Tuesday 30 August 2016

Android 7.0 Nougat: First look at the new update

 


 
Google has started seeding the new Android 7.0 Nougat update, but if you are reading this you already know that. The update is currently only available for the more recent Nexus devices including the Nexus 6, 5X, 6P, Nexus 9 tablet and the Pixel C. It is a gradual process so in case you haven’t got the update, don’t panic, eventually you will get it.
We haven’t seen the OTA (Over the Air) update notification popping up on Nexus devices here in India, but we did manage to get a flashable version of the new update. Google has been pushing beta builds since March, but this is the real deal.
nougat-1
For starters, the UI is the same and like the beta builds, it comes with a new wallpaper with a pink hue. Everything else is somewhat similar to Android 6.0 Marshmallow. We have been using the new update on a Nexus 6P and here are some of the key changes and features that make a difference to the Android experience.
Notifications
nougat-4
One of the first things that you will notice is how the notifications show up. They feel more informative and richer. The notifications, be it on the shade or the lock-screen, now show up in broader cards which take up more space on your screen. They are stacked on top of each other and are more interactive offering you options to bundle multiple notifications from the same app into one card or quickly reply from there itself, without having to open the app separately. Long pressing brings notification options for the particular app so you can quickly put an app on silent.
Multitasking
Google has finally added multi-window support. This feature has been long awaited especially since Android devices are getting larger. Samsung and a few other OEMs have been providing this feature through their own UIs but now this becomes a standard. Android Nougat allows you to run two apps simultaneously in a split screen mode.
nougat-3
It’s quite simple too, just long press the square multitasking button when you are in an app. The app will automatically move to upper part of the screen. At the bottom you will see the familiar multitasking tab with all the running apps. You can select any of the running apps, or just hit home and select the app of your choice. It is a useful feature as you don’t need to switch between apps.
Since its a new feature, it will take time to be perfect. Not every single app supports the multi-window view feature but since it is now at a system-level, more support will follow soon. Some apps work great and most of the time we saw a warning message. Another interesting addition is the clear all button on the multitasking view. Closing all apps is finally simple.
Other tweaks
There are small improvements and tweaks here and there. For instance the notification shade now shows a thin row setting toggles rather than taking larger space. When you further swipe down, they open to a bigger page along with the notifications. The Android Nougat easter egg has also been added here, but for some reason it didn’t do anything for us. There are some new emoticons as well.
nougat-2
The settings, like the notifications, is more informative. Right when you jump to settings, you get some suggestions, like adding an account, or using your voice to navigate the menu. Each setting also gives you relevant info right there. For instance under Sounds you will see volume level of the ringer and under storage you will see how much storage is used. There is also a Data Saver feature which can limit what apps can do when connected to a cellular connection. Even the Doze battery saver mode has seen an update, but for now we can’t say much as we need to test the update for a few days to actually see if there is an improvement in battery life.
nougat-15
Animations are slicker, menus open smoother and everything feels snappy. There isn’t a major overhaul of things, but more of improvements. For an early release, it is quite stable. Judging by what we have seen, we expect some more additions once the new Nexus devices are out. Until then, we suggest that you should update to the new Android 7.0 as the whole OS flows better than ever.

Intel takes the wraps off seventh gen Kaby Lake architecture


Intel takes the wraps off seventh gen Kaby Lake architecture

 
Intel has officially taken the veils of their seventh generation processor platform called Kaby Lake. This latest platform is built via an improved 14nm manufacturing; it’s also the third architecture based on this 14nm process.
What is Kaby Lake?
Intel Kaby Lake CPU architecture Tech2 720
Meet Intel’s new baby, Kaby Lake. Note the size of the GPU section
Kaby Lake is a third generation chip made on the 14nm manufacturing process and probably the last. It’s also the seventh generation chip in Intel’s Core i series of CPUs.
Intel PAO
Intel has been following a tick-tock upgrade cycle for a while now, but with the 14nm process, they shifted to a PAO (Process Architecture Optimisation) system. Tick-tock essentially introduced a new manufacturing process in the tick, then optimised it by the tock. The second, fourth and sixth generation Intel CPUs comprise the ‘tock’. With PAO, Intel adds another step, where they further optimise and enhance the results of a ‘tock’ cycle.
Intel Kaby Lake U series CPU Tech2 720
A representation of the Intel U series Kaby Lake chip
While Intel didn’t reveal a whole lot about their entire line-up, it did share a great deal about its Y and U series mobile platform. Intel sums it up best when they say that the focus of Kaby Lake is the “immersive web.”
All you performance junkies and overclockers can stop reading right now though. Intel made no mention of their K series SKUs and Kaby Lake’s desktop variants. The only thing you really need to know is that the CPUs are socket compatible with the previous generation. Details of an Intel i7 7700K were leaked some time ago, but performance figures are lacking.
The immersive web
Source: Intel
Source: Intel
Intel’s entire pitch for Kaby Lake revolves around the fact that the future’s the internet and the future of the internet is video. They expect to see an increasing amount of 4K content and even 4K 360-degree video. If everyone’s going to be watching video, one might as well optimise for video.
Source: Intel
Source: Intel
While Skylake, the previous generation architecture, was perfectly comfortable dealing with FHD video content, its power consumption for 4K decoding is very high. Kaby Lake incorporates dedicated hardware channels for HEVC 10-bit and VP9 video codecs, making for extremely efficient 4K rendering and editing.
Source: Intel
Source: Intel
Intel claims that Kaby Lake will be capable of rendering multiple 4K video streams while still allowing you to work or browse the web. They also add that the impact to battery life will be minimal. In fact, they claim up to 2.6 times the battery life when playing back 4K video, as compared to Skylake.
If the data they provided is correct, we’ve no reason to doubt them, Kaby Lake uses barely 0.5W on average to decode a 4K video, while Skylake would require 10W of power.
Performance boost
Source: Intel
Source: Intel
Compared to Skylake, Intel claims that Kaby Lake will offer a 12 percent bump in productivity and a 19 percent bump in web browsing tasks.
Source: Intel
Source: Intel
The boost in web browsing is notable because Intel says that Kabe Lake uses an improved version of Speed Shift, which is more efficient at boosting performance for very short periods, a performance burst if you will, which is extremely useful in web browsing scenarios.
Intel Kaby Lake vs Skylake speed shift boost Tech2 720
Source: Intel
Intel states that part of this bump is down to the higher clock-speeds that they’re able to extract from the new manufacturing process. The rest can be attributed to overall platform improvements (such as HEVC and VP9 support).
Most of these comparisons were between an Intel i5 6500U and an Intel 7500U. When compared to Sandy Bridge, a 5-year old architecture, Intel claims a performance bump of 70 percent and a 3.5x improvement in graphics performance.
Features
Along with the usual Intel bells and whistles, Kaby Lake will support PCIe Gen 3.0, which offers higher bandwidth when compared to the previous iteration (8GT/s vs 5GT/s). PCIe-based SSDs should benefit the most from this.
All Intel Core m series, Y series and U series chipsets will be dual-core CPUs with support for 4 threads via hyperthreading.
The rest of the architecture remains very similar to Skylake.
Source: Intel
Source: Intel
Judging by what Intel has told us so far, it’s clear that Kaby Lake will greatly benefit mobile devices (laptops, 2-in-1s, etc.). Since 4K content is still sparse, the average user will probably not notice much of an improvement in performance, especially if they’re on Broadwell or Skylake. For the high-end desktop user though, the performance rewards will also be less tangible, especially since battery life is not an issue.
Obviously, if you’re still on an older generation platform like Sandy-Bridge, it’s high time you upgraded.

Thursday 25 August 2016

Intel's New Joule In The Crown

Intel's New Joule In The Crown
Written by Harry Fairhead   
Wednesday, 17 August 2016
I couldn't resist the headline, but the news is perfectly serious. IoT hardware is becoming more and more like a full desktop computer. Move over Arduino, the Intel Joule might well crush you with its processing power.
Presumably Joule is named after James Prescott Joule, the physicist rather than the unit of energy in line, with the naming of previous products, Edison, Galileo and Curie.
joule1
The world of IoT has just received a small shock with the announcement at the Intel Developer Forum that the Joule is Intel's new 4-core embedded module. It is very small, 48 x 24 x 3.5mm, but the key specification is its processing power. It has a quad core Atom running at 1.7GHz with 4GBytes of RAM - this would be a reasonable spec for a desktop computer. It also has integrated HDMI GPU graphics, and USB for keyboard and mouse. You can take this device, add a monitor and keyboard and start using it. The operating system also has support for a graphical user interface. In addition it has a standard camera interface and Intel clearly expect it to be used in video processing applications such as computer vision.
As such it is part of Intel's new Augmented Reality push. The Joule API includes support for Intel's RealSense depth camera and these two technologies are key to the merged reality headset, project Alloy, also announced at the event.
The rest of the specification is:
Processor
  • Quad-core Atom processor, 64 bit, running at 1.7 GHz with 2.4GHz burst mode. 
  • 4 MB cache (2 MB per core-pair)
Memory
  • 4 GB of package-on-package (PoP) LPDDR4 RAM
  • 16 GB of eMMC 5.0
Graphics
  • On-chip Intel® HD Graphics
  • 450 MHz with burst up to 650 MHz
  • 4K video display
Wireless
  • Integrated Wi-Fi and Bluetooth
  • 802.11ac Wi-Fi dual band (2.4 and 5 GHz)
  • Bluetooth low energy compliant
  • MHF4 connectors for external antennae
Operating System and Board Support Package
  • Yocto Linux OS for IoT
  • Linux 4.4 kernel
  • Open-source UEFI-compliant BIOS
  • I/O configuration via EEPROM

Two Board-to-Board 100-Pin Connectors (J6 and J7)
  • 1 USB 3.0 On-The-Go (OTG)
  • 1 USB 3.0 or PCIe (muxed)
  • 2 USB 2.0
  • 1 SD* card 3.0 interface
  • 1 4-lane MIPI* display serial interface (DSI)
  • 1 1080p HDMI 1.4b
  • 1 I2S audio interface
  • 5 I2C bus
  • 2 SPI bus
  • 2 digital microphones
  • 4 Universal Asynchronous Receiver/Transmitters (UART)
  • Up to 48 GPIO (with 4 PWM)
  • 32.768 KHz and 19.2 MHz clock outputs
MIPI Camera Connectors
  • 2 4-lane MIPI* camera serial interface (CSI)
  • 1 programmable clock per MIPI CSI interface connector
  • 2 I2C bus
  • 4 GPIO per expansion board connector
  • 3 programmable voltage rails per connector
The system has been launched as two different modules that can be hosted on the same breakout board. The other is slightly less powerful with a 1.5Ghz Atom processor and 3Gbytes of RAM and, as yet,  no price for it has been announced.

joule2

The Intel Joule 570x developer kit with expansion board, costs $369.99 and is available now. This is expensive compared to a Raspberry Pi but the two are not really comparable. The Joule is a desktop class machine in a tiny package - of course not quite as tiny when you throw in the development board. The Tuchuck expansion board is open source, which makes creating custom hardware easier. It exposes all of the interfaces on the Joule plus a fan header and power sockets.
At the moment it isn't clear if the Joule is a 1.8V device although it seems likely that it is. However this might not be as important an issue as for the Intel Edison as the expansion board offers 3.3V (5V tolerant) connections. It also isn't an Arduino compatible expansion board, which is a great relief to anyone trying to do serious work

Tuesday 23 August 2016

Google Is Building A New Operating System Fuchsia For All Things Connected

 


Google has Android, it also has Chrome as an operating system. But it is not done yet. The search giant is developing another operating system called Fuchisa (Purple + Pink == Fuchsia). While the functionality and direction of the OS is still not clear, it can virtually run on any device.For this operating system, Google has taken a slightly different path for the build. Instead of using Linux as a base like in Android and Chrome OS they have chosen the Magenta Kernel. That is based on “LittleKernel” project used in embedded system. It means that the operating system is meant run on connected devices to the Internet.Furthermore, two of the engineers Christopher Anderson and Brian Swetland have been involved as embedded systems experts on projects such as Nexus Q and Android TV.
Even though being a Kernel for the embedded device Magenta supports users profiles and app permission-like security features.As pointed out by Android Police, there are three parts of the base of the operating system Google is building. The first one is the programming language called Dart, the second one is called user interface building platform called Flutter, and the third one is Escher which helps in light infusion and shadows. Escher is responsible for a lot of animations and transitions in Google’s material design structure.Right now Google has hosted the code of Fuchsia in their own code library and GitHub. The readme documents of any library don’t provide any substantial details about its features. There are speculations that this can finally be the merger of Android and Chrome OS.Right now the code is available for the PCs and even ARM devices to be compiled and run on. One of the Google developers on the project Travis Geiselbrecht has said that the OS will be coming to Raspberry Pi 3 as well.One of the post at hacker news even indicates to operating system being tested on various platforms such as Acer laptops.

Fuchsia: Google is Building a New Operating System, But No One Knows What it’s For

                              
In Brief
Google quietly uploaded the code for a mysterious new OS in their depository and on Github, but they're not saying what it's for.

Goodbye, Linux

Google appears to have started building a new OS. They quietly posted the pile of code under the name ‘Fuchsia’ in their code depository and on Github, but they haven’t said a word about what it’s for.
The odd part is that, while this OS can run on both phones and PCs, it isn’t related to Android or Chrome OS. And Google’s statement about the OS doesn’t help with figuring it out: “Pink + Purple = Fuchsia.” 

raspberri pi projects

simple Humanoid robot using raspberri pi

Personal robots may become just as common to us as personal computers. They can provide information to people, have conversations with the family and take care of children.
Maybe one day, we will be able to implement our digitized consciousness into robots, and our children will still learn from us and interact with us even we are already physically dead for a long time.
In case you ask yourself where the name of the robot comes from: the robot was named after the cyborg Alex Murphy from one of my favorite science fiction movies RoboCop (1987).
Humanoid robots walk through an uncanny valley, a hypothesis stated by Masahiro Mori that as the appearance of a robot is made more human, some observers' emotional response to the robot will become increasingly positive and emphatic, until a point is reached beyond which the response quickly becomes that of strong revulsion.
Following plot speaks volumes:
Source
That is why I put a lot of thoughts in Murphy's design, not making it too human like or worse - look like a zombie. Revised the design a few times. Not a replica of the human body, just loosely based. For the moment I will leave you here, providing you just some technical data, design goals, a basic system design and impressions of the robot during its development in this project details chapter. Please check out the project logs for more...So far:

Murphy's specs

Height: 60 cm
Weight: 3 kg
Drive system: DC gear-motors with encoders
Actuators: 2 x 2 DOF robot arm, 1 x 2 DOF head, 1 x base rotate unit
Output devices: 1 x Loudspeaker 2 W, 8 Ω
CPU's: 2 x Arduino Mega, 1 x Raspberry Pi
Sensors: 7 x HC-SR04, 2 x microphone, 1 x Raspberry Pi camera, 1 x RGB and Gesture Sensor - APDS-9960, 1 x PIR, 1 x flame sensor, 1 x humidity, barometric pressure and ambient temperature sensor, 1 x FIGARO TGS 2620 gas sensor, 1 x voltage sensor, 1 x 3-axis accelerometer
Power source: LiPo 11.1 V, 5500 mAh
Target environment: indoors

My design goals

  • Make everything Open Source
  • Use off-the-shelf components whenever possible
  • Make it as easy to build as possible
  • Cost less than USD 500 to build at home
  • Make it modular and easy to upgrade
  • Make it easy to customize and open to improvements
  • Build an actually useful robot companion, not just a toy
  • Step-by-step building instructions will be provided completely as 3-D drawings (like LEGO or IKEA)

System design

I will probably use a behavior-based robotic system, first suggested by Rodney Brooks (1986). Below figure shows an example of a behavior-based robot control architecture:
One can visualize a behavior-based architecture as a layered stack of parallel behaviors where the behaviors at the bottom deal with the survival of the robot and the behaviors at the top achieve desirable goals if opportunities exist. The design of a behavioral control system proceeds incrementally. The first layer is designed, tested on the robot and refined until it is satisfactory. At this point, the robot is already operational. Then, the second layer is designed, tested and refined. At this point the robot can make use of both types of competences. And so forth... [quote from Bio-Inspired Artificial Intelligence - Theories, Methods, and Technologies, The MIT Press].









for more details click this link
https://hackaday.io/project/10232-humanoid-robot-named-murphy