It has long made sense to combine a processor with an FPGA on a single piece of silicon. It’s the law of integration and it was inevitable. Embedded designers are accustomed to working with single-chip microcontroller units (MCUs) that combine non-volatile memory, subsystem, and analog blocks in a single chip. The next logical step was to fully merge an MCU and an FPGA, but this first required that FPGAs move to the MCUs’ flash-based process technology. Now that this has occurred, the industry has a customizable system-on-chip (cSoC) platform that delivers greater flexibility than traditional fixed-function microcontrollers, without the excessive cost of soft processor cores on traditional FPGAs.
Today’s cSoC platforms with flash-based FPGAs are being used by designers to create a wide variety of products with the features and capabilities they want, without having to compromise with off-the-shelf ASSPs and microcontrollers. These cSOCs integrate high-performance programmable logic, microcontrollers, programmable analog, large nonvolatile memory blocks, and comprehensive clock generation and management circuitry into a single, monolithic device. This enables embedded designers to optimize hardware and software tradeoffs on the fly, without board-level changes.
A key target application for these cSoCs is the industrial "sense and control" market. For instance, motor control systems must deliver high performance with outstanding power efficiency and a long lifetime in a small, low-cost design for high-volume applications. Integration and cost-efficiency are significantly more important in these applications than processing power and bandwidth. And yet, previously, there was no complete, customizable SoC solution for this market. Instead, most systems were built using a microcontroller, an FPGA for logic, and some discrete analog components. Without an integrated solution, hardware and software developers had to work on different platforms, and once your PCB was layed out, changes between the interfaces of these blocks were no longer possible.
Thus, the next logical step was to unify this architecture. Using this approach, designers no longer have to settle for an FPGA with a companion external microcontroller. Instead, they can use an industry-standard ARM processor with its well-known architecture and popular instruction set. This is important, because the more popular the processor, the easier it is to find software, tools, and talent (the so-called “ecosystem”) for it. Studies show that most engineers choose their operating system and code-development tools first, before deciding what processor to run them on. With their companion software, middleware, operating systems, debuggers, and hardware and software development tools, ARM processors are the de facto choice for developers around the world. It’s easy to hire programmers with ARM processor experience, and easy to get good support, software and service for ARM-based chips. In short, the ecosystem surrounding ARM CPUs is unparalleled.
Among ARM’s offerings, the ARM Cortex-M3 is an ideal processor for cSoCs. It is a powerful and economical implementation of the ARM architecture, and is available in a wide variety of settings. It also has been the platform upon which many designers have advanced from 8-bit to 32-bit architectures. As an example, a hard 100 MHz 32-bit ARM Cortex-M3-based microcontroller subsystem can be combined with 128 Kbytes of embedded flash, 16 Kbytes of embedded SRAM, plus programmable analog including one ADC, one DAC and two comparators in to a single cSoC as shown below:
Microsemi SmartFusion A2F060 cSoC
In this cSoC device, the processor is a hard core. This means that it is implemented in silicon as ASIC gates and is powerful enough to run complex algorithms, readily running precision motor control, or even multi-axial control of several motors. In an application such as system management, it can supervise all of the voltage monitoring, sequencing, fan control and associated "system housekeeping" tasks, while having ample capacity to run a higher, user-application-level task as well.
cSoCs are quickly being adopted in system and power management and industrial automation applications, across a variety of market segments spanning the industrial, military, medical, and telecommunications markets.
Today’s cSoC solutions improve flexibility while reducing cost, and offer a compelling answer for a wide variety of applications. By combining an ARM core, embedded flash memory and FPGA into a single, monolithic device, these solutions enable designers to avoid the compromises associated with using separate, off-the-shelf ASSPs and microcontrollers. Instead, designers can create the products they want with the features they need, without compromise. About the author
Rich Kapusta is Vice President of Terrestrial Products in the SoC Products Group at Microsemi Corporation
. Rich joined Actel (now Microsemi) in 2008, bringing with him more than 15 years of engineering, marketing and management experience.
Prior to joining Actel, Rich was vice president, business development for Exar Semiconductor and product line vice president and general manager at Sipex Semiconductor. Previously, Rich held strategic marketing, engineering and management positions at Cypress Semiconductor. He holds a bachelor’s degree in computer engineering from University of Illinois, Urbana-Champaign.Rich can be contacted at Rich.Kapusta@microsemi.com
If you found this article to be of interest, visit Programmable Logic Designline
where you will find the latest and greatest design, technology, product, and news articles with regard to programmable logic devices of every flavor and size (FPGAs, CPLDs, CSSPs, PSoCs...).
Also, you can obtain a highlights update delivered directly to your inbox by signing up for my weekly newsletter – just Click Here
to request this newsletter using the Manage Newsletters tab (if you aren't already a member you'll be asked to register, but it's free and painless so don't let that stop you [grin]).