1 2 3 Previous Next


511 posts

Inforce Computing will be exhibiting at the ‘Robotics as a Business’ (RaaB), conference organized by the Silicon Valley Innovation Center (SVIC) in San Mateo, CAhttps_proxy (962×426)On day-2 of the conference (Friday, July 31st, 2015), SVIC will host a select few companies, including Inforce Computing, in the exhibit area. Come visit us at the Silicon Valley Innovation Center between 10:00AM and 4:00PM to see two interesting robots performing cool tasks powered by Inforce’s ARM-based embedded computing platforms, such as the Inforce 6410Plus SBC. The open-source 3D-printed Snapdragon Micro-Rover running Android and the TurtleBot running ROS on Linaro Ubuntu Linux, are exciting platforms to start prototyping robotics projects. Also on display will be the latest 4K Ultra HD video capture capable Inforce 6501 Micro SOM, which is ideally suited for “SWaP” constrained commercial drones and UAVs in agriculture, surveillance, law enforcement, and other industrial applications.



From the Inforce Computing Desk
©2015, Inforce Computing, Inc. All rights reserved.

I have a very strong resistance to the practice of doing something [anything!] just because “it is the way we have always done it”. I love to play Devil’s Advocate [or “Devil’s Avocado” as I heard someone quip the other day] and propose change just to shake things up. It may be that the tried and tested approach is, indeed, the best, but you cannot be sure until you have considered that other options.

When it comes to embedded programming, it is easy – most people use C or similar languages. But is that the only way? …

Why is C [along with languages similar to it] the most common way to code embedded software, when other options are used somewhat more widely for desktop software development? I think that there are two main reasons. First, embedded developers are quite conservative and resist new ideas until they are proven and the benefits are clearly apparent – “if it ain’t broke, don’t fix it”. That is not such a bad attitude to take if you want to produce reliable code in a predictable time [or, at least, have a good shot at these laudable objectives]. The second reason is that many embedded developers have either started out [like me] in assembly language programming. The programming paradigm of procedural languages is much the same as assembly language and, hence, the native machine code of the CPU. So, C feels “right” and must be most efficient. Well, maybe.

There are other possibilities, which really should be considered when starting out on an embedded project:

To read the rest of this entry, visit the Colin Walls blog.


Key considerations for next-generation high-performance embedded system design


Have you wondered what would it take to implement a modern embedded system design with a system-on-module (SOM) as opposed to building your entire design from scratch? What are some of the clear advantages to designing with a SOM? Are you perennially stuck in the build-vs-buy argument on whether or notDesign_with_micro-SOM-flow-chart.png to design the most time-consuming and complex compute portions of an embedded system in-house? What's a carrier board and how to design a custom one that can be re-used by new SOMs that come down the line? Is your design going to be obsoleted because of the lack of a microprocessor roadmap and upward/cross-compatibility of your custom carrier boards? Are you missing your market windows and are looking for a solution that addresses time-to-market issues and cost escalations? Download white paper >>>>

Have you hit the limits of your current system's compute horsepower? Would you like to obtain access to high-performance and low-power state-of-the-art Qualcomm® Snapdragon™ processors without the need to license them?

To find out answers to the above questions and kick-start your embedded design with the latest ARMv7 ISA compatible Inforce 6501 Micro SOM, please download a whitepaper on the subject here.

Vasu Madabushi

Inforce Computing, Inc.


We had a very successful series of events at DAC in which we highlighted collaborative successes with our strategic partners.


The video is now available for the TSMC+ARM+MediaTek+Synopsys event, which included the following presentations:

  • ARM and Synopsys presented results from their collaboration to deliver a reference implementation for mobile computing with an ARM Cortex-A72 processor in TSMC's 16FF+ FinFET process
  • MediaTek talked about the design challenges to creating complex SoCs for multimedia and LTE, and then shared their successes using ARM GPU and CPU IP in TSMC 16FF+ process with Synopsys' Galaxy design tools, Virtualizer Development Kits and the ZeBu emulation system
  • ARM, Synopsys and TSMC technologists discussed their collaboration to enable design in 10nm FinFET technology

tsmc-arm-mediatek-speakers.jpg    tsmc-breakfast-video-thumb.jpg

While we had incredible attendance and interactions at the DAC breakfast event, we're happy to now be able to make the video of this event available for those who were unable to attend live!


Would love you hear your feedback and/or questions!

I am involved in an interesting project, with which I need some help. And that might be where you come in. I had an idea for a book and managed to persuade the inimitable Clive (“Max”) Maxfield to work with me on it. The plan is that he and I will write some sections and jointly edit the whole book. We have an agreement with Elsevier to publish the fruit of our labors, so we are now somewhat committed.

This is why I am now asking for help …

The concept for the book is both simple and wide reaching: it will be a comprehensive guide to the languages used in embedded system development. This includes software programming, hardware definition and modeling languages.

To read the rest of this entry, visit the Colin Walls blog via Mentor Embedded.


Thread Applications.jpgYou may have heard of the Thread working group whose goal is to solve the wireless networking challenges in the IoT (and embedded space) and last week they released the latest spec for the protocol.  In order for the IoT to deliver any of the benefits we are expecting the standards issues have to be addressed so I see this as a very significant milestone to get those 20 billion devices connected.


Thread now has 160 companies participating in the standard (including Bill Curtis from ARM) and it looks like they have made some real progress.  Thread is mainly involved in the development of standards for the smart home but many of the issues they are addressing also apply to industrial, medical and automotive applications (think security, low power and mesh networking) so its worth watching.  Further to that point any organization that has such diverse members as Qualcomm, Proctor & Gamble, Philips and Big Ass Fans participating in a standards body has to have some fun meetings!


I have attached the Thread overview below if you would like to get a taste of what they are up to and the new spec is detailed here.


Coincidentally the giant US retailer Target opened their IoT Smart home demonstration store in San Francisco last week so I stopped by to take a look.  Target is selling the 40 IoT products on display. They have a very interesting video which talks about how IoT products come to market.  With the rise of the Maker movement and Kickstarter we see new products emerging that come from new sources and could be game changers so you can see why its in Target's interests to incubate and encourage the market.  The store has well known devices from NEST but some new and more obscure products that might becomes hits.  By the way, Target shows a top 10 list of products sold and Tile is number one,  the Tile is a Bluetooth based tracking device you attach to your keys, TV remote or bag to locate them when they mysteriously crawl under the furniture,  not the killer app for the IoT yet but its a start.


Target are very amenable to having events in the space and showcasing new devices so if you want to launch something for the IoT give them a call (Target's David Newman who introduces the video spoke to me and Brian Fuller about the project and was very passionate about the project).


Is the IoT going to sneak up on us via Target?

I recently came across the Geppetto tool (actually version 2.0 of the tool), developed by Gumstix a producer of Linux-based computers-on-module (COMs). Geppetto is an on-line tool that lets you build out your application in a pretty simple and straightforward manner.Geppetto.png


The new version of the tool introduces TuxApproved recommended mappings for buses, ensuring the best compatibility between customer-created hardware and standard Linux images. In addition, Geppetto 2.0 offers an expanded module selection, improved dimensioning, faster UI, and video tutorials.


With a fairly low buy-in cost ($1999), Geppetto 2.0 gives designers a relatively inexpensive way to reduce their product time-to-market and design costs. This could be significant, as in many cases, this refers to custom embedded systems.


On the hardware side, the company is offering the AeroCore 2 MAV control board and the Pepper DVID single-board computer. The AeroCore 2 is built with an ARM Cortex M4-based CPU. Aimed at powering intelligent, next-generation microaerial vehicles (MAVs), the board gives MAV developers a greater selection in finding a custom computing solution, adding CAM, Spektrum RC, and GPS interfaces. The board sells for $149.


The Pepper DVID board is a relatively powerful, compact solution for embedded developers looking to work with ARM's Cortex A8 processors, specifically the Texas Instruments' Sitara AM3354 CPU. This board sells for $119 and offers high-definition video output, 512 Mbytes of RAM, WiFi, Bluetooth, a microSD card slot, audio connectivity, a console port, and two USB OnTheGo ports.

ROS_6410Plus.pngInforce Computing has made key contributions to Qualcomm® and Open Source Robotics Foundation’s (OSRF) efforts to port the Robotics Operating System (ROS) to ARM® ISA compatible Snapdragon™ processors. If you haven’t heard about ROS, it is basically a framework of libraries and tools that run on Linux and Android OS to enable developers create and program robotics applications. These libraries are fully baked now and available for download from the ROS wiki. Several key functionality such as perception, identification, segmentation, recognition, motion tracking, stereo vision, grasping, etc., constitute the library. It is geared towards reducing the time and effort required to develop a robot powered by an ARM architecture based SBC such as the Inforce 6410Plus™.


Last December, I wrote about Qualcomm demonstrating the first port of ROS on the Snapdragon 600 powered Inforce 6410 SBC. In the meanwhile, Inforce has qualified both ROS Indigo (Ubuntu 14.x) and ROS Jade (Ubuntu 15.x) releases on the latest Inforce 6410Plus SBC. Here’re a couple of articles that appeared recently that talks more on the subject:

Qualcomm Developer Network: ROS Support for Qualcomm® Snapdragon™ - Bringing ARM into Robotics

readwriteNow You Can Build Your Robots Using A Snapdragon Processor


Vasu Madabushi
Inforce Computing

Back in January, Atmel unveiled the brand new Atmel | SMART SAM S70 and E70 families. And if you’ve been waiting to get your hands on the new ARM Cortex-M7-based MCUs, you’re in luck. That’s because both are now shipping in mass production.


With 50% higher performance than the closest competitor, larger configurable SRAM, more embedded Flash and high-bandwidth peripherals, these devices offer the ideal mix of connectivity, memory and performance. The SAM S70 and E70 series allow users to scale-up performance and deliver SRAM and system functionality, all while keeping the Cortex-M processor family ease-of-use and maximizing software reuse.

“As a lead partner for the ARM Cortex-M7-based MCUs, we are excited to ship volume units of our SAM E70 and S70 MCUs to worldwide customers,” explains Jacko Wilbrink, Atmel Senior Marketing Director. “Our SAM E70 and SAM S70 series deliver a robust memory and connectivity feature set, along with extensive software and third party support, enabling next-generation industrial, consumer and IoT designers the ability to differentiate their applications in a demanding market. We are working with hundreds of customers worldwide on a variety of applications using the new ARM Cortex-M7-based MCUs and look forward to mass adoption of these devices.”


These boards pack more than four times the performance of current Atmel | SMART ARM Cortex-M based MCUs. Running at speeds up to 300 MHz and embedding larger configurable SRAM up to 384 KB and higher bandwidth peripherals, the new series offer designers the right connectivity, SRAM and peripheral mix for industrial and connectivity designs. Additionally, the SAM S70 and E70 boast advanced memory architectures with up to 384KB of multi-port SRAM memory out of which 256KB can be configured as tightly coupled memory delivering zero wait state access at 300MHz. All devices come with high-speed USB Host and Device with on-chip high-speed USB PHY and Flash memory densities of 512kB, 1MB and 2MB.

What’s more, the Atmel | SMART ARM Cortex-M7-based MCUs are supported by ARM ecosystem partners on development tools and real-time operating system (RTOS) board support packages (BSPs) accelerating time-to-market. Software development tools are available on Atmel Studio, the ARM Keil MDK-ARM and IAR Embedded Workbench. Operating system support include Express Logic ThreadX, FreeRTOS, Keil RTX, NuttX and Segger embOS. A comprehensive set of peripheral driver examples and open source middleware is also provided in Atmel’s Software Package.


“Atmel has developed a global network of ecosystem partners that deliver hardware and software solutions for the Atmel SMART Cortex-M7 MCU,” adds Steve Pancoast, Atmel Vice President of Software Applications, Tools and Development. “Atmel’s robust, easy-to-use development platform along with our partners’ advanced development platforms offer developers the opportunity to use the best tools and services to bring their designs quickly to market. Atmel continues to expand our partner program to bring the best tools and solutions to our customers.”

Interested? Production quantities of both the SAM E70 and S70 are now available. In order to help accelerate design and to support these devices, an Atmel Xplained development kit is shipping today as well. Pricing for the SAM S70 starts at $5.34 in 64-pin LQFP package and 512KB on-chip flash for 10k-piece quantities while the Atmel Xplained board will run you $136.25. Meanwhile, be sure to read up on the new MCU families here.

For me, the Embedded Systems Conference has been an important date on my calendar for many years. The conference has been given various different names over the years, but I am glad that its identity has finally been restored. The event has moved around the year somewhat and similar events have been organized at other locations around the world.

The Silicon Valley ESC, which takes place next week, is somehow the “real thing” …

Earlier this year, I enjoyed attending the Boston ESC, where I delivered a couple of classes and participated in another interesting session – The RTOS Smackdown.

To read the rest of this entry, visit the Colin Walls blog via Mentor Embedded.


Are you considering making the transition from an Intel-based board to an ARM-based board? While it may seem as simple as swapping one for the other, we all know that things are rarely as simple as they appear.


To that end, I asked two experts at WinSystems what a designer might encounter when making that change. First, I asked, "What are some of the key hardware differences that systems integrators need to consider when migrating from Intel-based boards to ARM-based boards?"


Here's the response I got from Mike Norton, the resident hardware expert:SBC35-C398-Q-468x468[1].jpg


A couple of the key hardware differences when migrating from Intel-based boards to ARM-based boards are associated with power and I/O configuration. In most cases, ARM processors are used when targeting low-power embedded and mobile applications. It's important to understand the power envelope to which you will be constrained.


You may want your design to be powered from battery, solar power,  or possibly Power-over-Ethernet (PoE). Providing sleep states and fast boot capabilities are typically expected for these applications and to further minimize power.


I/O configuration is another area of concern.  ARM chip manufacturers often design processors with a plethora of I/O capabilities and communication standard protocols. Many of these I/O and communication options are mutually exclusive. It's important to understand what features are available and what can be configured to work simultaneously.


Here's the response I got from Patrick Philip, the resident software expert:


When migrating from an Intel platform running some flavor of a Microsoft OS to an ARM platform running Linux, there's a moderate learning curve for software engineers and a paradigm shift when it comes to the way things are done in Linux.


If porting Linux to a new platform, the key software difference is the need to set up an ARM cross-development environment on the workstations where the OS is being developed and images created. Building Linux on an x86 platform for an x86 platform is straight forward. Building Linux on an x86 platform for an ARM platform requires new compilers, libraries, and source code. However, once solved, at the start of the project, this issue is rarely thought about afterwards.


If developing application software, there are minimal key differences. When running Linux, both x86 and ARM platforms will have the same standard C/C++ support and the compilers will be a version of GCC. On both platforms, software will open files and devices using the same API calls and  usually the same parameters.


The cross development host used to build the ARM operating-system image can also be used to build and (remotely) debug the ARM applications. However, if the ARM platform is stable and has sufficient resources to build and debug the application natively, the key application software development differences between Intel and ARM are nearly nonexistent.


Lastly, I asked Philip why, from a purely technology perspective, a developer who is comfortable working with Intel-based products, would consider moving to ARM-based products?


It really depends on the application. If a developer is working with Intel because the product will run Microsoft Windows 7/8/10, ARM should not be a consideration. However, if the product can run Linux or Android, the low power and scalability of ARM processors opens up new applications. It can go in a lightweight multi-core tablet running Android, or into an extremely low power, battery-operated blood pressure cuff. And if you're an embedded software or firmware engineer, acquiring ARM skills lets you work with a diverse set of ARM devices, from multi-core SoCs, to low-power MCUs.

Kicking off this week with news on the availability of Cortex-M7 dev boards from Atmel and STMicroelectronics.  The Cortex-M7 is the current top of the line M series microcontroller but its also got intriguing DSP capabilities that could be a game changer.


Atmel has a Cortex-M7 development board out that is available now (for $136) with an automotive grade version of their SMART SAMV71 processor.  STMicro has also got their version of the Cortex-M7 on the market (STM32F7) which also has a dev kit ($50) and many other resources here. If you are involved in audio DSP designs then check out Audio Weaver software (free download) which runs on the Atmel board and takes advantage of the DSP functions of the Cortex-M7 (ask Chin Beckmann for more info).  Audio processing is going through a renaissance right now with new MEMS microphones and audio devices like the Amazon Echo changing the way we can interact with everyday devices.  Audio might just give the IoT a boost because a natural interface like audio might overcome user fear of installing and using new devices.


Talking about intuitive design and user barriers I noticed that the electronics education (or toy) company littleBits just got $44m in financing.  littleBits is a fascinating company which has developed very simple magnetic computing blocks based on the Arduino.  They seem excited:


This is what the littleBits products looks like:


The base kit is $99 on Amazon


ARM CEO Simon Segars was the keynoter at last month's IMEC Technology Forum in Brussels. If you don't know IMEC they are an impressive EU funded research organization that is always doing interesting projects (follow them on Twitter).  Simon challenged the semi industry to address three key issues; power, memory and packaging.  You should read electronics journalist extraordinaire Peter Clarke's writeup on Simon's speech.


Finally this week I was struck by just how many prize challenges are out there right now, some with very significant prizes of $1m and more.  These challenges can be very profound like the X-Prize and some are more mundane big data hacks like finding the optimal price for a hydraulic hose to fit a Caterpillar bulldozer.  I wrote about this over on EETimes today.


That's it for this week, please follow me to get notifications when I post if you find this interesting.  If I'm missing the mark on anything please let me know.

How can you introduce modularity in embedded development to reduce time-to-market and development cost?

Check out my blog that introduces Computer on Modules and how can they can be included in embedded development.



Up until now the power of Software Defined Networking (SDN) was only available to the administrators of large corporate networks like Google and Facebook. However, one Australian company has shrunken those capabilities down to a palm-sized form factor of just 10cm x 8cm to create what they’re calling the world’s smallest OpenFlow-capable switch.


With aspirations of getting SDN into the hands of Makers, students and hobbyists, the Zodiac FX is the first OpenFlow switch meant to sit on your desk, not in a data center. The idea was initially conceived by Northbound Networks founder Paul Zanna after finding that there was a persistent gap between SDN controllers and simulation software and OpenFlow-capable hardware.

What’s nice is that the Zodiac FX packs many of the features of an OpenFlow switch all for a fraction of the cost and size. Based on an Atmel | SMART SAM4E Cortex-M4 MCU, the board includes four 10/100 Fast Ethernet ports with integrated magnetics and indicator LEDs along with a command line interface accessible via USB virtual serial port. Aside from all that, the Zodiac FX is equipped with the layer 2 and 3 switching capabilities of the Micrel KSZ8795 Ethernet controller.


Designed with the SDN development and Maker communities in mind, the Zodiac FX firmware is completely open source. This means that anyone can download its code and use Atmel Studio to produce their own custom version. From there, it can be reloaded onto the board via USB.

"The Zodiac FX firmware utilizes the Atmel Software Framework (ASF) for generic device drivers such as USB, SPI, etc. On top of this it then adds a custom written driver for the KSZ8795. FreeRTOS is used to provide task and memory management for the three core processes; Command (CLI), Switching and OpenFlow,” the team writes.

It should be noted that, although the Zodiac FX may be the company's first foray into the hardware world, Northbound Networks has been extensively involved with SDN development utilities. Interested? Head over to Zodiac FX’s Kickstarter page. While delivery for the beta version is expected to kick off in October 2015, the final units aren’t slated to ship until January 2016.

This blog originally appeared on Atmel Bits & Pieces.

Filter Blog

By date:
By tag: