Connect with us

Technology

Graphcore’s AI chips may not be as powerful as Nvidia’s GPUs, but may provide good bang for your buck • The Register

Voice Of EU

Published

on

In brief The latest results by benchmarking consortium MLPerf, tracking the best chips for training the most popular neural networks, are out and a new player has entered the game: Graphcore.

Each MLPerf release is pretty standard. A sprawling spreadsheet records the time various systems take to train or run a particular machine-learning model; these numbers are submitted by hardware vendors.

Nvidia and Google pretty much always lead the pack so the latest results aren’t particularly surprising. What is different this year is that Graphcore has joined for the first time. It’s a good sign; it signals their technology is maturing and that it’s willing to publicly compare itself to competitors.

Although Graphcore’s IPU-PODs weren’t as fast at training the computer vision model ResNet-50 and language model BERT as Nvidia’s A100 GPU or Google’s latest TPUs, the company’s hardware is much cheaper, so it may have a performance per dollar advantage. Google’s TPUs are only available via the cloud.

You can see the full results here, and more on Graphcore here from our sister site, The Next Platform.

Say bye-bye to Pepper, the robot

SoftBank has stopped producing its humanoid Pepper robot and is cutting jobs across its robotics unit. Pepper is instantly recognizable by its white body, complete with a head, two arms, torso, lower body on wheels and a screen. It’s about the size of a small child, and has two circular black eyes and a small smile on its face.

Launched in 2014, the machine was designed to do all sorts of tasks, such as greeting customers or showing useful information like menus or locations. But it hasn’t been popular, and SoftBank has struggled to shift the 27,000 units it made. Now, it has decided to stop making them altogether, according to Reuters, and hundreds of jobs across France, the US, and the UK have been cut.

Experiments to roll out the robot across supermarkets and offices haven’t always gone well. In 2018, a Scottish supermarket fired Pepper after it freaked shoppers out and often told them to look “in the alcohol section” for unrelated items, it was reported.

The World Health Organization’s AI ethics guideline

The WHO published a 165-page report outlining an ethics and governance framework for AI in health, this week.

It’s founded on six main principles that the org hopes “can ensure the governance of artificial intelligence for health maximizes the promise of the technology and holds all stakeholders – in the public and private sector – accountable and responsive to the healthcare workers who will rely on these technologies and the communities and individuals whose health will be affected by its use.”

Those six principles are:

  1. Protecting autonomy: Machines may automate tasks and generate results, but humans should always remain in charge of the systems and oversee all medical decisions.
  2. Promoting human safety and well-being and safety and the public interest: Make sure that the effects of computer algorithms are studied and regulated to make sure they don’t harm people.
  3. Ensuring transparency, explainability and intelligibility: The technology must be understandable to everyone that uses or is affected by it, whether it’s the developers, healthcare professionals, or patients.
  4. Fostering responsibility and accountability: Understand the limits of AI technology and where it may go wrong. Make sure that someone can be held responsible if that’s the case.
  5. Ensuring equity and inclusivity: AI should not be biased or perform less well against age, sex, gender, income, race, ethnicity, sexual orientation, and so on.
  6. Promoting tools that are responsive and sustainable: Machine learning software should be designed to as computationally efficient as possible.

Facebook upgrades its research dataset to help developers build house robots

Chores are mundane and no one in their right mind likes doing the dishes or laundry, really. Humans are going to have to keep doing them unfortunately until AI robots get nimble and smart enough to take over.

Simple tasks like picking up cups, putting them in the dishwasher, or in cupboards, might be easy for us but they’re incredibly difficult for machines. Roboticists can dream about building the perfect algorithm or neural net, but without any training data it’ll be no good.

That’s why Facebook released AI Habitat, a dataset containing multiple simulations of images modelling inside people’s houses to help developers in 2019. Now, it has upgraded that to Habitat 2.0, which contains 111 unique, 3D-rendered layouts of rooms containing 92 objects, like drawers, carpets, sofas, plants, fruit, and the like.

Future AI agents can be trained to complete a specific chore in simulation to gain enough experience before they’re tested in real-world conditions. What’s most interesting is that robots designed to tidy homes in the US will probably perform differently in other countries, where the style of homes or jobs vary.

“In the future, Habitat will seek to model living spaces in more places around the world, enabling more varied training that takes into account cultural- and region-specific layouts of furniture, types of furniture, and types of objects,” said Dhruv Batra, a Facebook research scientist.

You can read more about the dataset here. ®

Source link

Technology

How to keep a support contract • The Register

Voice Of EU

Published

on

On Call Let us take a little trip back to the days before the PC, when terminals ruled supreme, to find that the more things change the more they stay the same. Welcome to On Call.

Today’s story comes from “Keith” (not his name) and concerns the rage of a user whose expensive terminal would crash once a day, pretty much at the same time.

The terminal in question was a TAB 132/15. It was an impressive bit of kit for the time and was capable of displaying 132 characters of crisp, green text on a 15-inch CRT housed in a futuristic plastic case. Luxury for sure, unless one was the financial trader trying to use the device.

Once a day, at around 13:30, the terminal would hang. The user would have to reach behind it, power it off, wait a bit, and then fire it back up again. To placate the angry customer, a replacement was dispatched, and all was well. Until the problem started again. Another replacement was made. Another week or so went by with no complaints. And again, another call: the terminal was hanging. Same time. A few times a week.

“These terminals were in the thousand-dollar range,” Keith told us, so a monthly replacement cycle was not really an option. He even used one of the faulty units himself for a while and encountered no issues, which was odd in itself and, we reckon, planted a seed of suspicion.

As for the customer, he was raging by this point. “He was threatening to cancel our contract for his entire firm,” remembered Keith, which would hit the bottom line hard. A salesperson was sent out to see what was happening, but there was no failure.

A technician went out; again no failure. Was this a case of “Technician Syndrome”, where a problem cannot be replicated in front of service personnel? Maybe. Keith’s team were at their wit’s end while the customer had hit the end of his tether and gone beyond.

The solution to the problem was accidental. Keith was back on site, diagnosing an unrelated software issue, but could see the suspect terminal on the other side of the room. As he watched, the trader using the machine sat back for lunch, flipping through the pages of a financial newspaper. A phone call came through, and the trader slung the paper on top of the monitor, took the call, and then resumed work.

Oblivious to the newspaper.

A few minutes later there was uproar. The trader had stood and was slapping the side of the terminal, yelling all manner of not-safe-for-work oaths and casting aspersions upon the good name of Keith’s firm, the software, the programmers, and the computing industry in general. The cursing continued as the trader reached behind for the power switch, knocking the paper aside.

Keith had his solution. But was smart enough to know that a bland presentation of facts would probably not help. Instead, he arranged for his office to call the trader and tell him that a tech was on the way to help. He waited until the trader was distracted and sauntered over.

“Sure enough,” said Keith, “he said he was glad to see me but launched into a tirade again about the device’s many faults.”

He let the customer vent for a while, and surreptitiously placed the newspaper back on top over the heat vents on the terminal while pretending to examine the rear of the unit.

Now patience was needed. It wouldn’t take long – the terminal had, after all, only just recovered from its last overheating episode – and Keith encouraged the trader to unload all his woes and grievances.

The bug list was building as the screen suddenly flickered and locked up. “There! You see that?” exclaimed the user. Keith nodded and reached round the side of the terminal to cycle the power. Sure enough, it came back up.

Keith made a show of thanking the user for showing him the elusive bug and was staging a call with a co-worker, supposedly to prepare a replacement, when the terminal locked up again.

Keith wrinkled his forehead at the “mystery” before offering up an explanation.

“Ah!” he exclaimed, “Did you see how that flicker started from the top and moved to the down?”

Those familiar with the technology will know it was just following the raster pattern. The customer, on the other hand, did not.

“That is often a sign it is overheating,” said Keith, playing fast and loose with the truth, “but this office is cool?”

He pretended to be mystified until the penny dropped for the trader, who unleashed yet more expletives as he realised where he’d dropped his newspaper and snatched it away from the vents.

Feeling the volcanic heat spewing from the depths of the terminal, he turned to Keith, suddenly concerned: “Will it be OK?”

Of course it would. It had only been overheating for a short time every day. The apologies from the customer, who had “discovered” the problem, were profuse and copious. Keith excused himself, but not before rubbing a bit more salt into the wound by telling the user he needed to cancel the burn-in process of yet another expensive replacement.

As it turned out, rather than the customer cancelling the support contact, it ended up being extended.

“It was a good thing I’d let him ‘discover’ the fault,” said Keith. “If I had found it, he would have been very defensive and we still might have lost that contract.”

The minor bugs the user had reported while Keith had been waiting for the overheating to happen again were swiftly dealt with and the enhancement requests logged. Keith also reported back to his boss, who spent rather a lot of time laughing.

“It was a good day.”

Ever set the stage so the customer thinks they’re the hero of the hour? Or maybe you’ve wished all manner of unpleasantness upon your suppliers before realising the blame laid with you all along? Tell us about the time you picked up the phone with an email to On Call. ®

Source link

Continue Reading

Technology

NUIG to spend €5m on research to help address global issues

Voice Of EU

Published

on

Several key research areas have been identified by NUI Galway to work towards for 2026.

NUI Galway’s recently launched research and innovation strategy includes a €5m investment on support for its multi-disciplinary research teams as they grapple with several global issues.

The strategy, which lays out plans for the university’s next five years of research, focuses on six areas: antimicrobial resistance, decarbonisation, democracy and its future, food security, human-centred data and ocean and coastal health.

“As a public university, we have a special responsibility to direct our research toward the most pressing questions and the most difficult issues,” said to Prof Jim Livesey, VP for research and innovation at NUI Galway.

“As we look into the future, we face uncertainty about the number and nature of challenges we will face, but we know that we will rely on our research capacity as we work together to overcome them,” Livesey added.

The plan focuses on creating the conditions to intensify the quality, scale and scope of research in the university into the future. This includes identifying areas with genuine potential to achieve international recognition for NUI Galway. It also aims to continue to cultivate a supportive and diverse environment within its research community.

NUI Galway has research collaborations with 3,267 international institutions in 114 different countries. The university also has five research institutes on its Galway city campus, including the Data Science Institute, the Whitaker Institute for social change and innovation and the Ryan Institute for marine research.

Its research centres in the medtech area include Science Foundation Ireland’s Cúram and the Corrib Research Centre for Advanced Imaging and Core Lab.

The university will also continue to involve the public with its research and innovation plans through various education and outreach initiatives. It is leading the Public Patient Involvement Ignite network, which it claims, will “bring the public into the heart of research initiatives”.

Another key area identified in the strategy report is the development of partnerships with industry stakeholders. NUI Galway has spun out many successful companies in recent years, including medtechs such as AuriGen Medical, Atrian, Vetex Medical and Neurent.

According to MedTech Europe, Ireland has the highest number of medtech employees per capita in Europe along with Switzerland.

Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.

Source link

Continue Reading

Technology

France hails victory as Facebook agrees to pay newspapers for content | France

Voice Of EU

Published

on

France has hailed a victory in its long-running quest for fairer action from tech companies after Facebook reached an agreement with a group of national and regional newspapers to pay for content shared by its users.

Facebook on Thursday announced a licensing agreement with the APIG alliance of French national and regional newspapers, which includes Le Parisien and Ouest-France as well as smaller titles. It said this meant “people on Facebook will be able to continue uploading and sharing news stories freely amongst their communities, whilst also ensuring that the copyright of our publishing partners is protected”.

France had been battling for two years to protect the publishing rights and revenue of its press and news agencies against what it termed the domination of powerful tech companies that share news content or show news stories in web searches.

In 2019 France became the first EU country to enact a directive on the publishing rights of media companies and news agencies, called “neighbouring rights”, which required large tech platforms to open talks with publishers seeking remuneration for use of news content. But it has taken long negotiations to reach agreements on paying publishers for content.

No detail was given of the exact amount agreed by Facebook and the APIG.

Pierre Louette, the head of the media group Les Echos-Le Parisien, led the alliance of newspapers who negotiated as a group with Facebook. He said the agreement was “the result of an outspoken and fruitful dialogue between publishers and a leading digital platform”. He said the terms agreed would allow Facebook to implement French law “while generating significant funding” for news publishers, notably the smallest ones.

Other newspapers, such as the national daily Le Monde, have negotiated their own deals in recent months. News agencies have also negotiated separately.

After the 2019 French directive to protect publishers’ rights, a copyright spat raged for more than a year in which French media groups sought to find common ground with international tech firms. Google initially refused to comply, saying media groups already benefited by receiving millions of visits to their websites. News outlets struggling with dwindling print subscriptions complained about not receiving a cut of the millions made from ads displayed alongside news stories, particularly on Google.

But this year Google announced it had reached a draft agreement with the APIG to pay publishers for a selection of content shown in its searches.

Facebook said that besides paying for French content, it would also launch a French news service, Facebook News, in January – a follow-up to similar services in the US and UK – to “give people a dedicated space to access content from trusted and reputable news sources”.

Facebook reached deals with most of Australia’s largest media companies earlier this year. Nine Entertainment, which includes the Sydney Morning Herald and the Age, said in its annual report that it was expecting “strong growth in the short-term” from its deals with Facebook and Google.

British newspapers including the Guardian signed up last year to a programme in which Facebook pays to license articles that appear on a dedicated news section on the social media site. Separately, in July Guardian Australia struck a deal with Facebook to license news content.

Source link

Continue Reading

Trending

Subscribe To Our Newsletter

Join our mailing list to receive the latest news and updates 
directly on your inbox.

You have Successfully Subscribed!