Subscribe to:

Subscribe to :: ::

FDA Approves Digital Drug Tracking System For Meds

November 15, 2017 by  
Filed under Around The Net

Soon making sure medication is taken correctly will be easier to track.

The US Food and Drug Administration has approved the first drug in the US with a digital ingestion tracking system.

Abilify MyCite, an aripiprazole tablet embedded with an ingestible sensor, uses digital tracking to record whether the medication was taken. The tablet has been approved for the treatment of schizophrenia, acute treatment of manic and mixed episodes associated with bipolar I disorder, and for use as an add-on treatment for depression in adults, the FDA said.

The pill’s sensor sends a message to a wearable patch that transmits the information to an app, allowing patients to track the medication’s ingestion on their phone. Patients can also let their doctor or carer view the information through a portal online.

 Abilify MyCite’s sensor has been around since 2012, developed by Proteus Digital Health. In 2016, British Airways got in on the digital drug game, patenting a sensor-packed smart pill that measures your temperature, stomach acidity and more to help fight jet lag.

Is SAP Losing Steam

October 23, 2017 by  
Filed under Around The Net

The maker of expensive management software, which no-one really knows what it does, SAP has seen its profits take a dip.

The outfit missed market expectations for third quarter profit as it invested heavily to shift business customers into cloud computing.

SAP said it is in the middle of a transition to offering cloud-based services to its business customers and management had flagged that 2017 would see a trough in profit margins as it invested in datacenters and redeployed staff.

The outfit said it should see a recovery next year and had a “very good shot” at stabilizing margins in the fourth quarter. Chief Financial Officer Luka Mucic told a conference call:  “Going into 2018 we see a margin turnaround.”

Revenue for the German business planning software provider grew eight percent to 5.59 billion euro from a year earlier, falling short of the mean forecast of 5.71 billion euro from 16 analysts surveyed by Reuters.

Core profit excluding special items rose by four percent to 1.64 billion euro at constant currency rates, SAP said, below the 1.69 billion euro expected by analysts.

The euro’s strength sliced four percentage points off core profits, which was flat after taking currency moves into account. Analysts at Baader Helvea said they expected currency headwinds to continue for the next three quarters.

The company nudged up guidance for the full year core operating profits to 6.85-7.0 billion euro and said 2017 total revenue would range from 23.4-23.8 billion euro, marking year-to-year growth of around six to eight percent, excluding currency effects.

Cloud subscriptions and support revenue rose 27 percent in the third quarter to 938 million euro, excluding currency effects, compared with the 29 percent analysts had expected, on average.

This was offset by its classic software license and support business revenue, which rose four percent to 3.72 billion euro, slightly above the 2.2 percent growth rate expected by analysts.

Chief Executive Bill McDermott was bullish for the fourth quarter: “We are gaining share against our competitors. SAP is growing faster in the cloud – and we are doing it organically.” During a conference call, he contrasted his company with the the acquisition-fueled growth of its rivals.


Chrome 63 To Start Warning Users Of Man In The Middle Attacks

September 20, 2017 by  
Filed under Around The Net

Google Chrome 63 will warn users when third-party software is performing a Man-in-the-Middle (MitM) attack that hijacks their internet connection. 

A MitM attack happens when a communication between two systems is intercepted by a malicious actor through an application installed on a user’s computer, enabling them to send, alter and receive data meant for someone else.

It isn’t an easy attack to perpetrate, mainly because many MitM toolkits fail to correctly rewrite the user’s encrypted connections, causing SSL errors that Chrome can detect.

In Chrome 63, Google is introducing a new warning screen whenever the browser detects a large number of SSL connection errors within a short timeframe. This is a signal that an attacker is attempting to intercept the user’s web traffic, albeit with no success.

Errors can come from applications such as anti-virus software and firewalls, as well as from malware. But Chrome will filter the warning sign to only show up for software that has failed to rewrite SSL connections properly.

Chrome 63 is scheduled to be released on 5 December, according to the Chromium Development Calendar, and users can preview it through the Google Chrome development branch known as Google Canary.

The new security feature was developed by Sasha Perigo, a Stanford University student who interned at Google, working with the team responsible for Chrome. It isn’t enabled by default in Google Canary, but can be turned on manually.

Last month, Google revealed that it had developed a tool that lets users permanently mute websites that automatically play videos with sound.

The feature is currently only available in Google Canary as Google’s developers are still experimenting with it, but the company is likely to introduce the tool to Chrome users in the coming months. 

The features come as the so-called ‘browser wars’ start to hot up, once again, with Vivaldi offering a feature-packed alternative to Chrome and Opera, and Microsoft seeking to entice Windows 10 users to Edge.


Did Russian Hackers Breach U.S. And European Power Grids

September 15, 2017 by  
Filed under Around The Net

Symantec has claimed that Russian-linked hackers have targeted and successfully penetrated a power grid networks in the US and Europe.

The attacks bear the hallmarks of a hacking group that Symantec calls Dragonfly, which the company believes is a front for a state-led hacking operation. The company implied – but didn’t explicitly state – that Dragonfly is connected with Russia.

“The Dragonfly group appears to be interested in both learning how energy facilities operate and also gaining access to operational systems themselves, to the extent that the group now potentially has the ability to sabotage or gain control of these systems should it decide to do so,” claimed Symantec.

Symantec issued a research note on Dragonfly in June 2014, claiming that they had “managed to compromise a number of strategically important organisations for spying purposes and, if they had used the sabotage capabilities open to them, could have caused damage or disruption to energy supplies in affected countries”.

The company suggested that Dragonfly was targeting energy grid operators, major electricity companies, oil pipeline operators and industrial equipment providers to the energy industry. The majority of the victims were located in the United States, Spain, France, Italy, Germany, Turkey, and Poland, it added.

Symantec claims that Dragonfly activity died down after it had been exposed in 2014, but restarted in December 2015, ratcheting up from around April this year.

“As it did in its prior campaign between 2011 and 2014, Dragonfly 2.0 uses a variety of infection vectors in an effort to gain access to a victim’s network, including malicious emails, watering hole attacks, and Trojanised software,” claimed Symantec in its latest report.

“The earliest activity identified by Symantec in this renewed campaign was a malicious email campaign that sent emails disguised as an invitation to a New Year’s Eve party to targets in the energy sector in December 2015.”

The group conducted further malicious email phishing campaigns during 2016 and 2017. “The emails contained very specific content related to the energy sector, as well as some related to general business concerns. Once opened, the attached malicious document would attempt to leak victims’ network credentials to a server outside of the targeted organisation,” it added.

Intriguingly, perhaps, the Dragonfly group was observed attempting to subvert legitimate software in order to deliver malware to victims – a tactic deployed in June’s NotPetya malware outbreak in which the software update servers of a Ukrainian accounting software company were compromised to deliver a Trojanised software update.

That attack had also been linked with the Russian state, with the malware absorbing some of the leaked US National Security Agency (NSA) exploits before the Shadow Brokers group, which claimed responsibility for cracking the server on which they had been hosted, had publicly released them.

The group is using the evasion framework Shellter to develop Trojanised applications, Symantec added.

y enhete dmeidt ctiov dao wenclnoiavdn oacn outp dgantier efeonri gtnhee ilra iFcloass hg npilsauy eyrb. sSphaohrrtelpy— sakfrtoewrt evni stietgirnagt soptencoi fsirco oUdRkLcsa,b as ufoiilcei lnaamm eldl a”tisnnsit aoltl _dfelsaus he_bp lyaayme rs.eetxaed”p uw ahss asleFe n] eobno dvAi[c tsiam gcnoimdpaurteeurqss,a mf oslelloiwfe dt ashhto rttsleyg gbuys tohte eTcrnoejdainv.eK asraahg aonsyl.aB cbeatcnkadmoyoSr”.

nsit aoltl maedhdti twioolnlaal dtnoao lsss eicfc an eecteosmsearr ym.e hGto oedvoirg, oKta rsargeatnuyp.mBo,c amnidt cDiovr sohtenlo asrreo oedxkacmapbl eosw to fr ob aecnkod ololrast sunsie dl,l iawl osnrge kwciattht aT reohjta n,.yHlelraicpilpoyrT.””

While cyber attacks on infrastructure can be perpetrated with the intention of sabotage, the latest Dragonfly campaign appears to be reconnaissance, claimed Symantec.

“While Symantec cannot definitively determine Dragonfly’s origins, this is clearly an accomplished attack group.

“It is capable of compromising targeted organizations through a variety of methods; can steal credentials to traverse targeted networks; and has a range of malware tools available to it, some of which appear to have been custom developed. Dragonfly is a highly focused group, carrying out targeted attacks on energy sector targets since at least 2011, with a renewed ramping up of activity observed in the last year.”

y enhete dmeidt ctiov dao wenclnoiavdn oacn outp dgantier efeonri gtnhee ilra iFcloass hg npilsauy eyrb. sSphaohrrtelpy— sakfrtoewrt evni stietgirnagt soptencoi fsirco oUdRkLcsa,b as ufoiilcei lnaamm eldl a”tisnnsit aoltl _dfelsaus he_bp lyaayme rs.eetxaed”p uw ahss asleFe n] eobno dvAi[c tsiam gcnoimdpaurteeurqss,a mf oslelloiwfe dt ashhto rttsleyg gbuys tohte

nsit aoltl maedhdti twioolnlaal dtnoao lsss eicfc an eecteosmsearr ym.e hGto oedvoirg, oKta rsargeatnuyp.mBo,c amnidt cDiovr sohtenlo asrreo oedxkacmapbl eosw to fr ob aecnko

While cyber attacks on infrastructure can be perpetrated with the intention of sabotage, the latest Dragonfly campaign appears to be reconnaissance, claimed Symantec.

“While Symantec cannot definitively determine Dragonfly’s origins, this is clearly an accomplished attack group.

“It is capable of compromising targeted organizations through a variety of methods; can steal credentials to traverse targeted networks; and has a range of malware tools available to it, some of which appear to have been custom developed. Dragonfly is a highly focused group, carrying out targeted attacks on energy sector targets since at least 2011, with a renewed ramping up of activity observed in the last year.”


Is Data Mining Sending GPU Prices Through The Roof

September 5, 2017 by  
Filed under Computing

Consumer demand for graphics cards may be undermined by price hikes arising from the GDDR memory shortage.

According to Digitimes, first-tier vendors are expected to raise their Nvidia GeForce GTX 1080/1070/1060/1050 graphics card pricing by three to 10 percent at the end of August.

From April to mid-July, the cryptocurrency mining segment mainly contributed graphics card to sales.

With demand from the segment starting to cool off since mid-July and graphics cards’ supply and pricing both stabilizing, sales from the retail channel have started picking up.

But the stable pricing may not last long because price hikes caused by memory shortages could again deter consumer demand.

Samsung and SK Hynix have cut their memory supply for the graphics card segment, August quotes for RAM used in graphics cards have risen to US$8.50, up by 30.8 per cent from US$6.50 in July.

Both memory suppliers have allocated more of their production capacities to making memories for servers and handsets, reducing output for the graphics cards segment and fueling the price rally.


Are Russian Hackers Targeting Our Nuclear Sites

July 17, 2017 by  
Filed under Around The Net

US nuclear facilities, their suppliers and manufacturing plants using phishing methods, US authorities have said.

Last week the US Department of Homeland Security and the FBI released a joint report into recent attacks, including one on Kansas-based nuclear power station operator Wolf Creek Nuclear Operating Corporation. The report was obtained by the New York Times.

The networks of Wolf Creek and other key infrastructure companies were said to have been infiltrated. The attackers appeared to be on a reconnaissance mission, seeking to understand the workings of the networks, possibly laying the groundwork for a future assault.

The authorities blamed an “advanced persistent threat” actor for the activity, which is usually taken to mean a state-sponsored group.

However, quoting unnamed sources, the NYT says the methodology deployed by the attackers is similar to the modus operandi of the Russian group “Energetic Bear” which has been blamed for hacking energy facilities and other key targets including financial institutions since 2012.

In the recent wave of attacks, which began in May, the attackers deployed spear-phishing techniques, emailing fake CVs with a malware payload to senior control engineers authorized to access the industrial control systems. The malware was designed to harvest user credentials and passwords, the report says. Other techniques involved man-in-the-middle and watering hole attacks that are using compromised legitimate websites known to be visited frequently by the targets.

While the joint DHS-FBI report carries an ‘amber’ threat warning, the industry appears to be downplaying the seriousness of the hackers’ activities.

Nuclear Energy Institute spokesperson John Keeley said that nuclear facilities are required by law to report cyberattacks but that none of the 100 or so facilities covered by the Institute have said that their security was compromised.

Meanwhile, in a joint statement with the FBI, a spokesman for the Department of Homeland Security said, “There is no indication of a threat to public safety, as any potential impact appears to be limited to administrative and business networks.”

The US Department of Energy also said the impact appears limited to administrative and business networks.

“Regardless of whether malicious actors attempt to exploit business networks or operational systems, we take any reports of malicious cyber activity potentially targeting our nation’s energy infrastructure seriously and respond accordingly,” a spokesperson told Bloomberg.


Can Quantum Entanglement Stop Hacking

June 29, 2017 by  
Filed under Around The Net

A Chinese satellite has split pairs of “entangled photons” and transmitted them to separate ground stations 745 miles (1,200 kilometers) apart, smashing the previous distance record for such a feat and opening new possibilities in quantum communication.

In quantum physics, when particles interact with each other in certain ways they become “entangled.” This essentially means they remain connected even when separated by large distances, so that an action performed on one affects the other.

In a new study published online today (June 15) in the journal Science, researchers report the successful distribution of entangled photon pairs to two locations on Earth separated by 747.5 miles (1,203 km).

Quantum entanglement has interesting applications for testing the fundamental laws of physics, but also for creating exceptionally secure communication systems, scientists have said. That’s because quantum mechanics states that measuring a quantum system inevitably disturbs it, so any attempt to eavesdrop is impossible to hide.

But, it’s hard to distribute entangled particles — normally photons — over large distances. When traveling through air or over fiber-optic cables, the environment interferes with the particles, so with greater distances, the signal decays and becomes too weak to be useful.

In 2003, Pan Jianwei, a professor of quantum physics at the University of Science and Technology of China, started work on a satellite-based system designed to beam entangled photon pairs down to ground stations. The idea was that because most of the particle’s journey would be through the vacuum of space, this system would introduce considerably less environmental interference.

“Many people then thought it [was] a crazy idea, because it was very challenging already doing the sophisticated quantum-optics experiments inside a well-shielded optical table,” Pan told Live Science. “So how can you do similar experiments at thousand-kilometers distance scale and with the optical elements vibrating and moving at a speed of 8 kilometers per second [5 miles per second]?”

In the new study, researchers used China’s Micius satellite, which was launched last year, to transmit the entangled photon pairs. The satellite features an ultrabright entangled photon source and a high-precision acquiring, pointing and tracking (APT) system that uses beacon lasers on the satellite and at three ground stations to line up the transmitter and receivers.

Once the photons reached the ground stations, the scientists carried out tests and confirmed that the particles were still entangled despite having traveled between 994 miles and 1,490 miles (1,600 and 2,400 km), depending on what stage of its orbit the satellite was positioned at.

Only the lowest 6 miles (10 km) of Earth’s atmosphere are thick enough to cause significant interference with the photons, the scientists said. This means the overall efficiency of their link was vastly higher than previous methods for distributing entangled photons via fiber-optic cables, according to the scientists.

“We have already achieved a two-photon entanglement distribution efficiency a trillion times more efficient than using the best telecommunication fibers,” Pan said. “We have done something that was absolutely impossible without the satellite.”

Apart from carrying out experiments, one of the potential uses for this kind of system is for “quantum key distribution,” in which quantum communication systems are used to share an encryption key between two parties that is impossible to intercept without alerting the users. When combined with the correct encryption algorithm, this system is uncrackable even if encrypted messages are sent over normal communication channels, experts have said.

Artur Ekert, a professor of quantum physics at the University of Oxford in the United Kingdom, was the first to describe how entangled photons could be used to transmit an encryption key.

“The Chinese experiment is quite a remarkable technological achievement,” Ekert told Live Science. “When I proposed the entangled-based quantum key distribution back in 1991 when I was a student in Oxford, I did not expect it to be elevated to such heights!”

The current satellite is not quite ready for use in practical quantum communication systems, though, according to Pan. For one, its relatively low orbit means each ground station has coverage for only about 5 minutes each day, and the wavelength of photons used means it can only operate at night, he said.

Boosting coverage times and areas will mean launching new satellites with higher orbits, Pan said, but this will require bigger telescopes, more precise tracking and higher link efficiency. Daytime operation will require the use of photons in the telecommunications wavelengths, he added.

But while developing future quantum communication networks will require considerable work, Thomas Jennewein, an associate professor at the University of Waterloo’s Institute for Quantum Computing in Canada, said Pan’s group has demonstrated one of the key building blocks.


Anthem Agrees To Pay $115M Over Data Breach

June 27, 2017 by  
Filed under Around The Net

Anthem, the behemoth health insurance company, has agreed to settle a class action lawsuit over a 2015 data breach for a record $115 million.

The settlement still has to be approved by US District Court Judge Lucy Koh, who is scheduled to hear the case on August 17 in San Jose, California. And Anthem isn’t admitting any wrongdoing or that “any individuals were harmed as a result of the cyberattack.”

“Nevertheless, we are pleased to be putting this litigation behind us, and to be providing additional substantial benefits to individuals whose data was or may have been involved in the cyberattack and who will now be members of the settlement class,” an Anthem spokeswoman said in a statement confirming the settlement.

Assuming it’s approved, it would be the largest data breach settlement in history, according to the plaintiffs’ lawyers, who first announced the agreement Friday.

The funds would be used to provide victims of the data breach at least two years of credit monitoring and to reimburse customers for breach-related expenses. The settlement would also guarantee a certain level of funding for “information security to implement or maintain numerous specific changes to its data security systems, including encryption of certain information and archiving sensitive data with strict access controls,” the plaintiff attorneys said.

The 2015 breach resulted in the exposure and theft of nearly 80 million records, including client names, dates of birth, physical and email addresses, medical IDs and Social Security numbers. Using a stolen password, hackers were able to break into a database that contained information of former and current customers.

Although a mammoth breach at the time, the Anthem hack doesn’t compare in scale to breaches Yahoo has since reported. One of them, which occurred in 2014 and was revealed in September, affected 500 million user accounts. Then three months later, the company disclosed an even bigger breach that happened in 2013 and affected a billion user accounts.

Yahoo is facing its own data breach-related lawsuits. But for now, Anthem’s appears to be the most costly to date for a US company in terms of litigation payouts. In May, for example, Target agreed to pay $18.5 million to 47 states to settle claims stemming from a 2013 breach of credit card data. And Home Depot agreed to pay $19.5 million last year to settle a breach-related class action suit.

Ybrain Headband Treats Depression

June 20, 2017 by  
Filed under Consumer Electronics

The 21st century is increasingly acknowledging that depression as a serious illness, and now a team from South Korea wants to treat it with 21st century technology.

South Korean startup Ybrain’s Mindd headband sends weak electronic currents to the frontal lobe of your brain, reports The Korea Herald. The process, neuroscientifically referred to as transcranial direct current stimulation (tDCS), aims to stimulate the frontal lobe — where decreased activity is associated with depression.

Ybrain has received over $9 million in funding over the past four years, and says clinical trials have shown the Mindd headband to be effective and yield no side effects. It’s been used in 12 hospitals throughout South Korea already, Ybrain CEO Lee Ki-won said to the Herald, and the company hopes to bring the headband to Europe later this year and to the US in 2019.

Depression and suicide are a major problem in South Korea, with nearly 38 people killing themselves every day in the country in 2015, according to South Korea’s National Statistical Office. That’s about a third of the daily suicides in the US in the same year, a country which has roughly six-times the population.

The Mindd headband works alongside a phone app, in which patients can log their sleep, exercise and medical treatment.. All of that information will be sent to doctors for monitoring.

Ambulance Drones Shows Promise In Race To Save Lives

June 16, 2017 by  
Filed under Around The Net

The use of drones to reach patients could aid in slashing one crucial aspect of emergency care: response times.

A drone unveiled in 2014 offered the possibility of providing much quicker care to heart attack victims by coming equipped with a defibrillator — and now, the concept has passed the first test.

By staging a series of simulated cardiac arrests around the city of Stockholm, Sweden, a team of researchers led by Andreas Claesson of the Karolinska Institutet demonstrated that drones can cut response time by a median of 16 minutes. Their research has been published in the Journal of the American Medical Association.

“Some 800,000 people suffer a cardiac arrest in the EU every year, and only 8 percent survive,” said Alec Momont, who developed the original drone in 2014.

“The main reason for this is the relatively long response time of the emergency services (approx. 10 minutes), while brain death and fatalities occur within four to six minutes. The ambulance drone can get a defibrillator to a patient inside a 12 km2 zone within one minute. This response speed increases the chance of survival following a cardiac arrest from eight percent to 80 percent.”

The drone for Claesson’s tests was equipped with an automatic external defibrillator (AED) weighing 0.8 kilograms (1.7 pounds), a GPS, a high-definition camera and autopilot software. It was placed at a fire station, and deployed to 18 simulated cardiac arrests within a 10-kilometre (6.2-mile) radius, all at locations where real cardiac arrests had taken place between 2006 and 2014.

Whenever the remotely operated drone was deployed, the team would also send an ambulance.

The drone was faster. The median time from call to dispatch of the ambulance was three minutes, compared to three second for the drone. The median time from dispatch to arrive was 22 minutes for the ambulance, and just 5 minutes and 22 seconds for the drone. The median reduction in response time was 16 minutes and 39 seconds.

“Saving 16 minutes is likely to be clinically important. Nonetheless, further test flights, technological development, and evaluation of integration with dispatch centers and aviation administrators are needed,” the study reads. “The outcomes of out-of-hospitcal cardiac arrest using the drone-delivered AED by bystanders vs resuscitation by EMS should be studied.”

There are some limitations to the study. All the test flights, of which there were only 18, were conducted in good weather, and all were over short distances. And it didn’t take into account what happens after the drone reaches the patient, which would require a person present who is able to operate the defibrillator.

Momont’s 2014 system’s solution to this problem was to equip his drone with a livestream audio and video connection so that a trained operator could walk a bystander through the process of using the equipment.

Social Media Companies May Face New Fines, Regulations In Britain

May 15, 2017 by  
Filed under Around The Net

British Prime Minister Theresa May vowed to create new powers giving her authority to punish social media and communications companies that fail to look after users’ data, and to demand cash from firms to pay for policing the internet.

The election pledge comes after firms like Facebook and Twitter have been criticized the government for not doing enough to stop the spread of extremist content online or help victims of abuse.

May, who is expected to win a majority at the June 8 election, pledged to pass laws giving users new rights to access data held about them, and granting the government the power to enforce them with sanctions.

“The internet has brought a wealth of opportunity but also significant new risks which have evolved faster than society’s response to them,” May said in a statement.

“We want social media companies to do more to help redress the balance and will take action to make sure they do.”

Hospitals and doctors’ surgeries across England were forced to turn away patients and cancel appointments on Friday after a nationwide ‘ransomware’ cyber attack crippled some computer systems in the state-run health service.

The Conservative Party said it wanted to be able to tax the industry if it chooses to, citing similar plans already in force for the gambling industry.

“The Conservatives will also create a power in law for government to introduce an industry-wide levy from social media companies and communication service providers to support awareness and preventative activity to counter internet harms,” the party said in a statement.

Is nVidia Investing A Lot Of Money Into A.I. Start-Ups?

May 3, 2017 by  
Filed under Computing

Nvidia is telling the world before its GPU Technology conference in San Jose on May 8-May 11.

Writing in his bog, Nvidia’s vice president of business development Jeff Herbst said Nvidia will provide technical guidance, joint marketing help, strategic direction, and other aid. Nvidia made the investments through its GPU Ventures program.

The start-ups include Abeja, a Tokyo-based startup focused on AI-powered retail analytics systems; Datalogue, a New York AI data-mining platform; Optimus Ride an MIT spinoff looking at autonomous vehicles; SoundHound, which makes voice-enabled AI solutions; TempoQuest — Bouldera, which is creating GPU-accelerated weather forecasting and Zebra Medical an Israeli-based start-up using AI to read medical images.

Nvidia also made its third investment in MapD, which uses GPUs to query massive databases and is about to announce another AI startup investment.


Will The A.I. Space Heat-Up This Year?

April 13, 2017 by  
Filed under Computing

Major component manufacturers in the artificial intelligence (AI) market have all increased their efforts to develop more aggressive processors for AI-fueled markets in 2017 including autonomous vehicles, enterprise drones, medical care, smart factories, image recognition, and general neural network research and development.

Intel’s Nervana platform is a $400 million investment in AI

Back in November, Intel announced what it claims is a comprehensive AI platform for data center and compute applications called Nervana, with its focus aimed directly at taking on Nvidia’s GPU solutions for enterprise users. The platform is the result of the chipmaker’s acquisition of 48-person startup Nervana Systems back in August for $400 million that was led by former Qualcomm researcher Naveen Rao. Built using FPGA technology and designed for highly-optimized AI solutions, Intel claims Nervana will deliver up to a 100-fold reduction in the time it takes to train a deep learning model within the next three years.

The company intends to integrate Nervana technology into Xeon and Xeon Phi processor lineups. During Q1, it will test the Nervana Engine chip, codenamed ‘Lake Crest,’ and make it available to key customers later within the year. The chip will be specifically optimized for neural networks to deliver the highest performance for deep learning, with unprecedented compute density and a high-bandwidth interconnect.

For its Xeon Phi processor lineup, the company says that its next generation series codenamed “Knights Mill” is expected to deliver up to four times better performance for deep learning. Intel has also announced a preliminary version of Skylake-based Intel Xeon processors with support for AVX-512 instructions to significantly boost performance of inference tasks in machine learning workloads.

“Intel is committed to AI and is making major investments in technology and developer resources to advance AI for business and society,” said Intel CEO Brian Krzanich.

Nvidia partners with Microsoft on AI cloud computing platform

Earlier last month, Nvidia showed no signs of slowing its AI cloud computing efforts by announcing a partnership with Microsoft for a hyperscale GPU accelerator called HGX-1. The partnership includes integration with Microsoft Project Olympus, an open, modular, very flexible hyperscale cloud hardware platform that includes a universal motherboard design (1U/2U chassis), high-density storage expansion, and a broad ecosystem of compliant hardware products developed by the OCP community.

Nvidia claims that HGX-1 establishes an industry standard for cloud-based AI workloads similar to what the ATX form factor did for PC motherboards more than two decades ago. The HGX-1 is powered by eight Tesla P100 accelerators connected through NVLink and the PCI-E standard. Nvidia’s hyperscale GPU accelerator will, it claims, allow cloud service providers to easily adopt Tesla and Quadro accelerator cards to meet the surging demand for AI computing. The company plans to host another GPU technology conference in May 2017 where it is expected to unveil more updates on its AI plans.

On the consumer front, Nvidia’s Shield platform integrates with Amazon Echo, Nest and Ring to provide customers with a “connected home experience”, while Spot is its direct take on Amazon Echo and brings ambient AI assistance into the living room. For drivers, the company’s latest AI car supercomputer is called Xavier and is powered by an 8-core custom ARM64 processor, and a 512-core Volta-based GPU. The unit is designed with ASIL D safety rating, the highest classification of initial hazard, and can deliver 30 tera ops of double-precision learning in a 30W design.

Qualcomm’s acquisition of NXP signals investment in AI market

Back in October, San Diego-based Qualcomm bought NXP, the leader in high-performance, mixed-signal semiconductor electronics – and a leading solutions supplier to the automotive industry – for $47 billion. The two companies, joined into a single entity, have now represented what is considered a strong contender in automotive, IoT, and security and networking industries. Using several automotive safety sensor IPs from the acquisition, including radar microcontroller units (MCUs), anti-lock braking systems, MCU-enabled airbags, and real-time tire pressure monitoring, Qualcomm is now positioned to be a “one-stop solution” for many automotive customers.

With its Snapdragon and Adreno graphics capabilities, the company is well positioned to compete with Nvidia in the automotive market and stands a much better chance of developing its self-driving car platform with the help of NXP and Freescale IP in its product portfolios.

AMD targets AI learning workloads with Radeon Instinct accelerators

Back in December, AMD also announced its strategy to dramatically push its presence into the AI-related server computing business with the launch of new Radeon Instinct accelerator cards and MIOpen, a free, comprehensive open-source library for developing deep learning and machine intelligence applications.

The company’s Radeon Instinct series is expected to be a general-purpose solution for developing AI-based applications using deep learning frameworks. This includes autonomous vehicles, HPC, nano-robots, personal assistance, autopilot drones, telemedicine, smart home, financial services and energy, among other sectors. Some analysts note, however, that AMD is the only company uniquely positioned with both x86 and GPU processor technologies, allowing it to fulfill the role of meeting a variety of data center solutions on demand. The company has been developing what it claims is an efficient connection between both application processor types to meet the growing needs of AI’s technological demands.

The MIOpen deep learning library was expected to be released in Q1, but may have been delayed by a couple of weeks. Meanwhile, AMD’s Radeon Open Compute (ROCm) Platform lets programmers focus on training neural networks through Caffe, Torch 7, and Tensorflow, rather than wasting time doing mundane, low level, performance tuning tasks.


Is Samsung Going All-In On A.I.?

March 1, 2017 by  
Filed under Computing

Samsung may be about to write a billion dollar cheque to pick up some artificial intelligence technology.

The billion will not just be used for acquisitions, but also to invest in companies involved in AI.

This is in addition to what Samsung has already bought, including the acquisition of Viv Labs, an AI company from the team behind Apple’s Siri, plus the many references to its own AI assistant coming soon, which we currently know as Bixby.

Samsung also recently contributed to SoundCloud’s recent funding round, focusing on development of its Houndify AI platform. Joining the Catalyst program is Samsung Next, a $150 million fund for startups specialising in VR, the Internet of Things, and artificial intelligence.

We have heard all this before. Last year Samsung’s head of software research and development told Bloomberg:  “We are actively looking for M&A targets of all sorts in the software area. We are open to all possibilities, including artificial intelligence. Intelligence is no longer an option. It’s a must.”

However, this is first time we’re hearing about a cash figure linked to Samsung’s interest in AI, and it’s big enough to show that the outfit is serious. The first move will come with the Galaxy S8, which is expected to feature the Bixby, an AI assistant. Of course, the Tame Apple Press claims this is all because Samsung is envious of Apple’s super cool Siri, even if that AI is looking rather out of date now.


Researchers Show How Heartbeats Can Be Used As Passwords

January 23, 2017 by  
Filed under Around The Net

Researchers at Binghamton State University in New York believes your heart could be the vital to your personal data. By measuring the electrical activity of the heart, researchers say they can encrypt patients’ health records.

The fundamental idea is this: In the future, all patients will be outfitted with a wearable device, which will continuously collect physiological data and transmit it to the patients’ doctors. Because electrocardiogram (ECG) signals are already collected for clinical diagnosis, the system would simply reuse the data during transmission, thus reducing the cost and computational power needed to create an encryption key from scratch.

“There have been so many mature encryption techniques available, but the problem is that those encryption techniques rely on some complicated arithmetic calculations and random key generations,” said Zhanpeng Jin, a co-author of the paper “A Robust and Reusable ECG-based Authentication and Data Encryption Scheme for eHealth Systems.”

Those encryption techniques can’t be “directly applied on the energy-hungry mobile and wearable devices,” Jin added. “If you apply those kinds of encryptions on top of the mobile device, then you can burn the battery very quickly.”

But there are drawbacks. According to Jin, one of the reasons ECG encryption has not been widely adopted is because it’s generally more sensitive and vulnerable to variations than some other biometric measures. For instance, your electrical activity could change depending on factors such as physical exertion and mental state. Other more permanent factors such as age and health can also have an effect.

“ECG itself cannot be used for a biometric authentication purpose alone, but it’s a very effective way as a secondary authentication,” Jin said.

While the technology for ECG encryption is already here, its adoption will depend on patients’ willingness to don wearables and on their comfort with constantly sharing their biometrics.

Next Page »