Monday, January 18. 2016
Researchers at MIT’s Computer Science and Artificial Intelligence Lab (CSAIL) have created an algorithm they claim can predict how memorable or forgettable an image is almost as accurately as a human — which is to say that their tech can predict how likely a person would be to remember or forget a particular photo.
The algorithm performed 30 per cent better than existing algorithms and was within a few percentage points of the average human performance, according to the researchers.
The team has put a demo of their tool online here, where you can upload your selfie to get a memorability score and view a heat map showing areas the algorithm considers more or less memorable. They have also published a paper on the research which can be found here.
Here are some examples of images I ran through their MemNet algorithm, with resulting memorability scores and most and least forgettable areas depicted via heat map:
Potential applications for the algorithm are very broad indeed when you consider how photos and photo-sharing remains the currency of the social web. Anything that helps improve understanding of how people process visual information and the impact of that information on memory has clear utility.
The team says it plans to release an app in future to allow users to tweak images to improve their impact. So the research could be used to underpin future photo filters that do more than airbrush facial features to make a shot more photogenic — but maybe tweak some of the elements to make the image more memorable too.
Beyond helping people create a more lasting impression with their selfies, the team envisages applications for the algorithm to enhance ad/marketing content, improve teaching resources and even power health-related applications aimed at improving a person’s capacity to remember or even as a way to diagnose errors in memory and perhaps identify particular medical conditions.
The MemNet algorithm was created using deep learning AI techniques, and specifically trained on tens of thousands of tagged images from several different datasets all developed at CSAIL — including LaMem, which contains 60,000 images each annotated with detailed metadata about qualities such as popularity and emotional impact.
Publishing the LaMem database alongside their paper is part of the team’s effort to encourage further research into what they say has often been an under-studied topic in computer vision.
Asked to explain what kind of patterns the deep-learning algorithm is trying to identify in order to predict memorability/forgettability, PhD candidate at MIT CSAIL, Aditya Khosla, who was lead author on a related paper, tells TechCrunch: “This is a very difficult question and active area of research. While the deep learning algorithms are extremely powerful and are able to identify patterns in images that make them more or less memorable, it is rather challenging to look under the hood to identify the precise characteristics the algorithm is identifying.
“In general, the algorithm makes use of the objects and scenes in the image but exactly how it does so is difficult to explain. Some initial analysis shows that (exposed) body parts and faces tend to be highly memorable while images showing outdoor scenes such as beaches or the horizon tend to be rather forgettable.”
The research involved showing people images, one after another, and asking them to press a key when they encounter an image they had seen before to create a memorability score for images used to train the algorithm. The team had about 5,000 people from the Amazon Mechanical Turk crowdsourcing platform view a subset of its images, with each image in their LaMem dataset viewed on average by 80 unique individuals, according to Khosla.
In terms of shortcomings, the algorithm does less well on types of images it has not been trained on so far, as you’d expect — so it’s better on natural images and less good on logos or line drawings right now.
“It has not seen how variations in colors, fonts, etc affect the memorability of logos, so it would have a limited understanding of these,” says Khosla. “But addressing this is a matter of capturing such data, and this is something we hope to explore in the near future — capturing specialized data for specific domains in order to better understand them and potentially allow for commercial applications there. One of those domains we’re focusing on at the moment is faces.”
The team has previously developed a similar algorithm for face memorability.
Discussing how the planned MemNet app might work, Khosla says there are various options for how images could be tweaked based on algorithmic input, although ensuring a pleasing end photo is part of the challenge here. “The simple approach would be to use the heat map to blur out regions that are not memorable to emphasize the regions of high memorability, or simply applying an Instagram-like filter or cropping the image a particular way,” he notes.
“The complex approach would involve adding or removing objects from images automatically to change the memorability of the image — but as you can imagine, this is pretty hard — we would have to ensure that the object size, shape, pose and so on match the scene they are being added to, to avoid looking like a photoshop job gone bad.”
Looking ahead, the next step for the researchers will be to try to update their system to be able to predict the memory of a specific person. They also want to be able to better tailor it for individual “expert industries” such as retail clothing and logo-design.
How many training images they’d need to show an individual person before being able to algorithmically predict their capacity to remember images in future is not yet clear. “This is something we are still investigating,” says Khosla.
Wednesday, January 06. 2016
Via Slash Gear
Everyone on the internet has come across at least couple error codes, the most well-known being 404, for page not found, while other common ones include 500, for internal server error, or 403, for a "forbidden" page. However, with latter, there's the growing issue of why a certain webpage has become forbidden, or who made it so. In an effort to address things like censorship or "legal obstacles," a new code has been published, to be used when legal demands require access to a page be blocked: error 451.
The number is a knowing reference Fahrenheit 451, the novel by Ray Bradbury that depicted a dystopian future where books are banned for spreading dissenting ideas, and in burned as a way censor the spread of information. The code itself was approved for use by the Internet Engineering Steering Group (IESG), which helps maintain internet standards.
The idea for code 451 originally came about around 3 years ago, when a UK court ruling required some websites to block The Pirate Bay. Most sites in turn used the 403 "forbidden" code, making it unclear to users about what the issue was. The goal of 451 is to eliminate some of the confusion around why sites may be blocked.
The use of the code is completely voluntary, however, and requires developers to begin adopting it. But if widely implemented, it should be able to communicate to users that some information has been taken down because of a legal demand, or is being censored by a national government.
Monday, January 04. 2016
Via Tech Crunch
As humanity debates the threats and opportunities of advanced artificial intelligence, we are simultaneously enabling that technology through the increasing use of personalization that is understanding and anticipating our needs through sophisticated machine learning solutions.
In effect, while using personalization technologies in our everyday lives, we are contributing in a real way to the development of the intelligent systems we purport to fear.
Perhaps uncovering the currently inaccessible personalization systems is crucial for creating a sustainable relationship between humans and super–intelligent machines?
From Machines Learning About You…
Industry giants are currently racing to develop more intelligent and lucrative AI solutions. Google is extending the ways machine learning can be applied in search, and beyond. Facebook’s messenger assistant M is combining deep learning and human curators to achieve the next level in personalization.
With your iPhone you’re carrying Apple’s digital assistant Siri with you everywhere; Microsoft’s counterpart Cortana can live in your smartphone, too. IBM’s Watson has highlighted its diverse features, varying from computer vision and natural language processing to cooking skills and business analytics.
At the same time, your data and personalized experiences are used to develop and train the machine learning systems that are powering the Siris, Watsons, Ms and Cortanas. Be it a speech recognition solution or a recommendation algorithm, your actions and personal data affect how these sophisticated systems learn more about you and the world around you.
The less explicit fact is that your diverse interactions — your likes, photos, locations, tags, videos, comments, route selections, recommendations and ratings — feed learning systems that could someday transform into super–intelligent AIs with unpredictable consequences.
As of today, you can’t directly affect how your personal data is used in these systems.
In these times, when we’re starting to use serious resources to contemplate the creation of ethical frameworks for super–intelligent AIs-to-be, we also should focus on creating ethical terms for the use of personal data and the personalization technologies that are powering the development of such systems.
To make sure that you as an individual continue to have a meaningful agency in the emerging algorithmic reality, we need learning algorithms that are on your side and solutions that augment and extend your abilities. How could this happen?
…To Machines That Learn For You
Smart devices extend and augment your memory (no forgotten birthdays) and brain processing power (no calculating in your head anymore). And they augment your senses by letting you experience things beyond your immediate environment (think AR and VR).
The web itself gives you access to a huge amount of diverse information and collective knowledge. The next step would be that smart devices and systems enhance and expand your abilities even more. What is required for that to happen in a human-centric way?
Data Awareness And Algorithmic Accountability
Algorithmic systems and personal data are too often seen as something abstract, incomprehensible and uncontrollable. Concretely, how many really stopped using Facebook or Google after PRISM came out in the open? Or after we learned that we are exposed to continuous A/B testing that is used to develop even more powerful algorithms?
More and more people are getting interested in data ethics and algorithmic accountability. Academics are already analyzing the effects of current data policies and algorithmic systems. Educational organizations are starting to emphasize the importance of coding and digital literacy.
Initiatives such as VRM, Indie Web and MyData are raising awareness on alternative data ecosystems and data management practices. Big companies like Apple and various upcoming startups are bringing personal data issues to the mainstream discussion.
Yet we still need new tools and techniques to become more data aware and to see how algorithms can be more beneficial for us as unique individuals. We need apps and data visualizations with great user experience to illuminate the possibilities of more human-centric personalization.
It’s time to create systems that evaluate algorithmic biases and keep them in check. More accessible algorithms and transparent data policies are created only through wider collaboration that brings together companies, developers, designers, users and scientists alike.
Personal Machine Learning Systems
Personalization technologies are already augmenting your decision making and future thinking by learning from you and recommending to you what to see and do next. However, not on your own terms. Rather than letting someone else and their motives and values dictate how the algorithms work and affect your life, it’s time to create solutions, such as algorithmic angels, that let you develop and customize your own algorithms and choose how they use your data.
When you’re in control, you can let your personal learning system access previously hidden data and surface intimate insights about your own behavior, thus increasing your self-awareness in an actionable way.
Personal learners could help you develop skills related to work or personal life, augmenting and expanding your abilities. For example, learning languages, writing or playing new games. Fitness or mediation apps powered by your personal algorithms would know you better than any personal trainer.
Google’s experiments with deep learning and image manipulation showed us how machine learning could be used to augment creative output. Systems capable of combining your data with different materials like images, text, sound and video could expand your abilities to see and utilize new and unexpected connections around you.
In effect, your personal algorithm can take a mind-expanding “trip” on your behalf, letting you see music or sense other dimensions beyond normal human abilities. By knowing you, personal algorithms can expose you to new diverse information, thus breaking your existing filter bubbles.
Additionally, people tinkering with their personal algorithms would create more “citizen algorithm experts,” like “citizen scientists,” coming up with new ideas, solutions and observations, stemming from real live situations and experiences.
However, personally adjustable algorithms for the general public are not happening overnight, even though Google recently open-sourced parts of its machine learning framework. But it’s possible to see how today’s personalization experiences can someday evolve into customizable algorithms that strengthen your agency and capacity to deal with other algorithmic systems.
The next step is that your personal algorithms become a more concrete part of you, continuously evolving with you by learning from your interactions both in digital and physical environments. Your algorithmic self combines your personal abilities and knowledge with machine learning systems that adapt to you and work for you. Be it your smartwatch, self-driving car or an intelligent home system, they can all be spirited by your algorithmic self.
Your algorithmic self also can connect with other algorithmic selves, thus empowering you with the accumulating collective knowledge and intelligence. To expand your existing skills and faculties, your algorithmic self also starts to learn and act on its own, filtering information, making online transactions and comparing best options on your behalf. It makes you more resourceful, and even a better person, when you can concentrate on things that really require your human presence and attention.
Partly algorithmic humans are not bound by existing human capabilities; new skills and abilities emerge when human intelligence is extended with algorithmic selves. For example, your algorithmic self can multiply to execute different actions simultaneously. Algorithmic selves could also create simple simulations by playing out different scenarios involving your real-life choices and their consequences, helping you to make better decisions in the future.
Algorithmic selves — tuned by your data and personal learners — also could be the key when creating invasive human-computer interfaces that connect digital systems directly to your brain, expanding human brain concretely beyond the “wetware.”
But to ensure that your algorithmic self works for your benefit, could you trust someone building that for you without you participating in the process?
Machine learning expert Pedro Domingos says in his new book “The Master Algorithm” that “[m]achine learning will not single-handedly determine the future… it’s what we decide to do with it that counts.”
Machines are still far from human intelligence. No one knows exactly when super–intelligent AIs will become concrete reality. But developing personal machine learning systems could enable us to interact with any algorithmic entities, be it an obtrusive recommendation algorithm or a super–intelligent AI.
In general, being more transparent on how learning algorithms work and use our data could be crucial for creating ethical and sustainable artificial intelligence. And potentially, maybe we wouldn’t need to fear being overpowered by our own creations.
Friday, January 01. 2016
what3words is a geocoding system for the simple communication of precise locations. what3words encodes geographic co-ordinates into 3 dictionary words (for example, the Statue of Liberty is located at planet.inches.most). what3words is different from other alphanumeric location systems and GPS coordinates in that it displays 3 words rather than long strings of numbers or random letters or numbers. what3words has an iOS App, Android App, a website and an API that enables bi-directional conversion of what3words address and latitude/longitude co-ordinates.
Tuesday, November 24. 2015
Via The Daily Dot
People with Android devices might be a bit frustrated with Google after a report from the New York District Attorney's office provided detailed information about smartphone security, and Google's power to access devices when asked to by law enforcement. The report went viral on Reddit over the weekend.
Google can unlock many Android phones remotely when given a search warrant, bypassing lock codes on particular devices. The report reads:
Forensic examiners are able to bypass passcodes on some of those [Android] devices using a variety of forensic techniques. For some other types of Android devices, Google can reset the passcodes when served with a search warrant and an order instructing them to assist law enforcement to extract data from the device. This process can be done by Google remotely and allows forensic examiners to view the contents of a device.
When compared to Apple devices, which encrypt by default on iOS 8 and later, Google's seemingly lax protection is irksome. The report continues:
For Android devices running operating systems Lollipop 5.0 and above, however, Google plans to use default full-disk encryption, like that being used by Apple, that will make it impossible for Google to comply with search warrants and orders instructing them to assist with device data extraction. Generally, users have the option to enable full-disk encryption on their current Android devices, whether or not the device is running Lollipop 5.0, but doing so causes certain inconveniences, risks, and performance issues, which are likely to exist until OEMs are required to standardize certain features.
In October, Google announced that new devices that ship with the Marshmallow 6.0 operating system (the most recent version of Android) must enable full-disk encryption by default. Nexus devices running Lollipop 5.0 are encrypted by default as well. This means that Google is unable to bypass lock codes on those devices. However, because of the massive fragmentation of Android devices and operating systems, Google can still access lots of Android devices running older versions when asked to by law enforcement.
And despite the encryption updates to the Android compatibility documentation, a number of devices are exempt from full-disk encryption, including older devices, devices without a lock screen, and those that don't meet the minimum security requirements.
The number of devices that actually have full-disk encryption appears to be low. Just 0.3 percent of Android devices are running Marshmallow and more than 25 percent of Android devices are running Lollipop 5.0, but most of those aren't Nexus, according to ZDNet.
When compared to Apple, Google's security appears lacking. Apple made encryption mandatory in iOS 8 back in 2014, which of course extends to iOS 9, its most recent mobile OS update. Data shows that 67 percent of Apple users are on iOS 9, and 24 percent of devices are still on iOS 8. Just nine percent of devices run an older version of iOS.
Android users are often at the mercy of carriers who decide when to roll out Android updates, which is an obstacle for some Android owners who want the latest OS.
If you do have a compatible device and want to enable encryption, head over to your security settings and select "encrypt device."
Tuesday, October 06. 2015
Binaural audio consists in recording audio sounds in the exact same way as humans are naturally perceiving it. Listening to binaural audio may only be performed by the use of a headset, making the resulting experience hyper-realistic. Many more details may be found on the binaural recording wikipedia page.
Wednesday, September 09. 2015
We invariably imagine electronic devices to be made from silicon chips, with which computers store and process information as binary digits (zeros and ones) represented by tiny electrical charges. But it need not be this way: among the alternatives to silicon are organic mediums such as DNA.
DNA computing was first demonstrated in 1994 by Leonard Adleman who encoded and solved the travelling salesman problem, a maths problem to find the most efficient route for a salesman to take between hypothetical cities, entirely in DNA.
Deoxyribonucleaic acid, DNA, can store vast amounts of information encoded as sequences of the molecules, known as nucleotides, cytosine (C), guanine (G), adenine (A), or thymine (T). The complexity and enormous variance of different species’ genetic codes demonstrates how much information can be stored within DNA encoded using CGAT, and this capacity can be put to use in computing. DNA molecules can be used to process information, using a bonding process between DNA pairs known as hybridisation. This takes single strands of DNA as input and produces subsequent strands of DNA through transformation as output.
Since Adleman’s experiment, many DNA-based “circuits” have been proposed that implement computational methods such as Boolean logic, arithmetical formulas, and neural network computation. Called molecular programming, this approach applies concepts and designs customary to computing to nano-scale approaches appropriate for working with DNA.
It’s circuitry, but not as we know it. Caltech/Lulu Qian, CC BY
In this sense “programming” is really biochemistry. The “programs” created are in fact methods of selecting molecules that interact in a way that achieves a specific result through the process of DNA self-assembly, where disordered collections of molecules will spontaneously interact to form the desired arrangement of strands of DNA.
DNA can also be used to control motion, allowing for DNA-based nano-mechanical devices. This was first achieved by Bernard Yurke and colleagues in 2000, who created from DNA strands a pair of tweezers that opened and pinched. Later experiments such as by Shelley Wickham and colleagues in 2011 and at Andrew Turberfield’s lab at Oxford demonstrated nano-molecular walking machines made entirely from DNA that could traverse set routes.
One possible application is that such a nano-robot DNA walker could progress along tracks making decisions and signal when reaching the end of the track, indicating computation has finished. Just as electronic circuits are printed onto circuit boards, DNA molecules could be used to print similar tracks arranged into logical decision trees on a DNA tile, with enzymes used to control the decision branching along the tree, causing the walker to take one track or another.
DNA walkers can also carry molecular cargo, and so could be used to deliver drugs inside the body.
Why DNA Computing?
DNA molecules’ many appealing features include their size (2nm width), programmability and high storage capacity – much greater than their silicon counterparts. DNA is also versatile, cheap and easy to synthesise, and computing with DNA requires much less energy than electric powered silicon processors.
Its drawback is speed: it currently takes several hours to compute the square root of a four digit number, something that a traditional computer could compute in a hundredth of a second. Another drawback is that DNA circuits are single-use, and need to be recreated to run the same computation again.
Perhaps the greatest advantage of DNA over electronic circuits is that it can interact with its biochemical environment. Computing with molecules involves recognising the presence or absence of certain molecules, and so a natural application of DNA computing is to bring such programmability into the realm of environmental biosensing, or delivering medicines and therapies inside living organisms.
DNA programs have already been put to medical uses, such as diagnosing tuberculosis. Another proposed use is a nano-biological “program” by Ehud Shapiro of the Weizmann Institute of Science in Israel, termed the “doctor in the cell” that targets cancer molecules. Other DNA programs for medical applications target lymphocytes (a type of white blood cell), which are defined by the presence or absence of certain cell markers and so can be naturally detected with true/false Boolean logic. However, more effort is required before we can inject smart drugs directly into living organisms.
Future of DNA Computing
Taken broadly, DNA computation has enormous future potential. Its huge storage capacity, low energy cost, ease of manufacturing that exploits the power of self-assembly and its easy affinity with the natural world are an entry to nanoscale computing, possibly through designs that incorporate both molecular and electronic components. Since its inception, the technology has progressed at great speed, delivering point-of-care diagnostics and proof-of-concept smart drugs – those that can make diagnostic decisions about the type of therapy to deliver.
There are many challenges, of course, that need to be addressed so that the technology can move forward from the proof-of-concept to real smart drugs: the reliability of the DNA walkers, the robustness of DNA self-assembly, and improving drug delivery. But a century of traditional computer science research is well placed to contribute to developing DNA computing through new programming languages, abstractions, and formal verification techniques – techniques that have already revolutionised silicon circuit design, and can help launch organic computing down the same path.
Tuesday, June 30. 2015
The ancient Library of Alexandria may have been the largest collection of human knowledge in its time, and scholars still mourn its destruction. The risk of so devastating a loss diminished somewhat with the advent of the printing press and further still with the rise of the Internet. Yet centralized repositories of specialized information remain, as does the threat of a catastrophic loss.
Take GitHub, for example.
GitHub has in recent years become the world’s biggest collection of open source software. That’s made it an invaluable education and business resource. Beyond providing installers for countless applications, GitHub hosts the source code for millions of projects, meaning anyone can read the code used to create those applications. And because GitHub also archives past versions of source code, it’s possible to follow the development of a particular piece of software and see how it all came together. That’s made it an irreplaceable teaching tool.
The odds of Github meeting a fate similar to that of the Library of Alexandria are slim. Indeed, rumor has it that Github soon will see a new round of funding that will place the company’s value at $2 billion. That should ensure, financially at least, that GitHub will stay standing.
But GitHub’s pending emergence as Silicon Valley’s latest unicorn holds a certain irony. The ideals of open source software center on freedom, sharing, and collective benefit—the polar opposite of venture capitalists seeking a multibillion-dollar exit. Whatever its stated principles, GitHub is under immense pressure to be more than just a sustainable business. When profit motives and community ideals clash, especially in the software world, the end result isn’t always pretty.
Sourceforge: A Cautionary Tale
Sourceforge is another popular hub for open source software that predates GitHub by nearly a decade. It was once the place to find open source code before GitHub grew so popular.
There are many reasons for GitHub’s ascendance, but Sourceforge hasn’t helped its own cause. In the years since career services outfit DHI Holdings acquired it in 2012, users have lamented the spread of third-party ads that masquerade as download buttons, tricking users into downloading malicious software. Sourceforge has tools that enable users to report misleading ads, but the problem has persisted. That’s part of why the team behind GIMP, a popular open source alternative to Adobe Photoshop, quit hosting its software on Sourceforge in 2013.
Instead of trying to make nice, Sourceforge stirred up more hostility earlier this month when it declared the GIMP project “abandoned” and began hosting “mirrors” of its installer files without permission. Compounding the problem, Sourceforge bundled installers with third party software some have called adware or malware. That prompted other projects, including the popular media player VLC, the code editor Notepad++, and WINE, a tool for running Windows apps on Linux and OS X, to abandon ship.
It’s hard to say how many projects have truly fled Sourceforge because of the site’s tendency to “mirror” certain projects. If you don’t count “forks” in GitHub—copies of projects developers use to make their own tweaks to the code before submitting them to the main project—Sourceforge may still host nearly as many projects as GitHub, says Bill Weinberg of Black Duck Software, which tracks and analyzes open source software.
But the damage to Sourceforge’s reputation may already have been done. Gaurav Kuchhal, managing director of the division of DHI Holdings that handles Sourceforge, says the company stopped its mirroring program and will only bundle installers with projects whose originators explicitly opt in for such add-ons. But misleading “download” ads likely will continue to be a game of whack-a-mole as long as Sourceforge keeps running third-party ads. In its hunt for revenue, Sourceforge is looking less like an important collection of human knowledge and more like a plundered museum full of dangerous traps.
No Ads (For Now)
GitHub has a natural defense against ending up like this: it’s never been an ad-supported business. If you post your code publicly on GitHub, the service is free. This incentivizes code-sharing and collaboration. You pay only to keep your code private. GitHub also makes money offering tech companies private versions of GitHub, which has worked out well: Facebook, Google and Microsoft all do this.
Still, it’s hard to tell how much money the company makes from this model. (It’s certainly not saying.) Yes, it has some of the world’s largest software companies as customers. But it also hosts millions of open source projects free of charge, without ads to offset the costs storage, bandwidth, and the services layered on top of all those repos. Investors will want a return eventually, through an acquisition or IPO. Once that happens, there’s no guarantee new owners or shareholders will be as keen on offering an ad-free loss leader for the company’s enterprise services.
Other freemium services that have raised large rounds of funding, like Box and Dropbox, face similar pressures. (Box even more so since going public earlier this year.) But GitHub is more than a convenient place to store files on the web. It’s a cornerstone of software development—a key repository of open-source code and a crucial body of knowledge. Amassing so much knowledge in one place raises the specter of a catastrophic crash and burn or disastrous decay at the hands of greedy owners loading the site with malware.
Yet GitHub has a defense mechanism the librarians of ancient Alexandria did not. Their library also was a hub. But it didn’t have Git.
The “Git” part of GitHub is an open source technology that helps programmers manage changes in their code. Basically, a team will place a master copy of the code in a central location, and programmers make copies on their own computers. These programmers then periodically merge their changes with the master copy, the “repository” that remains the canonical version of the project.
Git’s “versioning” makes managing projects much easier when multiple people must make changes to the original code. But it also has an interesting side effect: everyone who works on a GitHub project ends up with a copy own their computers. It’s as if everyone who borrowed a book from the library could keep a copy forever, even after returning it. If GitHub vanished entirely, it could be rebuilt using individual users’ own copies of all the projects. It would take ages to accomplish, but it could be done.
Still, such work would be painful. In addition to the source code itself, GitHub is also home to countless comments, bug reports and feature requests, not to mention the rich history of changes. But the decentralized nature of Git does make it far easier to migrate projects to other hosts, such as GitLab, an open source alternative to GitHub that you can run on your own server.
In short, if GitHub as we know it went away, or under future financial pressures became an inferior version of itself, the world’s code will survive. Libraries didn’t end with Alexandria. The question is ultimately whether GitHub will find ways to stay true to its ideals while generating returns—or wind up the stuff of legend.
Friday, June 19. 2015
Google, Microsoft, Mozilla And Others Team Up To Launch WebAssembly, A New Binary Format For The Web
Via Tech Crunch
The idea is that WebAssembly will provide developers with a single compilation target for the web that will, eventually, become a web standard that’s implemented in all browsers.
Mozilla’s asm.js has long aimed to bring near-native speeds to the web. Google’s Native Client project for running native code in the browser had similar aims, but got relatively little traction. It looks like WebAssemly may be able to bring the best of these projects to the browser now.
As a first step, the WebAssembly team aims to offer about the same functionality as asm.js (and developers will be able to use the same Emscripten tool for WebAssembly as they use for compiling asm.js code today).
It’s not often that we see all the major browser vendors work together on a project like this, so this is definitely something worth watching in the months and years ahead.
Thursday, June 18. 2015
Via Tech Crunch
While companies like Facebook have been relatively open about their data center networking infrastructure, Google has generally kept pretty quiet about how it connects the thousands of servers inside its data centers to each other (with a few exceptions). Today, however, the company revealed a bit more about the technology that lets its servers talk to each other.
It’s no secret that Google often builds its own custom hardware for its data centers, but what’s probably less known is that Google uses custom networking protocols that have been tweaked for use in its data centers instead of relying on standard Internet protocols to power its networks.
Google says its current ‘Jupiter’ networking setup — which represents the fifth generation of the company’s efforts in this area — offers 100x the capacity of its first in-house data center network. The current generation delivers 1 Petabit per second of bisection bandwidth (that is, the bandwidth between two parts of the network). That’s enough to allow 100,000 servers to talk to each other at 10GB/s each.
Google’s technical lead for networking, Amin Vahdat, notes that the overall network control stack “has more in common with Google’s distributed computing architectures than traditional router-centric Internet protocols.”
Here is how he describes the three key principles behind the design of Google’s data center networks:
(Page 1 of 52, totaling 512 entries) » next page
Show tagged entries