Design News is part of the Informa Markets Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Raspberry Pi 3 Model B+ Is Ideal for Embedded Systems, Machine Learning

Raspberry Pi 3 Model B+ Is Ideal for Embedded Systems, Machine Learning
The latest model of the Raspberry Pi, The Raspberry Pi 3 Model B+ has a series of new features that should make it an attractive offering to embedded systems engineers and makers looking to explore advanced applications such as machine learning.

The Raspberry Pi 3 Model B has gotten an upgrade focused on making it an even better tool for serious embedded applications engineers and makers looking to explore more advanced applications such as AI and machine learning.

The Raspberry Pi 3 Model B+ boasts enhanced processor and connectivity performance that distributors believe will make it popular with embedded systems engineers. (image source: Raspberry Pi Foundation)

Here's a quick rundown of the Model B+ specs:

  • Broadcom BCM2837B0, Cortex-A53 (ARMv8) 64-bit SoC @ 1.4GHz
  • 2.4GHz and 5GHz IEEE 802.11.b/g/n/ac wireless LAN, Bluetooth 4.2, BLE
  • Gigabit Ethernet over USB 2.0 (maximum throughput 300 Mbps)
  • Extended 40-pin GPIO header
  • Full-size HDMI
  • 4 USB 2.0 ports
  • CSI camera port for connecting a Raspberry Pi camera
  • DSI display port for connecting a Raspberry Pi touchscreen display
  • 4-pole stereo output and composite video port
  • Micro SD port for loading your operating system and storing data
  • 5V/2.5A DC power input
  • Power-over-Ethernet (PoE) support (requires separate PoE HAT)

The newly released Raspberry Pi 3 Model B+, uses an updated version of the same processor as its predecessor, but in a new package that squeezes out even more performance. The Model B+ is built around Broadcom's BCM2837B0 64-bit processor, which now incorporates power integrity optimizations and a heat spreader, allowing the single-board computer to reach higher clock speeds, reduce power consumption, and better control the temperature of the chip.

The power supply is the new MaxLinear MxL7704 Universal PMIC, which enables the board to be more compact and flexible. The PMIC’s I2C interface communicates with the computer’s on-board SoC for dynamic voltage scaling, status monitoring, sequencing control and PGOOD routing. MaxLinear says these new features provide the Raspberry Pi 3 Model B+ with more knowledge and control of power status and performance than previous systems using discrete power management solutions. "We worked closely with Raspberry Pi all the way from product definition to product release,” Tuomas Hollman, MaxLinear Senior Director of Marketing, Power Management Products, said in a press statement. "This very open collaboration resulted in the MxL7704, which provides five rails pre-optimized for ease of use in single-board computer systems, with a host of features that allow monitoring, telemetry and additional flexibility."

Raspberry Pi CEO Eben Upton discusses the updates in the Raspberry Pi 3 Model B+. 

The Model B+ is also getting a boost in connectivity courtesy of a new Wi-Fi and Bluetooth combo solution from Cypress Semiconductor. Cypress' CYW43455 single-chip combo provides high-performance 802.11ac Wi-Fi, advanced coexistence algorithms for simultaneous Bluetooth and Bluetooth Low Energy (BLE) operations such as audio and video streaming, and low-power BLE connections to smartphones, sensors, and Bluetooth Mesh networks. Previous models of the Raspberry have used Cypress' CYW43438 802.11n Wi-Fi and Bluetooth combo SoC.

Chris Breslin, SVP Product at Newark element14, a top distributor of Raspberry Pi boards, told Design News, the new features of the Model B+ should make it an even more attractive offering for professional-level use cases. Breslin said that Newark element14 is expecting to see the Raspberry Pi increasingly adopted as an embedded device as engineers look for ways to bring products to market faster and with fewer resources. “We talk to our customers as part of our day-to-day relationships and we are seeing the Raspberry Pi increasingly adopted as an embedded device,” Breslin said. “We have recently held a couple of focus groups to help understand how our customers use the board in embedded applications, and what they would want to see in the next one. This board pretty much ticks every box on their wish list.”

Breslin said the new dual-band wireless will provide a more stable connection for city center IoT applications, “Combatting the challenges of a very crowded 2.4GHz band” as well as modular certification that will save time and money since design engineers will not need to re-certify this part of the board when it's embedded into a product. “The Power over Ethernet [PoE] HAT, available from the summer, also satisfies our customers’ requests for PoE – allowing the board to be booted and powered through one Ethernet wire, and removing the need for a memory card – also satisfying, in effect, customer requests for eMMC onboard,” he said.

The new Model B+ also comes at a rather exciting moment for makers looking to explore emerging applications such as AI and machine learning. Avid users will recall last year when Google announced it was planning to bring AI functionality to the Raspberry Pi and was surveying makers, engineers, and researchers on their needs. To that end, Google recently open-sourced its machine learning and data science workflow via a new project called Colaboratory, that allows users to create web-based machine learning models that can be run on a Raspberry Pi or similar single-board computer.

Design News' Don Wilcher recently wrote a tutorial on getting started with Colaboratory and a Raspberry Pi.

Are you excited for the latest Raspberry Pi model? Share your thoughts in the comments!

Chris Wiltz is a Senior Editor at Design News covering emerging technologies including AI, VR/AR, and robotics.

One of the major hassles of Deep Learning is the need to fully retrain the network on the server every time new data becomes available in order to preserve the previous knowledge. This is called "catastrophic forgetting," and it severely impairs the ability to develop a truly autonomous AI (artificial intelligence). This problem is solved by simply training on the fly — learning new objects without having to retrain on the old. Join Neurala’s Anatoly Gorshechnikov at ESC Boston, Wednesday, April 18, at 1 pm, where he will discuss how state-of-the-art accuracy, as well as real-time performance suitable for deployment of AI directly on the edge, moves AI out of the server room and into the hands of consumers, allowing for technology that mimics the human brain.
Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.