IBM launched its Power8 lineup of superscalar symmetric multiprocessors back in August 2013 at the Hot Chips conference, and the first systems became available in August 2014. The announcement was significant because it signaled the beginning of a continuing partnership between IBM and Nvidia to develop GPU-accelerated IBM server systems, beginning with the Tesla K40 GPU.
The result was an HPC “tag-team” where IBM’s Power8 architecture, a 12-core chip with 96MB of embedded memory, would eventually go on to power Nvidia’s next-generation Pascal architecture which debuted in April 2016 at the company’s GPU Technology Conference.
NVLINK, first announced in March 2014, uses a proprietary High-Speed Signaling interconnect (NVHS) developed by Nvidia. The company says NVHS transmits data over a differential pair running at up to 20Gbps, so eight of these differential 20Gbps connections will form a 160Gbps “Sub-Link” that sends data in one direction. Two sub-links—one for each direction—will form a 320Gbps, or 40GB/s bi-directional “Link” that connects processors together in a mesh framework (GPU-to-GPU or GPU-to-CPU).
NVLINK lanes upgrade from 20Gbps to 25Gbps
IBM is projecting its Power9 servers to be available beginning in the middle of 2017, with PCWorld reporting that the new processor lineup will include support for NVLINK 2.0 technology. Each NVLINK lane will communicate at 25Gbps, up from 20Gbps in the first iteration. With eight differential lanes, this translates to a 400Gbps (50GB/s) bi-directional link between CPUs and GPUs, or about 25 percent more performance if the information is correct.
NVLINK 2.0 capable servers arriving next year
Meanwhile, Nvidia has yet to release any NVLINK 2.0-capable GPUs, but a company presentation slide in Korean language suggests that the technology will first appear in Volta GPUs which are also scheduled for release sometime next year. We were originally under the impression that the new GPU architecture would release in 2018, as per Nvidia’s roadmap. But a source hinted last month that Volta would be getting 16nm FinFET treatment and may show up in roughly the same timeframe as AMD’s HBM 2.0-powered Vega sometime in 2017. After all, it is easier for Nvidia to launch sooner if the new architecture is built on the same node as the Pascal lineup.
Still ahead of PCI-Express 4.0
Nvidia claims that PCI-Express 3.0 (32GB/s with x16 bandwidth) significantly limits a GPU’s ability to access a CPU’s memory system and is about “four to five times slower” than its proprietary standard. Even PCI-Express 4.0, releasing later in 2017, is limited to 64GB/s on a slot with x16 bandwidth.
To put this in perspective, Nvidia’s Tesla P100 Accelerator uses four 40GB/s NVLINK ports to connect clusters of GPUs and CPUs, for a total of 160GB/s of bandwidth.
With a generational NVLINK upgrade from 40GB/s to 50GB/s bi-directional links, the company could release a future Volta-based GPU with four 50GB/s NVLINK ports totaling of 200GB/s of bandwidth, well above and beyond the specifications of the new PCI-Express standard.
Notebook vendors have largely moved to on-board memory designs and are killing off DIMMs so that they can make their Intel Apollo Lake-based notebooks as slim as possible.
According to Digitimes shipments of Intel’s 14nm Apollo Lake CPUs, which feature low-power consumption, high performance and reduced sizes is apparently on the increase and with that a move to on board memory designs. The reason is that the manufactures think an anorexic look is super fashionable.
Numbers of new notebooks using LPDDR4 has also increased as vendors have continued to make efforts to minimize power consumption, improve performance, while prolonging battery life.
Acer has two new ultra-thin notebooks, the Aspire S 15 and S 17, both featuring a thickness of less than 17mm. Other vendors, including Lenovo, Asustek Computer, HP and Dell are also expected to focus on ultra-thin notebooks and 2-in-1 models for the second half of 2016, Digitimes said.
Better-than-expected demand for Samsung Electronics Co Ltd’s new Galaxy Note 7 is creating supply problems worldwide, the South Korean tech giant said, suggesting strong initial sales for the new premium smartphone.
While robust demand could help deliver another solid quarter of earnings, Samsung also risks missing out on potential sales if it cannot boost supply quickly. Rivals such as Apple Inc are poised to launch new phones which could pull customers away from Samsung if a shortage persists.
“As pre-order results for the Galaxy Note 7 have far exceeded our estimates, its release date in some markets has been adjusted,” Samsung told Reuters in a statement without commenting on where launch delays could occur.
Production problems for the curved displays for the Galaxy S6 edge phone resulted in disappointing sales last year, and some investors fear a repeat if the world’s top smartphone maker does not move quickly to meet Note 7 demand.
Samsung said it was trying to boost production at the secret locations where the Notes are made, and aimed to meet demand “as early as possible”. It gave no further details.
A person familiar with the matter told Reuters there was no production issue for the curved screens used on the Galaxy Note 7 and that the shortage would not be a long-term problem.
“The party got more visitors than Samsung expected, so they just need to put more food out,” said Nomura analyst C.W. Chung, who said the supply situation was not a major risk given that Samsung made key parts such as displays and chips in-house.
Samsung could sell as many as 15 million Galaxy Note 7 phones this year, Chung said, compared with an estimated 9 million Galaxy Note 5 phones sold last year.
The phone went on sale on Aug. 19 in countries including the United States and South Korea, where it retails for 988,900 won ($882).
Security researchers from Bitdefender have found an IoT smart electrical socket which leaks your Wi-Fi password, your email credentials and is so poorly coded that attackers can use it to hijack the device and use it for DDoS attacks. In the good old days all the power point could do was turn electrical equipment on and off.
Bitdefender didn’t reveal the device’s manufacturer but said the company is working on a fix, which will release in late Q3 2016.
Smart electrical sockets are small electrical socket extenders, which you can plug into a regular wall socket. In this case the device comes with a module that allows users to manage power consumption using predetermined limits and schedule the socket to allow usage only between certain hours.
Bitdefender said that there were several major problems with this unnamed smart socket. When users set up the product, they also need to install one of the accompanying iOS or Android apps. These apps allow the user to connect to the smart electrical socket’s built-in hotspot and configure it by entering the local Wi-Fi network credentials.
The IoT socket uses these credentials to connect to the local network, and contact the vendor servers, where it sends a configuration file that includes several device details, such as model, make, device name, firmware version, MAC address, and others
All this networking is done without encryption, in cleartext, which an attacker can easily pick-up if sniffing the local network at the right time.
Additionally, the device’s default admin username and password is easy to guess, even without reading the device documentation.
The device also comes with a built-in feature to send users email notifications when a device scheduled task executes successfully. For this feature to function properly, users must fill in their email account username and password in the device’s configuration panel. The device improperly stores these details.
Bitdefender researchers say that an attacker that knows the device’s MAC address and default password can take control over the device, rescheduling it, or access data on the user’s email account and password.
Only ten percent of people respond to and deal with desktop security alerts immediately, suggesting that the rest of the population simply ignores them or just lets it happen automatically.
You know the sort of warning. It might be Chrome telling you that something is untrusted, something like that. They are very easy to become blind to, like cookie warnings for example, and the study, which comes from Brigham Young University (BYU) and Google engineers, said that most people just ignore them.
The study, entitled More Harm Than Good? How Messages That Interrupt Can Make Us Vulnerable, suggests that this seeming neglect is down to the fact that people can only do so many things at once.
“System-generated alerts are ubiquitous in personal computing and, with the proliferation of mobile devices, daily activity. While these interruptions provide timely information, research shows they come at a high cost in terms of increased stress and decreased productivity,” the study said.
“This is due to dual-task interference, a cognitive limitation in which even simple tasks cannot be simultaneously performed without significant performance loss.”
Multitasking, then. People struggle to comprehend alerts because they are busy closing windows, stopping videos, typing or uploading at their desk or while mobile. Some 87 per cent showed the most disregard when “transferring information”.
The researchers explained things better in an interview with Phys.org. “We found that the brain can’t handle multitasking very well,” said co-author and BYU information systems professor Anthony Vance.
“Software developers categorically present these messages without any regard to what the user is doing. They interrupt us constantly and our research shows there’s a high penalty that comes by presenting these messages at random times.”
A better time to alert, according to the researchers, is at more passive times, for example while punters are waiting for a page to load or have finished watching a video.
“Waiting to display a warning when people are not busy doing something else increases their security behaviour substantially,” said Jeff Jenkins, lead author of the study.
The new offering from Oracle Data Cloud includes more than 400 million business users and one million addressable U.S. companies, the company said. Factored into that mix are proprietary insights from Oracle BlueKai, Datalogix, and AddThis as well as data from Oracle partners Bombora, Dun & Bradstreet, FullContact, Gravy Analytics, HG Data, Infogroup, PlaceIQ, and TransUnion. Predictive analytics from Leadspace are included as well.
The data derives mostly from the U.S., but some international B2B segments are available, Oracle said.
“B2B marketers can now take advantage of more than 700 enhanced Oracle B2B audience segments, as well as a robust B2B audience marketplace boasting more than 4,000 pre-built audiences from partners,” Oracle said.
The offering aims to help marketers align digital expenditures with campaign objectives and sales outreach, providing a regular flow of relevant and qualified leads from target accounts.
The move is likely in part a result of Oracle’s $1.2 billion investment in Datalogix in late 2014, said Jim Wheaton, principal and cofounder with Wheaton Group.
Datalogix is one of four major cross-vertical “data co-ops” that focus on the acquisition, integration, and ongoing management of massive quantities of data about American consumers and businesses, Wheaton said via email. The others include Abacus, I-Behavior, and Wiland.
Oracle has long targeted the B2B market with sales tools designed to help organizations identify new opportunities, and it’s been “very useful for targeting and territory planning,” said Denis Pombriant, managing principal at Beagle Research Group.
“This marketing data provides the same kind of opportunity, and I expect there might be software add-ons to sell with it,” Pombriant said. Either way, “having a massive, up-to-date data source should be a boon to many companies.”
The company is building a mobile device strategy around Windows 10 Mobile and is slowly cutting its reliance on Android, once high on the company’s list for tablets and PCs.
HP has discontinued low-cost Android tablets, and two remaining enterprise tablets feature aging hardware and an old version of the OS. Company executives have said future mobile devices will be built around Windows 10 unless there’s significant new demand for Android.
HP is following the lead of Dell, which has cut Android devices to focus on Windows. Lenovo, meanwhile, still sells Android tablets and smartphones but is cutting its number of Android tablets and increasing its number of Windows 2-in-1s.
The goal for HP is simple: to unify products around one OS, much like Apple. That’s a challenge facing Samsung, with its PCs on Windows, tablets and smartphones on Android, and wearables and smart TVs on Tizen. Samsung is still working to put the pieces together to ensure all devices communicate flawlessly, but the company claimed progress during the recent launch of Galaxy Note 7.
HP is re-entering the smartphone market its Elite X3 handset, which runs Windows 10 Mobile. The company is building its smartphone strategy around Windows 10 Mobile, which had just a 0.7 percent market share in the first quarter, according to Gartner.
HP says Elite X3 can be a PC replacement with help from cloud services and accessories. Users will be able to run Universal Windows apps on PCs and smartphones. HP also plans to bring augmented reality apps on HoloLens to the Elite X3.
“We’re not trying to hit the volumes and scales of Android,” Park said. “We’re going after IT shops. There are a lot of people in the commercial domain who are not using Pokemon Go.”
HP has said it doesn’t want to sell low-cost devices and has cut many Android devices in the process. But the same strategy doesn’t apply to Windows — this week it announced low-cost Stream notebooks running Windows 10 starting at US$199.
Payments will be backed by a customer’s credit or debit card, the company said.
CVS Pay is currently available in New York, New Jersey, Pennsylvania and Delaware; a nationwide rollout at all 9,600 stores is expected to kick off later this year.
CVS doesn’t support Apple Pay or other NFC-based payment technologies, and its use of barcodes for payments is reminiscent of the way Starbucks customers pay for coffee. Working with the barcode technology was a faster way for CVS to bring forward technology for more convenient in-store payments, analysts said.
Other retailers have created in-store payments through their own apps. Walmart created Walmart Pay in December to allow payments through mobile device QR codes that can be read at checkout registers.
“There’s nothing really innovative here with CVS Pay,” said Gartner analyst Avivah Litan on Friday. “They are pretty much following the trend. It’s just mobile commerce with a credit card attached. It’s no big deal to put a credit card in a wallet.”
At one point, CVS was working with Walmart and dozens of other major retailers in the Merchant Customer Exchange, which was designed to process mobile payments electronically through bank accounts and not credit cards to cut out the card processing cost that merchants paid to banks. But MCX ended its pilot of its mobile app, CurrentC, in June. Analysts have predicted the concept will not continue.
Alibaba’s total revenue rose to 32.15 billion yuan, or $4.84 billion, in the quarter ended June 30 from 20.25 billion yuan a year earlier.
Mobile revenue increased 119.3 percent to 17.51 billion yuan, while monthly mobile active users increased 39 percent.
Net income attributable to shareholders fell to 7.14 billion yuan, or 2.94 yuan per share, from 30.82 billion yuan, or 11.92 yuan per share, in the year-earlier quarter.
Alibaba’s gross merchandise volume (GMV), the value of transactions carried out by third-party sellers on the company’s platforms, rose 24.4 percent to 837 billion yuan.
Founded in 2014, California-based Nervana offers a hosted platform for deep learning that’s optimized “from algorithms down to silicon” to solve machine-learning problems, the startup says.
Businesses can use its Nervana cloud service to build and deploy applications that make use of deep learning, a branch of AI used for tasks like image recognition and uncovering patterns in large amounts of data.
Also of interest to Intel, Nervana is developing a specialty processor, known as an ASIC, that’s custom built for deep learning.
Financial terms of the deal were not disclosed, but one estimate put the value above $350 million.
“We will apply Nervana’s software expertise to further optimize the Intel Math Kernel Library and its integration into industry standard frameworks,” Diane Bryant, head of Intel’s Data Center Group, said in a blog post. Nervana’s expertise “will advance Intel’s AI portfolio and enhance the deep-learning performance and TCO of our Intel Xeon and Intel Xeon Phi processors.”
Though Intel also acquired AI firm Saffron late last year, the Nervana acquisition “clearly defines the start of Intel’s AI portfolio,” said Paul Teich, principal analyst with Tirias Research.
“Intel has been chasing high-performance computing very effectively, but their hardware-design teams missed the convolutional neural network transition a few years ago,” Teich said. CNNs are what’s fueling the current surge in artificial intelligence, deep learning and machine learning.
As part of Intel, Nervana will continue to operate out of its San Diego headquarters, cofounder and CEO Naveen Rao said in a blog post.
The startup’s 48-person team will join Intel’s Data Center Group after the deal’s close, which is expected “very soon,” Intel said.
Troubled mobile phone maker BlackBerry has decided to make a bit more money by suing those it thinks stole its ideas.
A patent lawsuit has been launched against internet telephony outfit Avaya. However in making its case that Avaya should pay royalties, BlackBerry appears to be looking at what it has done rather than what it is doing. The firm argues that it should be paid for its history of innovation going back nearly 20 years.
The court papers say:
“BlackBerry revolutionised the mobile industry. BlackBerry… has invented a broad array of new technologies that cover everything from enhanced security and cryptographic techniques, to mobile device user interfaces, to communication servers, and many other areas.”
BlackBerry claims Avaya infringes eight US Patents:
Nos. 9,143,801 and 8,964,849, relating to “significance maps” for coding video data;
No. 8,116,739, describing methods of displaying messages;
No. 8,886,212, describing tracking location of mobile devices;
No. 8,688,439, relating to speech decoding and compression;
No. 7,440,561, describing integrating wireless phones into a PBX network;
No. 8,554,218, describing call routing methods; and
No. 7,372,961, a method of generating a cryptographic public key.
The oldest is 1998 and the most recent is 2011..
Products targeted by Blackberry include Avaya’s video conferencing systems, Avaya Communicator for iPad, a product that connects mobile users to IP Office systems, and various IP desk phones. .
The BlackBerry complaint states that the company notified Avaya of its alleged infringement of those specific patents in a letter dated December 17, 2015, which must have come as a bit of a surprise. It has been filed in the Northern District of Texas, which is less because the region is more patent friendly (like East Texas) but because it is where Avaya does business and maintains a two-story office.
BlackBerry has hired top patent lawyer Quinn Emanuel. The firm defended Samsung in the high-profile Apple v. Samsung case and has taken on various cases for Google.
Last year Cisco paid a “license fee” to Blackberry. Details were few and far between but it seems to have been to make Blackberry lawyers go away. In May, BlackBerry CEO John Chen told investors on an earnings call that he was in “patent licensing mode,” eager to monetize his company’s 38,000 patents.
Google has purchased Orbitera, a startup that aims to make it easier for software vendors to sell cloud-based products to enterprise customers. The startup gives software vendors a suite of tools for deploying and managing cloud applications and for billing businesses that use them.
Orbitera supports deploying applications on Amazon Web Services and Microsoft Azure, not Google Cloud Platform. Google said it will continue to support software deployments on platforms other than its own. That’s similar to its approach to Stackdriver, a cloud monitoring tool that works with GCP and AWS.
One of Orbitera’s key features is a service that lets companies try out enterprise software using the cloud. Vendors can set up profiles for proof-of-concept environments that are then automatically deployed on the cloud platform of their choice when a user requests a trial.
Google’s announcement explicitly states that it’s aiming to support businesses with multi-cloud deployments. That’s a somewhat different approach to Amazon and Microsoft. While they both support multi-cloud deployments, their marketing is focused on getting customers to standardize on their own services.
DC analyst Al Hilwa said the move is a way for Google to build credibility with enterprises.
“Google is aggressively building its enterprise credibility in the cloud, so perhaps they believe that this will allow them to build on the ecosystem side, which is always a great area to increase community engagement and adoption,” he said.
Because Orbitera doesn’t require customers to use GCP, Hilwa said it’s possible the deal might not drive any new growth for Google’s cloud.
The deal also seems to mesh nicely with Google’s acquisition of Bebop, the startup helmed by Diane Greene before she became head of the company’s cloud division. Bebop was building tools to help businesses more easily build cloud apps. If Google is able to fuse the two, it could help companies build cloud apps and then turn them into commercial products.
Nissan Motor Co is holding discussions with Panasonic Corp and overseas companies including Chinese firms over the possible sale of its controlling stake in a car battery manufacturing venture, sources said.
Two people with knowledge of the matter said that the Japanese automaker wants to sell its 51 percent stake in Automotive Energy Supply Corporation, which makes lithium-ion batteries for electric vehicles. The company is jointly owned by NEC Corp.
The Nikkei daily reported that Nissan was looking to sell the company because it would be cheaper to buy batteries for its electric vehicles including its Leaf model from other makers. The newspaper did not say where it obtained the information.
Talk of the sale “is speculation, and is not based on any announcement by us”, Nissan said in an email. Spokesmen for Panasonic and NEC declined to comment.
Competition to supply batteries for electric vehicles is heating up due to expectations that a growing number of lower emission cars will be produced in the coming years.
Tesla Motors, which currently procures batteries for its electric vehicles from Panasonic, is planning to boost its total vehicle production to 500,000 in 2018 – two years earlier than its original target.
Nissan and Renault SA, under Carlos Ghosn, who heads both companies, have bet more heavily on electric cars than mainstream competitors. In 2009 the two companies pledged to invest 4 billion euros ($4.43 billion) to build models including the Nissan Leaf compact and as many as 500,000 batteries per year to power them.
Sales of the Leaf and those of other electrical vehicles, however, have been disappointing, meaning Nissan and NEC have been unable to lower battery costs through mass production.
Reuters reported in 2014 that Ghosn was preparing to cut battery production by AESC and instead use packs made by LG Chem.
Nissan is also in the process of selling its 41 percent stake in auto parts supplier Calsonic Kansei Corp, sources have told Reuters.
In May, the automaker agreed to buy a 34-percent stake in Mitsubishi Motors Corp for about $2.2 billion as it seeks to better compete with bigger rivals such as Toyota Motor Corp and Volkswagen.
Security firm Malwarebytes sponsored the study, which found in June that 41 percent of U.S. businesses had at least encountered between one to five ransomware attacks in the previous 12 months.
Another 6 percent saw six or more attacks.
The study surveyed corporations in the U.S., Canada, U.K. and Germany to gauge how ransomware affected their operations.
The malware, which can infect a computer and take the data hostage, can be bad for business. Thirty-four percent of the victim corporations in the countries surveyed reported losing revenue because the ransomware had prevented access to important files.
U.S. businesses victimized by the malware generally didn’t suffer a heavy toll and only 6 percent of them reported losing revenue. In most cases, the malicious code only affected personal files.
The survey also looked at how the ransomware was affecting these enterprises, and found that generally the malware had been designed to affect desktop PCs or laptops. The infection often came through links and attachments inside emails, or from a website or web application.
The response of companies to the threat varied across countries. In the U.S, only 3 percent of the businesses hit by the ransomware decided to pay the hackers.
That’s a big difference from the Canadian businesses surveyed, of which 75 percent said they agreed to pay the ransom.
The survey said this was probably because the ransomware attacks in the U.S. often target lower-level employees and tend to only infect a few computers.
More amateur cybercriminals are probably indiscriminately spreading ransomware in the U.S. like spam, the survey added. Low-level ransom demands of up to $500 are prevalent in the U.S. However, high ransom demands of more than $10,000 are more common in Germany.
Malwarebytes sponsored Osterman Research to conduct the study by surveying 540 CIOs, CISOs and IT directors across the four countries.
Microsoft has opened up sales of the HoloLens development kit to anyone with $3,000 to spare, a move designed to significantly expand the community of developers building apps for the augmented reality headset.
Until now, HoloLens was available only to developers and companies through Microsoft sales reps, but as of Tuesday, anyone in the U.S. or Canada can purchase up to five headsets online through the Microsoft Store. There was no word about availability in other countries.
HoloLens can overlay three-dimensional images onto the real world. By tracking the user’s movements, the 3D images can be continually redrawn to match the user’s perspective, making them appear like objects in real space.
One such application was demonstrated at last month’s Worldwide Developer Conference in Toronto. Japan Airlines showed a HoloLens app that lets engineers get up close with a full scale, computerized model of a jet engine. The app can highlight engine parts and show how different components work together in a way not possible with a real engine.
he HoloLens now on sale is the same developer edition that has been offered to Microsoft partners, and buyers are asked to acknowledge before completing purchase that they understand it’s not a finished product intended for consumers.
Microsoft also asks buyers to agree not to resell the product and acknowledge that no refunds are available.
Difficulty in obtaining the HoloLens has led to a healthy markup on the device on Ebay. On Tuesday afternoon, units were selling for up to $5,000.
In June, Microsoft said it was opening up the HoloLens platform, allowing other device makers to build their own versions of the product.