Monthly Archives: May 2017

Orange Pi 2G-IoT Challenges the Raspberry Pi

If you are looking for an alternative for the ubiquitous Raspberry Pi (RBPi) or one of its siblings, give the Orange Pi 2G-IoT a second look. The Shenzhen-based maker of the Orange Pi developer board has made this one to rival the RBPiZW, the RBPi Zero W.

The Orange Pi 2G-IoT is a new design for a single board computer, available for sale of AliExpress. The device can run Android, Debian, Ubuntu, or Raspbian. It gives builders a 2G antenna to run applications for the Internet of Things, and that is where it gets the IoT in the name, while offering wireless LAN and Bluetooth for the same price as that of the RBPiWZ.

Featuring a 1 GHz ARM Cortex-A5 processor running at 32 bits, the Orange Pi 2G-IoT has a 256 MB RAM, and GC860, a Vivante graphics processor. The board supports 802.11 b/g/n Wi-Fi, Bluetooth 2.1, and 40-pin GPIO connector that matches the RBPi GPIO layout. Additional features include audio and video outputs and inputs, and USB 2.0 ports. There is a slot where you can insert a SIM card, while the 2G antenna supports GPRS/GSM data connections. The only two points of difference with the RBPiZW is it has double the RAM, that is 512 MB, while the Orange Pi 2G-IoT has 500 MB of on-board NAND flash to go along with the SD card slot.

With the Orange Pi 2G-IoT, you do not get a Display port, HDMI port or a VGA port. However, an LCD connector is present, where you can connect an external screen.

However, as the Orange Pi 2G-IoT is still new in the market, anyone who plans to use these devices must proceed with caution on two points. One, check if there is software support for the devices. The RBPiZW is a known entity and has a huge array of operating systems and software to run on it.

Second, those buying the board should check if there are carriers still supporting 2G. For instance, in the US, AT&T and many other carriers have killed off their 2G network, and many are planning to do so very soon. The situation is very similar in the UK. However, there are still some in other parts of the world who continue to support 2G and may do so for years to come.

These ultra-small single board computers offer a lot of options, the list continuing to grow at the lowest prices. After considering the shipping costs, we can call these the sub-$20 boards. This includes the RBPIZW that has the Wi-Fi and Bluetooth, the Orange Pi Zero, and now the Orange Pi 2G-IoT.

Even with all the impressive features at below $10, especially the addition of GPRS/GSM, the Orange Pi 2G-IoT is not likely to kill the sales of RBPiZW. This is mainly because of the community support the RBPiZW currently enjoys. Of course, the Orange Pis are great little computers, but if you run into a problem with them, you are likely to find less support online, as compared to what you can expect from the huge online community supporting RBPiZW.

Stylus: The Future of the Electronic Pencil?

So far, people have been rather disappointed in using Electronic Pencils (styluses) for the purposes of taking notes on their smartphones. However, a new combination of the MyScript Nebo note-taking app and the Apple Pencil on the iPad Pro seems like a winner. This is really helpful to people who are coming up with ideas at odd times of the day, and have to look for a piece of paper and pencil to jot them down.

The trouble with the paper and pencil approach is eventually every hand-written note has to be laboriously transcribed to the word processor running on a PC. Additionally, a drawing program such as Inkscape or Visio must recreate any hand drawn sketches. Many also have the habit of capturing on paper random thoughts as they crop up in their heads, and later, going back to flesh things out, moving paragraphs around, deleting some, editing others, and adding new material. At the end, there is invariably huge amounts of crossing outs, intermittent notes, and forward/backward pointing arrows.

Presently, those with smartphones, capture their notes by typing things out at a snail’s pace on the soft keyboard, although some are adept at using the same keyboard at lightning speeds. However, the result is nowhere near those achieved with a piece of paper and pencil.

The Apple app store offers two apps free for use—the MyScript Calculator and the MyScript Nebo. You can use the calculator on your iPad by writing your problem using your finger or a simple stylus. Therefore, this becomes a handy app for those always doing simple calculations.

However, the MyScript Nebo app is different. It will not allow you to proceed unless you have an Apple Pencil. Although the Apple Pencil is expensive compared to other styluses available in the market, it is worth spending on this intelligent stylus.
Apple’s Electronic Pencil has a much finer nib than most other styluses do, along with pressure and tilt sensors inside. Once you have mated the stylus with your iPad Pro using Bluetooth, using the stylus becomes a simple affair—just detach the magnetic cap and insert the end of the pencil into the lighting connector on the iPad. A prompt will come up asking if you want to mate this pencil. Simply touch the ok button, and the link is established.

By scanning the signals from the pencil almost 240 times every second with almost zero latency, the iPad Pro achieves results that are close to actual writing with a pen on paper. One can think of this as using crayon or paint brush on paper, depending on the application. The app MyScript Nebo recognizes handwriting very efficiently, and it is much better than other handwriting recognition software available on the market. It also has a spelling correction feature, which works as you write.

A simple scribbling motion is all you need if you wish to delete a letter, sentence, or a paragraph. In the same way, inserting additional material is also possible wherever it is necessary. MyScript Neo is good at deciphering bad handwriting, and recognizing sketches. A simple eraser tool is available to rub out unwanted parts.

RS485 & Raspberry Pi: Monitoring Power

Commercial data centers, lighting controls, utility rooms for buildings, and others need to keep a tab on their power consumption. The normal way to do this is by using electronic voltage meters and multi-branch current monitoring circuits. Vytas Sinkevicius wants to monitor power consumption using the ubiquitous single board computer, the Raspberry Pi (RBPi) as the main controller and the RS485 interface in a Branch Current Monitor (BCM) system.

The heart of the power monitoring system is an RBPi 3. Other parts the system uses are a Pi-SPi-RS485 Interface, a VP-EC-BCM Interface, a breakout PCB for an 18-Channel Current Sense Transformer, and a few Current Sense Transformers. Vytas will be writing the software in C, using the Geany compiler.

Electrical engineers use two types of current sense transformers for measuring current. The first type has a continuous hollow core, with the wire carrying the current passing through the hollow of the core. This type of current transformer is suitable for new constructions and requires the main power to be turned off for installations. The breaker wire has to be removed and re-connected after the current transformer is attached.

The second type of current transformer has a split hollow core, where one-half of the core may be separated from the other. Split cores are ideal for applications where the power wiring to the breakers cannot be switched off. By separating the top half of the core, the breaker wire can be placed in the hollow of the lower part, and the top half of the core replaced thereafter. Vytas is using a split-core current transformer, model type CR3110-3000, and CR Magnetics manufacture it.

The Pi-SPi-RS485 Interface provides power to the VP-EC-BCM Interface and communicates with the RBPi. As the RBPi and Pi-SPi-RS485 combination uses the Modbus RTU and RS485 protocols, they can be located as far as 4000 feet away from the actual area where power is being monitored.

The Pi-SPi-RS485 is a perfect fit for the RBPi3, as its ports match the GPIO port on the RBPi3. Moreover, as it duplicates the GPIO expansion port on the other sides of the Pi-SPi-RS485 module, additional modules are easy to add. You can fit the module directly on the back on an RBPi3, or use optional mounting hardware to connect and keep them alongside. All RS485 signals are duplicated on terminal blocks on the board, and on the RJ45 connectors as well.

Each RS485 module has its own power input (9-24 VDC) for powering remote transmitters, and its LDO regulator operating from the 5 VDC bus provides the 3.3 VDC. Therefore, this does not load the 3.3 VDC bus of the RBPi. There are on-board LED indicators for indicating the status of power and RS485 signals. Termination resistors can be selectively switched in using jumper settings provided. The module provides power to the VP-EC-BCM Interface over a CAT5e cable via the dual RJ45 connectors.

The VP-EC-BCM Interface made by VP Process Inc. does the actual power monitoring. This is a converter unit for current sense transformer with 36 channels. It has a 3-kVAC isolation between the primary circuits and the Power/RS485 Interface.

Orange Pi Prime – Another Rival for the Raspberry Pi 3

There is another Orange Pi among the branches of the highly productive Orange Pi tree belonging to Shenzhen Xunlong. This is the Orange Pi Prime—another rival to the most popular Raspberry Pi 3 (RBPi3). According to the year-end Linux hacker SBC roundup, half a dozen individual Orange Pi models were already existing, and if all the new variants are to be included, that number almost doubles. Proceeding at this rate, the company’s engineers will have checked out almost all possible combinations possible with size, RAM, I/O, and hacker board layout for an Allwinner processor.

Similar to the recent releases of Orange Pi Win and its sibling, the Orange Pi Win Plus, which are built on the quad-core, Cortex-A53 Allwinner SoC, the $30 Orange Pi Prime is also a fully open source SBC. In addition, similar to the more minimalistic Orange Pi PC 2 and its sibling the Orange Pi Plus H5, the Prime tab means they are using the newer SoC, the Allwinner H5. However, compared to the Allwinner A64 of the Win boards, the Orange Pi Prime has the more powerful Mali-450 MP2 graphics processor. The H5 processors typically run at clocks of 1.2 GHz.

Whenever a Linux capable hacker board surfaces in the market, people refer to it as an RBPi competitor, which could be erroneous—considering their features, size, and prices vary considerably. However, in the case of the Orange Pi Win and the Orange Pi Prime boards, including some boards such as the Odroid-C2 and NanoPi A64, the comparison with the RBPi3 is downright correct, given the close approximation of their feature set, performance, and price. Even their 40-pin expansion connector is pin compatible to that of the RBPi3.

The Orange Pi Prime is very similar to the Orange Pi Win board. Both have the generous 2 GB RAM, same as that available on the Odroid-C2. Also, just as the Orange Pi Win does, the Prime too has a micro SD card slot, Bluetooth, 802.11b/g/n Wi-Fi, a GbE port, HDMI port, AV, microphone inputs, MIPI-CSI, and a 3.5 mm audio output.

Apart from the above, there are other common features as well. These include the 40-pin connector, debug, GPIO, IR interface, and an operating temperature range of -10 to 65°C. The Prime has a footprint of 98 x 60 mm, which is only slightly larger than the 93 x 60 mm footprint of the Win. Among the differences with the Win, the Prime has only three USB 2.0 host ports, and does not have a battery connector, optional eMMC, or PMIC.

Unlike the Win boards, the Prime will not be supporting Windows 10 IoT in the future. The Linux distributions for the Prime are also somewhat different. They include Android 4.4, Ubuntu Desktop, Debian Desktop, and Arch Server.

The hardware specifications for the Orange Pi Prime include the Allwinner H5 processor, which is a 4x Cortex-A53 and an ARM Mali-450 MP2 GPU. The board runs on a 2 GB DDR3 SDRAM, has 2 MB NOR flash memory, and a micro SD slot with up to a maximum capacity of 64 GB.

Redefining MEMS with 3-D Interactive Projection

At the Mobile World Congress 2017, Bosch introduced a combo of a micro-scanner and projector, capable of turning any surface into a virtual user interface. Bosch is the world’s oldest and biggest manufacturer of micro-electro-mechanical systems (MEMS), and in its combo projector, it is using infrared for scanning and laser for projecting.

Currently, engineers are using MEMS devices for a variety of gadgets, especially where a human-machine interface (HMI) is necessary. These include in-car heads-up displays, infotainment, medical devices, robotics, industrial equipment, and on the factory floor. With the new microscanner BML050, Bosch Sensortech has extended its portfolio to include optical microsystems. This move also expands Bosch’s market from being only a component supplier to becoming a system supplier as well.

To sense where the user has placed his finger on the projected interactive display, the new Bosch BML050 uses a combination on two MEMS scanning mirrors. One of the mirrors tracks the X-direction, while the other scans the Y-direction. Sensing the finger also makes use of an infrared laser and an RGB laser.

The integrated module for infrared, red, green, and blue (IR-RGB) is only 6 mm high, and is capable of HD resolution. The two MEMS scanning mirrors are capable of both projecting images as well as collecting the reflected light, thereby determining accurately where the user’s finger is touching the projected image. According to Bosch, this technique is adaptable to 3-D scanning as well, where they can apply time-of-flight calculations using the reflected light from an object.

A major advantage over Digital Light Processors (DLPs), the Bosch laser-based MEMS scanner is always in focus, even when the projection surface is uneven. According to Stefan Finkbeiner, the chief executive officer of Bosch Sensortech, DLPs require thousands of mirrors that need focusing, and the entire outfit is expensive.

At present, the reference design of the Bosch BML050, although containing all the technicalities for use in almost any application, is much larger than the expected OEM circuit board. Finkbeiner informs that despite this, customers are already integrating the BML050 into their products, and they will be in the market by Christmas this year.

The BML050 has a two-mirror system, with one hinged in the X-direction and the other in the Y-direction. The mirror system projects from the module, which measures only 6 x 24 mm, and uses 30-lumen lasers. This arrangement allows Bosch to alter the size when using low-power lasers, or when using high-power lasers for instance, for industrial sized images. The reference design for the BML050 contains all required drivers and processors. This includes ASICs for driving the mirrors, processing the video, managing all colors, managing the system and laser power with two PMICs.

According to Finkbeiner, the two-MEMS mirror architecture is very simple to integrate. Therefore, for the future, Bosch is planning to use a sealed module design after further miniaturization. The design will then be suitable for use in tiny gadgets such as for IoTs and smartphones. Very soon, you may find virtual human-machine interfaces on everything from toys to industry equipment on the factory floor, robotics, medical devices, and infotainment such as in in-car heads-up displays.

What is Special about the SkyX Drones?

SkyX is a drone maker from Markham, Ontario. This startup has some unique designs for industrial drones. For instance, the SkyOne drone of the company can take off and land without needing a runway. That is, its takeoff and landing is more like that of a helicopter, but in flight, the drone resembles an airplane more closely. In technical terms of the drone industry, the SkyOne has both Vertical Take Off and Landing (VTOL) and fixed-wing elements, while flying more than 40 km on one charge.

SkyOne has a plethora of sensors and cameras on-board, enabling the drone to collect data about anything below it. It then sends the collected data to cloud-based applications for analysis. For launching and landing the drone, SkyX provides proprietary charging stations, which the company calls xStations. When the drone is not moving, the xStation closes a shell over the UAV, protecting them from theft, and charging them. The charging stations charge the batteries within the drone directly, rather than removing and replacing them.

Other drone producing companies such as Matternet allow their UAVs to land on charging stations, where their depleted batteries are swapped with fresh ones. Charging stations can be positioned along a route, giving the drone a virtually unlimited range. This scenario is likely to continue unless battery technology and other power systems improve significantly.
Didi Horn, the founder and CEO of SkyX, had earlier worked for the Israel Air Force. As he always wanted to develop drones and aviation for commercial use, he went in for the long-range UAV consumer market, where the demand was huge, and the products scarce. According to Horn, the world already has millions of kilometers of oil and gas pipelines, all at the risk of leaks and/or terror attacks.

The oil and natural gas industry faces its biggest challenges when inspecting its pipelines for leakages or damages. Its critical infrastructure can be difficult to monitor, especially when the lines cover several kilometers, often crossing inhospitable terrain. For instance, the Internal Energy Agency claims the expenditure on pipeline monitoring alone costs energy companies more than $37 billion every year.

Initially, SkyX is targeting the energy industry, since its UAVs and charging stations can then be configured to cover long distances from one pumping station to the next along pipelines carrying oil and gas. According to Horn, the drones can also cover vast farms carrying acres of solar panels installed on them. Likewise, the UAVs can also be used for inspecting wind turbines installed in remote areas. As the xStations are capable of plugging into typical electrical outlets, they can be connected to solar panels or any other freestanding generators as well.

While the drones being used in the field today have some features still under development, SkyX is working with several energy companies for conducting pilot projects and safety tests in the US. Although the drones can fly autonomously, the company has to secure permissions for flying beyond the line of sight of human observers on the ground.

In the long run, apart from improving the efficiency, energy companies may find using drones from SkyX to be less expensive.

OLED Lighting in the Auto Industry

In recent years, a number of industries have started using Organic Light-Emitting Diodes (OLEDs) in diverse ways. The automotive industry, in particular, has seen a huge potential in OLEDs. For instance, very soon Audi will be coming up with OLED taillights. At present Audi has presented prototypes of the taillights. At the LOPEC Congress, Audi provides advanced insights into the needs of the automotive industry that the deployment of OLEDs will require to meet, and the future of automotive lighting.

So far, there have been plenty of developments. At LOPEC, Audi demonstrated prototypes of their OLED taillights, which they claim have reached production stage. However, using OLEDs in vehicles has always been a challenge, although OLED lighting installations and table lamps have been around for a while, and these are in use in museums, clubs, and restaurants.

Difficulties of Using OLED in Automobiles

Major hurdles OLEDs have to cross when in use in automobiles are they have to withstand humidity, heat, cold, UV radiation, and constant vibration. All these can reduce the life span of OLEDs drastically. Audi claims to have solved this problem by encapsulating their displays hermetically, which they claim will make the displays as stable as LEDs.

Why Use OLED in Place of LEDs?

Regular LEDs act as point sources of light, and it requires substantial development work for generating an even light from them. On the other hand, OLEDs are evenly radiating sources of light, and they naturally produce a uniform illumination. Moreover, their thickness is only about a millimeter, which makes OLEDs more suitable for automotive design.

Designers find OLED appearance is high quality, both when off and on. This is because it has a simple and clean surface. As design is an important aspect of the automotive industry, it makes OLEDs ideal for such use. Most automobile owners expect a certain lifestyle from their vehicles, apart from its functional use of transportation from point A to point B.

However, for use as turn signals and brake lights, the light intensity from OLEDs is not adequate, and will have to be increased. The automotive industry is also working on using flexible OLEDs. At present many are using glass-based OLEDs, but these are rigid, and using plastic foil substrates as the base for OLED is opening up a whole new world of opportunities for the designers.

Audi is expecting LOPEC will open up a huge bandwidth of business and research institutes for them. They expect to hold discussions with specialists using this breadth of activity, and to meet other OLED manufacturers and materials developers.

What the Future Holds?

In about a decade from now, the world will be witnessing innovations in vehicle lighting that most can only dream about today. As it is, a vehicle’s lighting system already functions as a form of communication—hazard lights, turn signals, brake lights, for example. In the future, driverless cars will need to interact with others on the road with even greater sophistication. One of the visions Audi has is of a three-dimensional OLED display extending the entire tail of the vehicle, on the panel of the body, and integrated OLED within the windshield.

What is a QLED?

Recently, Samsung has announced their new TV technology using QLEDs to counter the OLED TVs that LG and others have put on the market. QLED stands for Quantum dot LED, and though Samsung has been using the concept of quantum dots in its TVs for quite some years now, they claim they will be bringing out several flavors of the QLED technology.

According to Samsung, QLEDs are transmissive, as LCDs are, and light goes through several layers to create an image on the surface of the screen. The company claims to be working on the ability of the QLEDs to overcome the challenges currently plaguing the OLEDs.

Although the Q part is currently demanding a premium in the price of Samsung TVs on the market, it will likely decrease in the future. According to Samsung, the QLEDs are bringing several advantages with reference to picture quality, such as higher light output and brighter colors. Samsung claims the light output in highlights is now 2,000 nits, a relative quick loss of peak luminance, and improvement of the delayed ramp-up.

Samsung compares their QLED performance with OLEDs and points out that the new quantum dots offer superior color, providing rich, fully saturated colors even for bright images. However, there is yet no independent testing to substantiate the claims of the company. Moreover, the claims cover only the high-dynamic range as against the standard dynamic range, where the OLED would be a superior performer.

While many observers claim to see better clarity and improved colors in the new TV technology, others fail to notice any difference. You can see QLEDs in Samsung TV model UE55KS9000, and in tablets such as Amazon Kindle Fire HDX 7, and HDX 8.9.

QLEDs contain quantum dots or microscopic molecules between two and 10 nanometers in diameter, which emit their own, differently colored light according to their size, when struck by photons or light particles. In the QLED TVs from Samsung, the dots are restricted to a film, and the LED backlight provides the illumination to light them up. This light then goes through other layers inside the display, which includes an LCD layer, ultimately creating the picture. As the light from the LED source passes through different layers before reaching the screen surface, the process is said to be transmissive.

The advantage of QLEDs is they can emit brighter, more vibrant, and more diverse colors—capable of making HDR content really shine—mainly due to their ability to achieve high peak brightness levels.

Compared to OLED TVs, it is more cost-effective to manufacture quantum dot TVs, which translates to better picture quality at a lower price. However, OLED displays still produce the deepest blacks, which means that OLEDs offer better contrast ratios. Therefore, while OLEDs offer true blacks, quantum dots offer great bright images.

QLED technology replaces the photoluminescent quantum dots with electroluminescent nanoparticles. Therefore, rather than coming from the LED backlight, light now comes directly to the display. Although the process is a lot similar to the light transference process within an OLED TV, within the QLED TV, individual pixels emit the light, thereby combining the best of quantum dots and OLED technology.

Multicolored LEDs Create Secondary Colors

Any student of physics knows mixing two primary color light sources produces a secondary color. For instance, mixing the primary colors red and green creates the secondary color yellow. There are three primary colors—Red, Blue, and Green. This process is easily seen in tricolor and RGB LEDs.

There is a disadvantage in this method. As two primary colors are necessary for generating a secondary color, two LEDs must remain turned on at the same time. Therefore, generating a secondary color means consuming twice the current a primary color requires. In battery powered circuits, the operating current of the LED indicator may be a significant fraction of the total current, and using the same current for generating both primary and secondary colors would be an advantage.

Using a sequencing method can generate balanced secondary colors from RGB, tricolor and bicolor LEDs, while using the operating current of a single LED. The sequencing method offers uniform intensities between the primary and secondary colors, and lower power dissipation. An added advantage of using the sequencing method with bicolor LEDs is keeping a simple pc-board layout with two pins while it produces three colors. Using the sequence method with RGB LEDs produces white light while consuming the operating current of a single LED.

The sequencing method works because it takes advantage of a property of the human eye. This is called persistence of vision, wherein images in the human eye persist for about sixty milliseconds after light from the object ceases to enter the eye. For instance, when a glowing coal is moved about in the dark, the eye sees a continuous red line.

When the human eye sees different primary colors flashed sequentially and quickly from one point, they appear to overlap in time, while the brain interprets the colors to be secondary colors, or, depending on the color components, even white.

Experiments with multiple primary-colored LEDs show that the above flash sequence should repeat every 25 milliseconds or lower, for the eye to treat the effect as a solid secondary color. In fact, the flash rate can go down to one microsecond, before the human eye can detect the degradation of the secondary color. Therefore, any clock source, say a convenient 40 Hz, should be adequate for creating secondary colors.

For the eye to properly see the mixed colors, the primary-color LEDs must be physically very close together, such as on a semiconductor chip. As an added advantage, diffused lenses are better, as this offers a wider viewing angle.

When using bicolor LEDs, the driver has to be bidirectional, as the LEDs are placed back-to-back in the chip. Moreover, currents for the three LEDs may have to be adjusted to achieve color balance between the primary and secondary colors. In addition, color balancing may be required also as LEDs have different intensities and efficiencies as the human eye sees them.

This correction can be done in one of two ways. As each LED has a current limiting resistor in series, the value of these resistors may be tweaked to achieve the necessary differentiation in individual currents. The other option is to keep the same current but tweak the duty cycle.

Is there a 64-bit Raspberry Pi?

Although the arrival of the Raspberry Pi 3 (RBPi3) heralded a huge speed boost for the Linux hacker board, this $35, wireless-enabled single board computer did not signal a switch over to 64-bit ARM computing. Even though the hardware, following so many other SBCs at the time, was 64 bits, the default Linux distribution from the Raspberry Pi Foundation is still 32-bit.

Eventually, there will be a changeover to 64-bit ARM firmware, as the technology offers significant improvements in performance. More power-efficient chips, such as the 64-bit x86 are also piling on the pressure. However, the Raspberry Pi Foundation is still not committing itself beyond considering a change in the coming months to the 64-bit for the default Raspbian distribution as the reworking of the code required for the changeover is going to be extensive.

The RBPi3 has advanced to the new quad-core of Cortex A53 BCM2837 SoC from Broadcom. Architecturally, this SoC is quite similar to the BCM2836 that the predecessor RBPi2 uses—the quad-core Cortex A7. The Pi Foundation claims that even while operating in 32-bits, the RBPI3 delivers more than 50% better performance than delivered by the RBPi2. This is because of two improvements, one due to the superior architecture of the Cortex A-53, and the other due to the higher clock rate of 1.2 GHz of the RBPi3, as compared to that of 900 MHz of the RBPi2.

While comparing the RBPi3 with the RBPi2, we find the BCM2837 on the RBPi3 is paired with the same VideoCore IV GPU from Broadcom, similar to that in the RBPi2. However, in the RBPi3, the GPU is clocked at a higher rate of 400 MHz. That precludes any video performance at 4K, deep learning projects, or any high-end VR from the RBPi3. On the other hand, the Odroid-C2, being equipped with a Mali-450 GPU, supports 4K video decoding.

Eben Upton, the CEO for the Foundation’s commercial arm, the Raspberry Pi Trading, has explained this. According to Upton, the VideoCore IV 3D is the only 3-D graphics core for the ARM-based SoCs that has been documented publicly and the Foundation wants to make the RBPi more open over time.

Apart from the new SoC, the RBPi3 has also added a wireless chip from Broadcom, the BCM43438, and this enables it with 2.4 GHz, 802.11n Wi-Fi, and Bluetooth 4.1 BLE. With this addition, the RBPi3 steals a march over the Odroid-C2, which lacks wireless, operates on a Cortex-53 Amlogic S905 SoC, and costs $5 more.

Whether to have wireless onboard, or to let users select their own wireless options via Ethernet or USB adapters, has been the subject of an intense debate. As earlier, cost was the main consideration, and the deciding factor came from the dropping prices of wireless chips. Further, the single antenna of the Broadcom chip can be soldered directly onto the board, rather than be used as a module.

Other than the slight shift in the placement of the LED, the new processor, and the wireless capability, the RBPi3 is identical to its predecessor, the RBPi2. They share the same dimensions, amount of RAM, and the 40-pin expansion connector.