Samsung plans to share more tech to lure more semiconductor business

Exynos8_PR_Main_1

Its devices foothold slipping in the country, Samsung is clawing back into China’s mobile market by putting more weight behind its semiconductor arm. Its foundry business is undergoing slow growth, but Samsung is expected to heat up talks with Chinese firms to change that.

The Korean tech firm is set to host what it’s calling the Samsung Foundry Forum on August 30, during which its semiconductor chiefs will lobby representatives from roughly 100 Chinese firms. The forum will welcome the likes of ZTE, MediaTek, HiSilicon and others, according toreports.

It isn’t the creation of the event that’s telling of Samsung’s intentions here, as the company regularly engages potential customers in China. It’s the effort that’s going into this forum and what’s expected to be a slight pivot in its strategy. It plans to take on more customers instead of leaning on a handful of them.

During Samsung’s upcoming forum, the firm is expected to court new clients by offering them access to its 10nm and 14nm FinFET process designs along with its 28nm fully depleted silicon on insulator processes.

Probably one of the most infamous setbacks in Samsung’s semiconductor business was whenApple began to offload some of its orders for iPhone processors to TSMC (Taiwan Semiconductor Corporation) two years ago.

Last year Apple turned backed to Samsung for production of the iPhone’s A9 chips. But the company’s semiconductor business isn’t running as optimally as Samsung’s execs would like. And Samsung would probably not have to rely so heavily on its frenemy, Apple.

Samsung’s foundries are still growing, as evidenced by its semiconductor group’s 8 percent-quarter over-quarter growth and 6 percent year-over-year growth in operating profit last quarter. It’s just that its foundries’ older lines are stagnating because they aren’t suited for the production of newer chips its current clients want.

 

 

[Source: Androidauthority]

Titanfall 2’s first multiplayer tech test gave Respawn’s servers a workout

Titanfall 2

Respawn Entertainment has posted up some notes on how last weekend’s multiplayer tech test for its upcoming sci-fi shooter Titanfall 2 held up on the server and backend side.

 

In a blog post, Respawn talked about how they ran the thousands of dedicated servers during the tech test:

We made sure to have a lot of datacenters for Titanfall 2, to give us the ability to scale very rapidly as well as to move games to another datacenter if we hit issues. We actually tested datacenters going offline completely and we learned how quickly we could move 100% of users from one datacenter into another datacenter. We weathered network attacks as well using this same system. After having done it a few times, it gives us a process to follow so that we can move players around even faster in the future. And for launch, we’ll have even more datacenters available!

The blog also talked about how the tech test was used to try out the other backend services for the game, such as the new matchmaking system and the Networks player group feature:

We had some times where we were fixing issues with the brand new Invite Network functionality, and we could actually see the impact of those in the player counts in real-time! We saw a huge amount of players who ONLY joined games using the Invite Network popup in lobbies. With the Network Invites, Happy Hour, clan tags, and Inbox functionality, it seems like people really embraced our new Networks system. This was a huge success for us, and they were something that was hardest to verify before this weekend. Networks were created for real humans to use, and no simulation can prove that it would be successful like real people can.

Xbox One and PlayStation 4 owners will be able to check out a portion of Titanfall 2 once again this weekend with the second of Respawn’s two tech tests, from August 26-28. The full version of the game is due for release on October 28.

 

 

[Source: Windowscentral]

What to do if the Characters palette won’t appear in Safari

macos emoji input

In some recent troubleshooting, I tried to get the Characters palette to appear by selecting the Input menu’s Emoji & Symbols item. This worked in every app except Safari: switching from, say, the Finder to Safari hid the palette. Choose Hide and Show from the Input menu didn’t fix it. Something was corrupted.

Sometimes, this bad boy won’t show up.

If you have this problem, follow these steps:

  1. Quit Safari.
  2. In the Finder, select Go > Go to Folder or press Command-Shift-G.
  3. Now enter exactly the following: ~/Library/Preferences/
  4. You’ll now be in the Preferences folder for the currently logged in user. Find the file named com.apple.HIToolbox.plist and move it to the Trash.
  5. Launch Safari.
  6. Select Input menu > Show Emoji & Symbols.

You should now be able to see the Characters palette floating over any app you’re using.

Ask Mac 911

We’ve compiled a list of the most commonly asked questions we get, and the answers to them: read our super FAQ to see if you’re covered. If not, we’re always looking for new problems to solve! Email yours [email protected] including screen captures as appropriate. Mac 911 cannot reply to email with troubleshooting advice nor can we publish answers to every question.

 

[Source: Macworld]

Tuning Workload Performance with Query Store

Image result for Tuning Workload Performance with Query Store

As your database grows, it is not unusual to experience performance problems caused by queries that once performed well, but now perform poorly. Without the ability to see all the changes in the query execution over time, it is difficult to figure out why regressions happen and what you can do to prevent them. Troubleshooting these performance issues can take hours or even days.

That’s exactly what Query Store, a new feature available in Azure SQL Database and SQL Server 2016, aims to improve.

Similar to an airplane’s flight data recorder, Query Store constantly collects information about all queries and greatly simplifies performance forensics by reducing the time to diagnose and resolve issues. Query Store also allows you to force query plans from the history, which makes the process of fixing problems with plan choice regressions extremely easy.

 

As a result, time to detect and mitigate performance issues is now in the range of minutes.

Typical usage scenarios

Query Store can be used in a wide set of scenarios when tracking and ensuring predictable workload performance is critical. It is equally useful when you need to react to an immediate problem as well as when you want to ensure optimal performance proactively for the long term (after the database upgrade, for other maintenance operations, or upon new application roll-out, for example).

 

 

[Source: Blogs Technet]

The advanced security techniques of criminal hackers

Image result for The advanced security techniques of criminal hackers

Staying secure online is an essential concern, for individual users, businesses, and cybercriminals alike. That’s right: Basic IT security applies whether you’re protecting sensitive data at an upstanding, ethical organization, or you’re in the business of stealing data from those same organizations.

After all, the business may be cybercrime, but cybercriminals are still operating a business, with all the associated worries. Criminals rely on operations security (opsec) to stay ahead of law enforcement and security researchers intent on dismantling their operations, but also to protect their criminal enterprises from competitors planning on sabotage.

I spoke with Rick Holland, former Forrester analyst and VP of strategy at Digital Shadows, at Black Hat about the security tools and techniques currently in use by cybercriminals. What Holland had to say was both illuminating and, when considering the business of cybercrime, not entirely unfamiliar.

“Opsec is a really good topic to be talking about in Vegas [Black Hat conference]. A lot of the things individuals are doing here, some of that applies to the bad guys,” Holland says.

Take, for instance, the fine line between increasing revenue opportunities and protecting yourself and your organization. In the legitimate business world, security measures can be an impediment to the nimbleness necessary to capitalize on some business opportunities. The same can be said in the criminal underworld. A seller on a carder forum doesn’t want to sell stolen credit card numbers to someone an undercover law enforcement officer or to someone who won’t pay. On the other hand, putting up a lot of barriers and identification checks will get in the way of making sales. Too much security can hurt the business.

“For cybercriminals, they want to make money. Opsec can be a barrier to making money, if they have too much,” Holland says. They have to find that right balance where they have enough opsec to keep themselves safe from law enforcement but free enough to be able to do business quickly.

A term from the military world, “opsec” refers to tactics used to protect privacy and anonymity. The irony is that criminals rely on many of the same tools that defenders and internet users on the good side adopt to stay secure online — tools such as Tor and VPNs for network connectivity, bulletproof hosting, and Jabber and OTR protocol for online chats, according to Holland. Snowden popularized Tails, the Linux distribution for the paranoid, Holland adds.

“The majority of criminal threat actors are using commercial/open source tools,” Holland says. There is no need to create their own tools when there are plenty to choose from.

Technology aside, criminals rely on other opsec measures such as disassociating their criminal enterprise entirely from their personal identity. Mixing up the two, even once, can result in a shutdown. For example, law enforcement authorities were able to track down Dridex botnet operator Andrey Ghinkul because he associated his nickname Smilex with his real name. Failing to keep the personal life and criminal activities separate is a common opsec mistake, Holland says.

Or thinking the VPN is on when it actually isn’t, as Sabu, Hector Xavier Monsegure from hacktivist collective Anonymous, learned to his chagrin when agents from the Federal Bureau of Investigation showed up at his door.

Defenders can capitalize on weak attacker opsec to gain insights into the people, process, and technology the attackers are using. Lapses in opsec give defenders the clue they need to stop the attack. In the new world where everything is online, any break is helpful.\

 

[Source: Javaworld]

Third Xiaomi Redmi 3S Prime, Redmi 3S flash sale today at 12 PM

Xiaomi Redmi 3S launch

If you’ve missed your chance at picking up the Xiaomi Redmi 3S or Redmi 3S Prime in the first two flash sales, you will have another opportunity to do so once again, today at 12 PM.

You only have to look back at the last two flash sales to see the popularity of these ultra-affordable smartphones from Xiaomi. In the first sale held two weeks ago, the Redmi 3S Prime sold out in just eight minutes, and in the second, which also saw the availability of the Redmi 3S, stocks lasted till only around 5 PM that evening.

If you’re wondering about which device is better suited to your needs, it is worth noting that most specifications and features are the same between the two versions, including the huge 4,100 mAh battery that is found with both. As far as differences go, the cheaper Redmi 3S comes with 16 GB of storage and 2 GB of RAM, while the Redmi 3S Prime offers 32 GB of internal storage and 3 GB of RAM, and also provides an additional layer of security with a fingerprint scanner.

The Xiaomi Redmi 3S is priced at Rs 6,999 (~$104), while the Redmi 3S Prime is only slightly more expensive, with its price tag of Rs 8,999 (~$134). The two devices will available from bothmi.com and Flipkart, with Flipkart also providing consumers with an exchange offer. So if you have an old Android smartphone lying around, you can always trade that in and pick up these affordable smartphones for even cheaper.

Let us know if you were able to pick up the Xiaomi Redmi 3S or Redmi 3S Prime during today’s sale, and if you already have the phone, do share your thoughts on what your experience has been with it so far in the comments section below!

 

 

[Source: Androidauthority]

How to use the Sonos Controller app on Windows 10

Image result for How to use the Sonos Controller app on Windows 10

Command and control your tunes right from your desktop with the Sonos app for Windows.

While you can control your music’s volume directly from any Sonos player, the Sonos app for Windows gives you full control over all aspects of your home audio. If you have more than one room set up, then you’ll definitely want to take advantage of the options the app gives you.

 

[Source: Windowscentral]

Twitter rolls out night mode in its iOS app

twitter night mode ios app

Now it’s easier for iPhone users to tweet past their bedtimes.

On Monday, Twitter added a night mode feature to its iOS app. When night mode is activated, the Twitter interface changes to darker hues so that Tweets are easier to read in low light, when you don’t want to be squinting at a bright white screen. The feature was originally introduced in Twitter’s Android app back in July.

You can tap night mode on or off whenever you like after pressing on the gear icon in the “Me” section of the iOS app. You can revert from night mode by following the same instructions. In addition, night mode can be toggled on or off by going into Settings then “Display and sound” options.

Twitter’s night mode does not activate automatically at sunset like the iPhone’s Night Shift mode. So, you have to set it manually, and can do it anytime (even if it’s light out–gasp!). This makes Twitter’s night mode more like a theme or skin. Its competitor, Tweetbot ($10 in the App Store), does have a setting to automatically switch between day and night mode depending on the time in your location.

 

 

[Source: Macworld]

Celebrating with Hadoop and Red Hat communities

Image result for Celebrating with Hadoop and Red Hat communities

This week I had the opportunity to represent Microsoft in keynotes at both the Hadoop and Red Hat Summits in San Jose and San Francisco, and during that time speak to many customers and members of the open source and big data communities. The energy in these communities is incredible and I’m proud to see Microsoft engage as an active participant and contributor. Below I want to share a quick summary of recent news related to both these events and communities.

Hadoop Summit

This year we celebrated 10 years of Hadoop and along with the rest of the community we’re working hard to make big data easier for customers, no matter the type of data, what they need to do or what platform they’re on. At Hadoop Summit, Hortonworks also announced Microsoft Azure HDInsight as its Premier Connected Data Platforms solution to deliver Apache Hadoop in a cloud environment. This announcement culminates a long-standing partnership with Hortonworks that started in 2011 when Hortonworks was three-months-old. It’s been rewarding to see how far Hadoop has come, and it’s now deployed in thousands of organizations. As an example, Jet.com is using Hadoop with HDInsight to help redefine the e-commerce category by providing consumers with completely transparent pricing that dynamically changes based on the actual costs of the transaction – warehouse location, payment method, and number of items shipped.

Looking back at our journey with Hadoop, it’s also been gratifying to see our contributions accelerate its adoption. Members of Microsoft have been contributing to the development of Apache YARN since its inception. We’ve also been leading or contributing to projects like bringing Hadoop onto Azure and Windows, speeding up the query processing of Hive, making cloud-based stores accessible via WebHDFS, and making Spark execution available through a REST endpoint. Recently, we also announced our commitment to Apache Spark at Spark Summit 2016 including:

  • Spark for Azure HDInsight General Availability, is a fully managed Spark service from Hortonworks that is enterprise ready and easy to use
  • R Server for HDInsight in the cloud powered by Spark, in preview today and will be generally available later this summer makes Spark integration easy no matter if you are working on premises or in the cloud.
  • R Server for Hadoop on-premises now powered by Spark, as the leading solution in the world to run R at scale, R Server for Hadoopnow supports both Microsoft R and native Spark execution frameworks made available this week. Combining R Server with Spark gives users the ability to run R functions over thousands of Spark nodes letting you train your models on data 1000x larger and 100x faster than was possible with open source R and nearly 2x faster than Spark’s own MLLib.
  • Free R Client for Data Scientists, a new free tool for data scientists to build high performance analytics using R.
  • Power BI support for Spark Streaming, General Availability, Spark support in Power BI now allows you to publish real-time events from Spark Streaming.

With our investments in R combined with Spark & Hadoop, statisticians and data scientists can rapidly train a variety of predictive models on large-scale data, limited only by the size of their Spark clusters. With Spark, R Server’s compiled code algorithms and transparent parallelization of regression, clustering, decision trees and other statistical algorithms speeds analysis 100x faster on terabytes of data.

 

Red Hat Summit

At Microsoft, we’re serious about building an intelligent cloud through a comprehensive approach that includes the open source ecosystem. Today, our cloud offerings range from support for Linux in Azure Virtual Machines – and nearly 1 in 3 VMs are running Linux today – to a Hadoop solution in HDInsight, or deep integration of Docker Swarm and Apache Mesos in Azure Container Service that represent our commitment to the ecosystem and highlight the value of our partnerships. In November, Microsoft and Red Hat announced a partnership to add value to the open source investments in the enterprise. At Red Hat Summit, we announced a number of important partnership milestones, including:

  • The general availability of .NET Core 1.0 and ASP.NET Core 1.0, a platform for creating modern applications for Windows, Linux and Mac OS X
  • In partnership with Red Hat and 21Vianet, this week we are extending support for Red Hat Enterprise Linux to Azure China operated by 21Vianet
  • Red Hat is making CloudForms 4.1 generally available, with deep support for Azure including state analysis, metrics, chargeback and retirement, making Azure the best supported cloud in CloudForms
  • Availability of a new OpenShift solution template on GitHub that makes it simple to deploy OpenShift in Azure

In March, we announced our plans to bring SQL Server to Linux, starting with a private preview. In the research note Microsoft Diversifies With Linux Support for SQL Server, Gartner wrote “SQL Server on Linux represents a bold statement that the company understands there is more to the overall IT world than just Windows and this flexibility is necessary to compete in the DBMS market.” Today, at the Red Hat Summit, I will show SQL Server running on Red Hat Enterprise Linux. Our goal is to make SQL Server the platform of choice to support any data, any application, on-premises or in the cloud, and providing you with platform choice. Bringing SQL Server to Red Hat Enterprise Linux will provide enterprise Linux customers with SQL Server’s mission-critical performance, industry-leading TCO, the least vulnerable database,[1] and hybrid cloud innovations like Stretch Database to access data on-premises or in the cloud.  We’ll first release the core relational database capabilities on Linux targeting mid-calendar year 2017.

MongoDB on Microsoft Azure

Additionally, this week, MongoDB announced MongoDB Atlas, a new elastic on-demand cloud service that will provide comprehensive infrastructure and management for its popular database. MongoDB Atlas will become available for Azure customers via a strategic partnership between the two companies. This partnership with MongoDB further reinforces Microsoft’s commitment to providing customers with open source solutions and the most comprehensive cloud platform on the market.

 

 

[Source: Blogs Technet]

Microservices, Java-on-Java top Oracle’s JVM design goals

Oracle's JVM goals: Microservices, value types, and more language support

Oracle’s plans for the Java virtual machine include greater language support, as well as accommodations for microservices and a heavier reliance on Java programming itself.

“We want a VM that will continue to be polyglot, that will interoperate with unmanaged languages, implement well-managed languages,” said John Rose, Oracle JVM architect, during the company’s recent JVM Language Summit conference in Silicon Valley. “Maybe in the next decade or so we’ll see C programs or C++ programs running in managed mode on top of the JVM. I wouldn’t be surprised.” In recent years, the JVM has become a home for a quite a few languages, including Scala, Clojure, JRuby, and Groovy.

Rose also cited a trend toward layered software implementations involving strong abstractions, a simplified data model, and Java-on-Java, in which Java is implemented on top of itself. The Graal Project, leveraging Java for writing JVM runtimes and compilers for other languages, and Project Panama, for interoperability between Java and C/C++, help drive Java-on-Java. Instead of having a lot of C++ code and a bytecode interpreter atop a stack, there would be various factored layers, perhaps low layers of C++, midlayers of low-level Java, specially compiled ahead-of-time Java, and Java components.

Java-on-Java “heals the rift between having to code in C++ and Java at the same time,” according to Rose. “It lets you use one high-leverage point, the Java compiler and the Java JiT and the Java type system, to manage more of the important parts of your system.”

Density and scaling goals for the JVM include enabling it to work with microservices. “We want those terabyte heaps,” Rose said. “We also want to be able to do tiny microservices, maybe running a few of them in the JVM or maybe a million of them in the JVM but independently.” Scaling will entail a fast startup, data sharing, and big heaps, making immutability more important. Aside from its JVM plans, Oracle recently cited work to retool Java EE for microservices and cloud deployments.

Honing in on language capabilities, Rose was bullish on value types, to mend a rift between classes and primitives. Value types enable huge changes through the stack and depend on parametric polymorphism, making generalizations across all values. “This is the last big thing that Java needs, in my opinion,” unifying primitives and objects but with everything looking like a class, he said. Rose sees value types as comparable to the impact of generics or lambdas in Java.

Rose listed eight specific goals for the JVM for the next 19 years, saying they had been initially expressed in 2015. These include having a uniform model, with objects, arrays, values, types and methods looking similar; memory efficiency, with tuneable data layouts and efficient code; optimization, in which shared code is mechanically customized; and post-threading, with confined, immutable data and granular concurrency.

Oracle wants Java to be broadly useful for running languages like Scala and Clojurewhile maintaining the compatibility to run even 30-year-old “dusty” JAR (Java Archive) files. The company also is emphasizing performance and interoperability. “Our JVM is more slowly changing than the hardware over which it runs, which means that we are always getting the latest chips and trying to figure out how to do the trick again on the new chips,” Rose said.

 

 

[Source: Javaworld]