Shrinking transistors have powered 50 years of advances in computing—but now other ways must be found to make computers more capable.
Mobile apps, video games, spreadsheets, and accurate weather forecasts: that’s just a sampling of the life-changing things made possible by the reliable, exponential growth in the power of computer chips over the past five decades.
But in a few years technology companies may have to work harder to bring us advanced new use cases for computers. The continual cramming of more silicon transistors onto chips, known as Moore’s Law, has been the feedstock of exuberant innovation in computing. But it looks to be slowing to a halt.
“We have to ask, is this going to be a problem for areas like mobile devices, data centers, and self-driving cars?” says Thomas Wenisch, an assistant professor at the University of Michigan. “I think yes, but on different timescales.”
Moore’s Law is named after Intel cofounder Gordon Moore. He observed in 1965 that transistors were shrinking so fast that every year twice as many could fit onto a chip, and in 1975 adjusted the pace to a doubling every two years.
The chip industry has kept Moore’s prediction alive, with Intel leading the charge. And computing companies have found plenty to do with the continual supply of extra transistors. But Intel pushed back its next transistor technology, with features as small as 10 nanometers, from 2016 to late 2017. The company has also decided to increase the time between future generations (see “Intel Puts the Brakes on Moore’s Law”). And a technology roadmap for Moore’s Law maintained by an industry group, including the world’s largest chip makers, is being scrapped. Intel has suggested silicon transistors can only keep shrinking for another five years.
The computers in our pockets will probably feel the effects later than other types of computing devices, Wenisch guesses. Mobile devices are powered by chips made by companies other than Intel, and they've generally been slightly behind in transistor technology. And mobile processors don’t make full use of some design techniques well established in more powerful processors for non-roving machines, he says.
“You probably have a generation or two more runway in mobile,” says Wenisch. However, many useful things that mobile devices can do rest on the power of billion-dollar data centers, where the end of Moore’s Law would be a more immediate headache. Companies such as Google and Microsoft eagerly gobble up every new generation of the most advanced chips, packed more densely with transistors.
Wenisch says companies such as Intel, which dominates the server chip market, and their largest customers will have to get creative. Alternative ways to get more computing power include working harder to improve the design of chips and making chips specialized to accelerate particular crucial algorithms.
Strong demand for silicon tuned for algebra that’s crucial to a powerful machine-learning technique called deep learning seems inevitable, for example. Graphics chip company Nvidia and several startups are already moving in that direction (see “A $2 Billion Chip to Accelerate Artificial Intelligence”).
Microsoft and Intel are also working on the idea of running some code on reconfigurable chips called FPGAs for greater efficiency (see “Microsoft Says Reprogrammable Chips Will Make AI Smarter”). Intel spent nearly $17 billion to acquire leading FPGA manufacturer Altera last year and is adapting its technology to data centers.
Horst Simon, deputy director of Lawrence Berkeley National Laboratory, says the world’s most powerful calculating machines appear to be already feeling the effects of Moore’s Law’s end times. The world’s top supercomputers aren’t getting better at the rate they used to.
“For the last three years we’ve seen a kind of stagnation,” says Simon. That’s bad news for research programs reliant on supercomputers, such as efforts to understand climate change, develop new materials for batteries and superconductors, and improve drug design.
Simon says the coming plateau in transistor density will stir more interest in redrawing the basic architecture of computers among supercomputer and data-center designers. Getting rid of certain design features dating from the 1940s could unlock huge efficiency gains (see “Machine Dreams”). Yet taking advantage of those would require rethinking the design of many types of software, and would require programmers to change their habits.
Whatever kind of computer you’re interested in, the key question is whether the creative avenues left open to computing companies can provide similar payoffs to Moore’s Law after it ends, says Neil Thompson, an assistant professor at MIT Sloan School. “We know that those other things matter, but the question is, are they of the same scale?” he says.
One reason to think they might not be is that companies will have to work together in new and complicated ways, without the common heartbeat that used to keep the industry’s product and R&D plans in sync.
“One of the biggest benefits of Moore's Law is as a coӧrdination device,” says Thompson. “I know that in two years we can count on this amount of power and that I can develop this functionality—and if you’re Intel you know that people are developing for that and that there's going to be a market for a new chip.”
Without that common music to dance to, advances in computing power that benefit all kinds of companies, not just ones with mutually strong incentives to collaborate, could be less common.
The first of Amazon’s new private-label brands could start appearing online in the coming weeks.
Amazon.com Inc. in the coming weeks is set to roll out new lines of private-label brands that will include its first broad push into perishable foods, according to people familiar with the matter.
The new brands with names like Happy Belly, Wickedly Prime and Mama Bear will include nuts, spices, tea, coffee, baby food and vitamins, as well as household items such as diapers and laundry detergents, these people said.
The first of the brands could begin appearing on Amazon’s namesake site as soon as the end of the month or early June, said one of the people.
Amazon has been working to develop the new private-label lines for several years and had approached branding consultants and manufacturers including TreeHouse Foods Inc., The Wall Street Journal reported last year.
Consumers have warmed to private-label brands since the days of generically named products sold in plain white packaging. Today, retailers from Wal-Mart Stores Inc. to Sephora to Dean & DeLuca sell a range of in-house brands that some may even view as higher quality.
Store brands reached $118.4 billion in U.S. sales last year, up about $2.2 billion from the prior year, according to the Private Label Manufacturers Association.
Amazon’s latest lineup is aimed at winning sales in niches with generally higher profit margins, as well as giving the Seattle retailer a potential edge in crafting new products ahead of its own vendors.
“Amazon is ‘carpet-bombing’ the market with new products,” said Bill Bishop, chief architect of brand consultancy Brick Meets Click. “Private label allows them to test out new prices and distinctive flavors with less risk.”
Mr. Bishop said private-label goods boast higher profit margins than name brands because companies save costs on marketing and brand development. And with Amazon’s rich trove of data, it may better predict which products will sell well to its customers.
Amazon only will offer the private-label products to members of its $99-per-year Prime membership, this person said, potentially giving the program a boost.
The retailer has been spinning out new private labels for years, including its Pinzon linens and towels and Elements baby wipes.
Its AmazonBasics line features hundreds of items such as cellphone cases, computer mice, batteries, dumbbells and dog crates. Recently, it has begun selling its own fashion lines such as Lark & Ro dresses and North Eleven scarves.
Amazon’s new brands will include the Happy Belly line of food stuffs including nuts, trail mix, tea and cooking oil, the people said. The Wickedly Prime brand will feature snack foods.
While the full lineup couldn’t be ascertained, Amazon previously applied for trademark protection for foods including pasta, granola, potato chips and chocolate, as well as razors and air deodorizers.
Amazon’s plans also include the Presto! line of household products such as laundry detergent and the Mama Bear brand for baby products like diapers, baby-food jars and gentle detergent, these people said.
It wasn’t clear how Amazon might price its new food and household goods products relative to brand names.
Food production carries particular risks. For its new brands Amazon will depend on manufacturers that may have varying quality controls. Any health-related recalls could damage Amazon’s reputation.
Amazon has stumbled in private labels before. Its Elements line, which promised greater transparency about where and how goods were made, initially included diapers, but Amazon pulled them weeks after launching in late 2014, citing design flaws.
It also has discontinued a tool line and cookware endorsed by a Seattle chef.
In a twist, Amazon may graduate some of its coming products to the Elements label based on a formula including ratings and sales, suggesting it plans to make it a premium line, one of the people said.
The Elements private label has been sold only through Prime since it was released in late 2014. Despite the restriction, its Elements wipes represent about 9% of all baby-wipes sales on Amazon.com, according to analytics firm One Click Retail.
Amazon occasionally designates some products for special discounts available solely to Prime members. For instance, some customers were surprised last month to find certain videogames such as Assassin’s Creed Syndicate could only be purchased by Prime members.
By some estimates, Amazon has 50 million or more Prime members. The company covets them because they spend more on the site on average and may watch streaming videos such as its “Transparent” TV series.
In addition to bolstering its Prime service, the new lineup of private-label products may feed Amazon’s Fresh grocery delivery business, which is available in several U.S. cities.
"If I can hack my lights I can probably do the same for my neighbour's room..."
Purism CTO Zlatan Todorić: 'The Internet of Things is really horrible'
Hacking a website is so 1990s. Today, hacking Internet of Things (IoT) devices is where the cool hackers are hacking.
And, in many cases, it's almost an open door. A short while ago, while staying at a London hotel, security developer Matthew Garrett hacked the light controls in his room, then blogged about it.
He's not the only, though. "I did the same thing," said Zlatan Todorić, chief technology officer at Purism, a developer of laptops and open source software, speaking to V3.co.uk sister publication Computing at the Privacy Advantage event in London last week.
"I noticed that there was this sort of smartphone to control all the lights and I thought 'ah, this is some sort of IoT device'. If it can send and receive signals it must have a network interface, and if it has a network interface then awesome, it's exploitable."
With a little effort Todorić hacked the system and succeeded in controlling the lights in his room from his laptop.
"Then I thought, if I can hack my lights I can probably do the same for my neighbour's room and I gave him a special light show and I could hear him shouting and complaining ‘what the hell's going on?'" he laughed.
"So, then I thought let's take it to the next step. The servers must be on the same network, so I looked around and found them in the clear, unprotected. So I hacked their servers and then I went down and told them what I'd done and said 'you really need to change this'."
So far, so geeky hi-jinks, but what this shows is just how vulnerable such systems are to those with a bit of technical know-how if they are not properly secured. As "smart" devices proliferate, they open up a huge range of potential entry points for hackers.
"The Internet of Things is really horrible," Todorić said. "Everyone's excited by their toaster being smart, but it's not smart, it's stupid. There's no such thing as a smart device. [A human being] creates software that says computationally how that thing will work, and that's all it does. It's stupid.
"Don't give information to your fridge. Don't give your information to a toaster. If I hack your toaster and it's connected to your phone I'm going to hack your phone. If I hack your phone I can get to your inbox and I can create fake data about you if I want to."
In Computing's latest research, device and data security (or the lack thereof) and a need for proper data protection and privacy frameworks were found to be the chief impediments to the wider adoption of connected IoT devices
At least this was better than the comedy meeting in which his chair collapsed
On-Call Welcome again to On-Call, our Friday folly in which readers share stories of their professional adventures.
This week, we bring you the tale of reader “Kelly” who encountered a previous On-Call titled Outsourcer didn't press ON switch, so Reg reader flew 15 hours to do the job.
“I see you that story,” wrote Kelly, “and raise it with 'I was flown to Hong Kong for a week to do nowt'.”
That's got us interested, Kelly. Do tell more.
“This was back in the late 80's,” Kelly wrote. “ I was in the third line support team for a Mainframe DBMS based in the UK. We had a customer in Hong Kong who were running a business critical system based on this DBMS and I'd been in part responsible for the successful commissioning of that system the previous year when I'd flown out and helped diagnose a fault that was preventing the user acceptance test from being completed.”
“So, when they hit a hard fault that was stopping a critical job from running they fired off the distress flare and I found myself being fast tracked through an international trip authorization and was climbing onto a 747 on a Saturday morning.”
Kelly was a bit worried about the job, because “the fault they were hitting exhibited the same symptoms as one for which a fix was already available, and I'd already pointed this out and been assured that the fix in question was already in place.”
“This must be something new,” Kelly told himself after landing at the “late and not much lamented Kai Tak Airport”* and enduring “the customary wait on the plane while the ground crew presumably removed all of the laundry that had been collected during the final approach from the leading edges.”
Upon landing Kelly says “I was met by an old mate who'd been flown up from Melbourne to look into the same problem (did I mention that this customer had some clout?), who greeted me thus:
"You were right. They were loading the code from a library that didn't have the fix applied that you've been telling them about since Thursday. I spotted it about the time you were taking off yesterday. Shall we go to the Pub?" Kelly got to stay there for a week: his cheap flight could not be rescheduled “and so I spent the next week languishing in the Excelsior Hotel, never quite getting overt the jet lag, and generally doing bugger-all.”
That wasn't Kelly's only pointless work journey. In recent years he says he was flown to Hanoi from Kuala Lumpur for a meeting in which it was thought his expertise would come in handy.
It didn't. “I spent a three hour meeting on Monday morning trying to stay awake while a bunch of guys discussed a network roll-out in Vietnamese,” Kelly wrote. “My only contribution to the meeting was having my chair collapse underneath me, which added some light relief.”
Parenting retailer Kiddicare has suffered a data breach that exposed the names, addresses and telephone numbers of some of its customers.
The company said it had emailed 794,000 people who may have been affected by the incident. It said the data had been taken from a version of its website set up for testing purposes. Security researchers have warned that the details could be used by criminals to try to scam those affected. The firm said it had reported itself to the UK's Information Commissioner.
Suspicious messages UK-based Kiddicare is a baby and child specialist that trades online and from its flagship store in Peterborough. The company said it became aware of the data breach after customers reported suspicious text messages that had not been sent by Kiddicare.
It was then contacted by a security company with further information and was able to link the breach to a "test" website it had been using in November 2015.
The company has not detailed the breach on its website "Kiddicare used real customer data on its test site," said security researcher Graham Cluley in a blogpost. "It shouldn't be forgotten that this was a test site and things are expected to go wrong." The company stressed that payment details such as credit card information, which can easily be changed, had not been stolen. However, customers' names, postal addresses, email addresses and telephone numbers had been exposed and that information could be used by scammers.
Risk Mr Cluley criticised the company for neglecting to post details of the breach prominently on its website, although they have answered some questions on the subject. "There is currently no mention of the data breach on the Kiddicare website's homepage or on its Twitter account," he wrote. "I'm not sure that's offering the best service for customers who, through no fault of their own, might now be at risk. "One clear risk is that Kiddicare customers might be contacted by fraudsters pretending to be the baby specialist retailer, in an attempt to trick unsuspecting consumers into handing over payment information." The company apologised to customers in a statement sent to the BBC. "We are very sorry for the potential stress and anxiety this incident may have caused our customers," it said. "We want to reassure everyone that the problem has been fixed, increased security measures have been implemented and we have a dedicated team to here to help with any further concerns."
Uber began as a luxury product, making a black car available at the press of a button.
Now six years later, the company has decimated the taxi industry and may have a new target: public transportation.
On Monday, Uber announced that it was changing how its carpooling product, UberPool, works in New York City.
The typical UberPool ride means a driver will pick a passenger up wherever they are and drop them off at their final destination — the difference being that along the way, the driver might be picking up and dropping off other riders during a trip.
Starting Monday in a pilot phase, New York City's new version of UberPool is keeping the carpool aspect, but instead moving the pick-up and drop-off points to street corners. Instead of showing up right in front of your doorstep, the app will direct you to a corner nearby to wait for your ride.
During peak commuter hours (7-10 am and 5-8pm, Monday through Friday), Uber will only charge $5 for these UberPool rides starting and ending in Manhattan. And that price will never increase from surge pricing.
On nights and weekends — the corner pick-ups are only between 7am-8pm — it will revert back to dropping you off at your door.
It's a lot of changes, but if you add together the low-cost flat fare pricing and ushering riders to stand together on corners, the UberPool changes start to sound a lot more like a private bus line.
And it should. The same team that worked on the project also built Uber's practically-identical-to-a-bus product, UberHop. In Seattle, Manila, and Toronto, UberHop runs fixed, low-price routes for commuters.
The only difference between UberHOP and the new pilot in New York City is that New Yorkers don't have to follow a fixed route.
At $5 a ride, it's almost double the cost of a $2.75 bus fare in New York, but to some New Yorkers, the extra $2 may be worth a semi-private and mostly direct ride to work.
Facebook has launched its facial-recognition-powered photo-sharing app in the EU and Canada.
The program - Moments - was released in some countries in 2015, but withheld elsewhere because of local data privacy rules.
The company has created a different version of the software to get around these restrictions.
But it acknowledged the new edition required "a little bit of work" for users to get the most out of it.
Moment's core features are that it automatically groups together photos featuring the same friend or friends, and then makes it easy to share the pictures with them if they have installed the same app.
In the original version, the snaps are automatically tagged with people's names, because Facebook is able to match them to other photos in its wider database.
But data protection watchdogs in the EU and Canada had expressed concern their citizens would have no way to opt out of the process.
To address this, the adapted app now links together photos of similar-looking faces but requires the user to identify who they are.
Moments is not the only app to use facial recognition to sort images.
Google Photos is the most popular alternative to do so.
But the search giant has yet to extend the facility to Europe, to avoid falling foul of the Irish Data Protection Commissioner.
EU and Canada-based users will have to manually identify each of their friends
Facebook has not disclosed how many people have signed up to Moments since its release on the US's iOS and Android stores 11 months ago.
However, the company has said more than 600 million pictures have been shared via the app so far.
"Our primary purpose is to solve a problem that we know that people have, where they never get the photos that their friends take of them," the app's product manager Will Ruben told the BBC.
"We view that as a pretty different type of sharing than might happen on Facebook, where people share photos more broadly with a large group of friends or even publicly.
Moments allows its users to keep track of whom they have shared their pictures with
"Moments is closer to the type of sharing that might happen these days on Whatsapp or other [private] messaging apps - but it places the photos together into a collection."
Users decide which photos are shared with the people labelled in them, and can withdraw access at a later point. They can also use the app to turn selected photos into slideshows that can be shared to their Facebook wall and elsewhere.
The technique - which Facebook refers to as "facial clustering" - still relies on some processing being done beyond the user's handset, but Facebook said it had taken great lengths to comply with the EU and Canada's privacy rules.
Google's Photos app also identifies different friends - but the facility is not offered in Europe
"A cropped low-resolution of the photo is uploaded [to the cloud] so that your phone gets a numerical representation of that face," Mr Ruben said.
"But that number is not stored anywhere on our servers, and it is only used to compare against the other photos on your phone.
"No comparison is being done on the server."
Copies of the images are, however, stored at Facebook's data centres as soon as they are shared with someone else.
"Facebook has notified this office of the Moments app and advised us that within the EU version of the Moments app they do not control or initiate the use of any feature recognition technology," said a spokeswoman for the Irish Data Protection Commissioner. "Consideration of this development is ongoing and we will more closely look at the technical details of the app following its release."
Mr Ruben said Facebook believed the original version of Moments remained the "best version" but the new edition was still "easy to use. "You don't need to label all the faces on your phone," he said. "The idea is to share with the people closest to you, so usually it's just the top 10 people or so."
Moments allows its users to keep track of whom they have shared their pictures with.
One industry-watcher said the lack of auto-tagging might slow the app's adoption. "Any additional effort that people have to put into a service creates a barrier," said Ben Wood, from the tech consultancy CCS Insight. "The seamless experience in other markets is therefore more compelling. "However, once people invest a little time, they will see the benefits, and it could catch on, on that basis."
Facebook's use of facial recognition has, however, caused controversy in the US. Last week, a US judge refused to block a class action case in which it is claimed the technology violates Illinois's Biometric Privacy Act. The law states biology-based identifiers - including facial maps and fingerprints - cannot be collected without their owners' explicit consent. Google is also being sued over the matter.
Online retail giant Amazon has launched an online video service to rival Google's YouTube.
Amazon Video Direct will allow users to post videos and earn royalties.
Viewers can rent and buy content or watch it for free with adverts. An ad-free version is offered to Amazon Prime members - a subscription service with other benefits.
Initially, videos will be viewable in the US, Germany, Austria, Japan and the UK.
The firm already offers access to professional TV shows and films via Prime Video, a rival to Netflix.
It also streams user-generated clips about video games via Twitch.
"There are more options for distribution than ever before and with Amazon Video Direct, for the first time, there's a self-service option for video providers to get their content into a premium streaming subscription service," said Jim Freeman, Vice President of Amazon Video
"We're excited to make it even easier for content creators to find an audience, and for that audience to find great content."
The news comes barely a month after online talent firm Fullscreen launched its own subscription video platform.
Previously the firm had published content on free platforms like YouTube, Instagram and Snapchat.
YouTube offers an ad-free subscription service of its own - YouTube Red - which costs $10 per month.
There's no shortage of places to watch video online, and Amazon is already a major player.
But what it's trying to do here poses a lot of questions.
How is this new idea different from YouTube? How do you beat the world's biggest video site at its own game without, on the face of it, offering any extra incentive to post or view content?
Unless, of course, Amazon does.
It may need to woo big name content creators over to the new platform with lucrative offers. But that's an expensive proposition and one that will take a long time to come to fruition, if at all.
Vimeo, Ted, Vine, Funny or Die and other popular video sites have found success by attacking a niche. It's hard to see what Amazon Video Direct's is here.
The co-founder of SoundCloud has told Newsbeat that launching a subscription service is the streaming site's "biggest achievement".
Eric Wahlforss said creating SoundCloud Go has been a "huge step" but admitted it hasn't been easy.
He said one of the biggest challenges was getting major labels to sign up after rows over artist payment.
"It's taken many hundreds of conversations across many different companies over a couple of years."
Users will have the option to keep using the site for free or pay £9.99 a month for features including access to the full range of 125 million tracks as well as ad-free and offline listening.
Some tracks by bigger artists will only be available for preview in the free version.
"It's great for us to be able to do the next important step which is all about revenue for artists," said Eric.
In May last year, Sony pulled all its music from the site after accusing SoundCloud of not providing enough options for it to make money from the music it hosts. It took until March this year to agree a deal with Sony. "If I'm an artist and I have a lot of followers, I want to be able to make a living doing what I do and that's what we're solving here," said Eric. "That's been a long conversation."
The site also has deals with two other major labels, Warner Music and Universal, as well as hundreds of independent labels. DJ Target
With SoundCloud being a key destination for music discovery, Eric suggested these new deals mean SoundCloud Go "captures the full range of music culture" from established artists to DJs and up-and-coming talent.
But with anyone able to upload content, copyright infringement is another issue SoundCloud has had to face. "We take copyright very seriously," said Eric.
"We've given them [labels and artists] flexibility to be able to put stuff in the free or paid tier.
"We want the creators and rights holders to be in control over their content and they should be able to decide what can stay up on the platform and what can't.
"What we broadly agreed with the industry is the fact that generally something like putting up a DJ mix or remix is something that's great and addictive."
The new features are important improvements for current users, said Eric.
Despite music fans signing up to the likes of Spotify, Apple Music and Tidal, he thinks there is "room for a couple of different streaming providers".
"We're not focused on trying to get people to move from other services. We already have a very large user base [an estimated 175 million].
"We are at the very early stages of streaming.
"In 10, 15, 20 years, we're going to see everything probably moving over to streaming and SoundCloud will be one of the driving forces in that transition."
Happy Ed Balls Day! Who knew that five years ago, a Twitter fail would still be remembered.
And now Ed Balls and his family are joining in the celebrations. It all started on 28 April 2011 when Ed Balls was shadow chancellor. He was urged by his aide to look on Twitter for articles mentioning his name, but instead of doing a simple search he tweeted his own name in error.
Since then on 28 April every year Twitter rejoices in the madness of the internet gaffe and marks Ed Balls Day.
As Ed Balls Day draws to a close, his wife MP Yvette Cooper has been urging him to join in. We've insisted he bake a cake. How else would you celebrate #EdBallsday?@YvetteCooperMP
Finally after thousands of tweets Ed Balls himself responded to his Twitter fans with a photo of a cake to mark the occasion. Good grief...but how could I say No? RT @YvetteCooper We've insisted he bake a cake. How else would you celebrate #EdBallsday?@edballs/Twitter
The hashtag #EdBallsDay has been the top trend in the UK and has been used over 26,000 times.