nguyen asserted that version 54 uses “significantly less RAM” than rivals such as Chrome, Apple’s Safari and Microsoft’s Edge. Elsewhere, Ryan Pollack, a product marketing manager at Mozilla, argued that the four-process limit is the correct compromise between low and high memory use. He even cited the Goldilocks and the Three Bears fairy tale to declare that the balance between performance (lots of processes) and memory consumption (few processes) is perfect.

“Firefox uses four content processes because it’s the ‘just right’ number for many Firefox users,” said Pollack. “With four content processes, your computer should have plenty of memory left to run apps besides Firefox.”

Chrome has relied on a multi-process model since its 2008 launch. Because it devotes a separate process to each tab, and each process requires memory, Chrome generally consumes much more memory than other browsers. (Safari uses a similar, but not identical, multi-process model that ultimately eats less RAM than Chrome. Edge, too, uses multiple processes.) So, it wasn’t surprising that Pollack compared Firefox 54’s memory appetite primarily to Chrome’s, and charged that in a 30-tab test the latter required up to 2.4 times the RAM of Firefox.

Users with devices boasting larger amounts of RAM — more than 8GB, Pollack said — can boost the number of processes Firefox 54 consumes by typing about:config in the browser’s address bar, then changing the number for the dom.ipc.processCount setting.

While e10s has been a focus of Mozilla engineers for two years, the project also illustrated how far Firefox had fallen behind other browsers, notably Chrome but even, in areas, Edge. Mozilla has suffered several massive defeats in recent years, including a drubbing over mobile operating systems and a lesser beating from a stab at in-browser advertising. Lately, it has rededicated itself to Firefox, but the jury remains undecided, with some, including a former CTO, maintaining that the browser has no chance of unseating Chrome.

Last month, Firefox accounted for 12% of all browsers used worldwide, about a fifth of the share owned by Chrome and half that of a combined Internet Explorer (IE) and Edge. That May number was the highest of the year so far, but it was also nearly identical to Firefox’s share of 24 months earlier, showing how mired the browser had become.

More Businesses Focusing On Own Mobile App Development

June 15, 2017 by  
Filed under Mobile

Companies are wading into the mobile app development waters as a way to improve business.

The number of enterprises now building custom mobile apps — many of them simple apps designed to handle business processes — rose significantly in 2016, according to Gartner’s annual study of mobile app development platforms.

In 2015, about 60% of organizations were engaged in mobile app development. Last year, that number jumped to about 73%, according to the study, which also evaluated 35 mobile app development platform (MADP) vendors for this year’s “Magic Quadrant.”

MADPs provide tools, technologies, components and services that become the key building blocks used by enterprises to create custom mobile apps, mobile web apps or websites, according to Gartner.

The majority of custom app development is still geared toward more code-centric custom apps intended for the development of customer-facing apps as well as business critical apps for a partners and distributors. The average number of mobile apps being deployed per company remains relatively small: eight apps.

Enterprises are increasingly using MADP to create self-service apps for human resources to track things such as employee vacation time or to get approvals or enrollments, as well as testing daily operations and for collaboration, according to Jason Wong, a research director at Gartner.

“Field service and sales continue to see a majority of the investment as well because of new capabilities that are still emerging — things like augmented reality, integrating list sensors and IoT, using voice and voice control,” Wong said. “That’s all new stuff that needs to be incorporated into apps.”

Increasingly, MADPs are including support for wearables, chatbots, virtual personal assistants (VPAs) and conversational user-interface endpoints through the same services and APIs used for mobile apps and the web.

“Close to 90% of those companies saying they were doing some kind of mobile app development were doing customer mobile apps…in order to create some type of differentiating solutions,” Wong said.

Delphi Automotive To Offer On-Demand Self-driving Shuttle Service

June 8, 2017 by  
Filed under Around The Net

Delphi Automotive PLC plans on joining forces with Paris-based Transdev Group, a public transport service controlled by the French government, to develop an automated on-demand shuttle service in Europe, according to announcements made by both companies.

It is the latest in a growing web of global alliances aimed at putting self-driving vehicles on the road over the next four years.

In a joint statement, Delphi and Transdev will test driverless vehicles in Normandy and outside Paris, in advance of building a commercial service that could be deployed in other markets.

The Delphi-Transdev partnership will provide “a clear path to commercializing automated mobility on demand,” said Glen De Vos, Delphi’s chief technology officer, in a media briefing.

Delphi is contributing a self-driving system that it has been developing with Israeli mapping and vision expert Mobileye NV, which is being acquired by U.S. chipmaker Intel Corp.

Transdev will provide dispatch, routing and remote control-command services, the companies’ statement said.

Delphi has been building its expertise and capability in self-driving vehicles through partnerships, investments and acquisitions.

Several of its affiliates will participate in the project with Transdev, including Ottomatika (vehicle control software), Control-Tec (real-time data analytics) and Movimento (over-the-air software updates).

In May, German automaker BMW AG announced that Delphi will join a self-driving partnership that includes Intel and Mobileye.

Transdev is a mobility services provider that is controlled by Caisse des Depots, an investment arm of the French government. Veolia Environnement SA, the French waste management company, holds a 30-percent stake.

Transdev operates public and private transport services in 19 countries.

Earlier this year, it formed a research partnership with Renault SA and Nissan Motor Co to develop a driverless system for public transport and on-demand use.

ARM Dives Into The A.I. Market With The Cortex A75 And A55

June 8, 2017 by  
Filed under Computing

ARM has taken to Computex to unveil the Cortex-A75 and A55 CPUs, which it claims are its most powerful to date.

The flagship Cortex-A75, which is based on ARM’s flexible and scalable Dymaniq microarchitecture and destined for high-end smartphones, offers a 22 per cent improvement in single-threaded performance compared to its Cortex-A73 predecessor and up to 50 per cent more performance in multithreaded use cases. What’s more, according to ARM, the CPU can use just 2W of power and offer a 30 per cent increase in performance on large-screened devices.

A timely leak suggests that Qualcomm’s as-yet-unannounced Snapdragon 845, set to debut inside the Galaxy S9, will be powered by Cortex-A75 cores.

Nandan Nayampally, VP and general manager of ARM’s CPU group, remarked: “I have been at ARM for over a dozen years and can’t remember being this excited about a product delivering such a boost to single-threaded performance without compromising our efficiency leadership.

“The Cortex-A75 delivers a massive 50 percent uplift in performance and greater multicore capabilities, enabling our partners to address multiple high-performance use cases including laptops, networking and servers, all within a smartphone power profile.”

The Cortex-A55 is described by ARM as its “most versatile high-efficiency processor”, and claims the new CPU offers 15 percent more power efficiency compared to its predecessor, in addition to 10 times more scalability, and two times the memory efficiency. 

As well as touting performance upgrades, ARM is – unsurprisingly – touting its new CPUs as optimised for artificial intelligence (AI) and machine learning. Both the Cortex-A75 and A55 put ARM feature dedicated instructions for AI performance tasks and set ARM, it claims, on a trajectory to deliver “50x AI performance increases over the next three to five years.”

ARM also on Monday launched the Mali-G72 GPU, which it claims further optimises its SoCs for AI and machine learning tasks. Touting it as its “most efficient yet”, the G72 claims a 1.4 times performance if the Mali-G71 before it and a 25 per cent boost in efficiency.

ARM expects that the first implementations of its new CPU and GPU technologies to appear either in Q4 2017 or Q1 2018. 

Courtesy-TheInq

Toyota Infotainment System Puts Weight Behind Linux’s Open Source Technology

June 1, 2017 by  
Filed under Around The Net

Toyota Motor Corp announced that the infotainment system of its re-designed Camry sedan to be sold in the United States will run on a Linux-based, open-source technology platform as it tries to keep up with tech firms in developing software for cars.

With the Automotive Grade Linux (AGL) system in a mainstay model, Toyota aims to have the flexibility to customise its software, while it would also keep user data that could otherwise be captured by CarPlay from Apple Inc or Android Auto from Alphabet Inc’s Google – applications which enable users to access smartphone data through vehicle infotainment systems.

Toyota is among 10 global automakers working with suppliers and technology companies to jointly build AGL, a basic, open-source platform for vehicle applications which automakers can customise, eliminating the need to code systems from the ground up for each vehicle model.

Developing the platform in collaboration with Mazda Motor Corp, Suzuki Motor Corp, Daimler AG and others will reduce development time and costs, Toyota said, and create an industry standard platform to operate in-vehicle features including music and navigation applications.

The platform can also be used to support future advanced technologies, including self-driving functions and connected car services.

“It’s very necessary to reduce the overhead of duplication work among our suppliers so they can spend more time to create new things rather than maintaining fragmentary codes,” said Kenichi Murata, group manager of Connected Strategy and Planning at Toyota.

Cars typically require over 100 million lines of computer code as automakers pack as much technology as possible to attract buyers.

So much so that coding has become an increasingly cumbersome part of vehicle development, which takes years, compared with the mere months it takes for tech firms to develop apps.

In addition, the process requires constant updating to keep up with technology developments and which results in disparate interfaces between automaker’s products.

The latest Camry sedan to be launched in coming months will use AGL to operate its suite of in-vehicle apps, and the Japanese automaker said it planned to expand the platform to other Toyota and Lexus vehicles in North America and elsewhere.

Roughly 70 percent of the operating platform for the latest system consists of largely generic coding, while the remaining 30 percent was customized for the Camry, Murata said.

At the moment, automakers make vehicles compatible with CarPlay and Android Auto. While this enables users to connect smartphones to cars, Dan Cauchy, general manager of automotive at the Linux Foundation, said it makes it difficult for automakers to have control over customizing their platforms.

“It comes down to an automaker wanting to customize their operating platform to their liking and not having a third party dictating what the applications are going to be for the vehicle,” he said.

“A lot of automakers want that control.”

Next Page »