feature article
Subscribe Now

Fusion Finds its Groove

Actel Sees Applications for FPGA/Analog Hybrid

Many times, we create a new technology with the knowledge that it has tremendous potential but without a clear idea of how it will ultimately be used. Or, we create a solution to one problem only to find later that our invention was even more useful in some completely different area. We throw our new creation out to the public and are amazed when someone comes up with a far more clever way to use it than we had ever envisioned.

This may be the case with Actel’s “Fusion” family of analog-capable FPGAs.  Sure, the company had a pretty good idea that engineers would find the combination of programmable digital fabric and real analog I/O useful, but we always suspected that the “killer apps” for Fusion would be discovered by the engineers that started using the devices, not by the factory.

Fusion has been around for awhile now, and it seems to be finding a unique groove in the industry. Applications such as smart motor control take advantage of the family’s particular combination of features and properties and provide solid economic justification for their design-in.  Happily, this kind of application is also directly in line with Actel’s power-centric mantra of the past couple of years.

For those of you that haven’t been tracking Fusion carefully, let’s take a brief look at where these devices come from.  Actel produces two major categories of programmable logic – anti-fuse style one-time-programmable FPGAs for use in space, mil/aero, and other high-reliability applications where extreme radiation tolerance is important; and flash-based FPGAs for low-cost, low-power applications. Fusion is a variant of the latter – the flash-based FPGA technology that includes ProASIC3, Igloo, and Fusion.  Each of these families is a non-volatile, in-system-re-configurable FPGA.  Unlike other FPGAs and CPLDs that use flash, the Actel devices use flash cells for the actual configuration.  Other vendors’ architectures utilize SRAM configuration cells that can be re-configured from on-chip flash memory. Actel’s approach has both advantages and disadvantages, but one of the biggest things in the “plus” column is extremely low static current. 

SRAM-based FPGAs (and SRAM FPGAs configured from on-chip flash) have to keep all those configuration SRAM cells live all the time (or they end up losing their configuration and have to be re-configured).  Configuration cells account for the majority of the transistors in any FPGA fabric, so that means most of the transistors on a device are typically active and leaking current.  At today’s tiny geometries, this leakage is even more problematic.  Because Actel’s flash-based FPGAs use flash cells for the actual configuration, there is no “keep-alive” power required (except for registers and memories, of course).  This means that these devices have extremely low standby current and very little leakage. Standby is typically in the range of microwatts compared with milliwatts for an SRAM FPGA.  

Actel has made this property of flash one of the cornerstones of their identity, pushing hard with their “Power Matters” marketing and targeting applications and sockets where power is at a premium.  On top of that, however, they are now pushing their devices into applications where much more than the power consumption of the FPGA is at stake.  Applications like system power management and motor control promise to save far more energy than just the difference in FPGA power consumption.  Since Actel’s devices occupy the lower density ranges (compared with the big SRAM FPGAs), they fit nicely into these types of applications.  You will often today see an Actel device doing the power management on a board that also boasts big Virtex or Stratix devices. 

The second corner of Fusion’s heritage is the ARM processor connection that Actel forged a few years back.  By building devices with an ARM core license included in the price of the part, they allow designers to pull the processor or MCU function onto the FPGA, saving cost, space, and often power. The combination of these features gives a very inexpensive device with a capable processor and a lot of very low-power programmable fabric for – whatever other digital functions you need.  If you’ve looked at many micro-controller catalogs, you’ll appreciate this “build-your-own” MCU.  Often, a catalog will have hundreds or even thousands of variations of controllers and peripherals – and sometimes none of them has exactly the combination you need.  

The final (and most differentiated) post in Fusion is the inclusion of programmable analog.  By offering even a modest amount of analog capability on the Fusion device, Actel has enabled a host of interesting applications in the “control” genres, giving the family a unique place in the designer’s catalog.

Now that customers are jumping on the Fusion bandwagon, Actel is doing what many other FPGA companies are also starting – building pre-engineered “almost done” applications that get design teams past the FPGA-fear-factor.  By including domain-specific IP, reference designs, and carefully-crafted development boards, many designers can get to a working application that does 90% of what they need without ever writing a line of HDL or messing with a single synthesis or place-and-route option.  For a design team whose mission may be something like “motor control” and whose expertise lies in areas like analog signal optimization, motor performance and efficiency, and software control algorithms – this can make the difference between designing in an FPGA and steering clear.

Given the very low price-point of Fusion, it’s clear that Actel will need to win a lot of sockets and a number of very high volume ones to really score a home run with the family.  However, their experience so far with the family seems to show a very wide range of target applications, and with their fine-tuning of development kits and tools, these unique, mixed-signal FPGAs are likely to capture a large number of design-ins where FPGAs have not previously landed.

Leave a Reply

featured blogs
Aug 17, 2018
Samtec’s growing portfolio of high-performance Silicon-to-Silicon'„¢ Applications Solutions answer the design challenges of routing 56 Gbps signals through a system. However, finding the ideal solution in a single-click probably is an obstacle. Samtec last updated the...
Aug 17, 2018
If you read my post Who Put the Silicon in Silicon Valley? then you know my conclusion: Let's go with Shockley. He invented the transistor, came here, hired a bunch of young PhDs, and sent them out (by accident, not design) to create the companies, that created the compa...
Aug 16, 2018
All of the little details were squared up when the check-plots came out for "final" review. Those same preliminary files were shared with the fab and assembly units and, of course, the vendors have c...
Aug 14, 2018
I worked at HP in Ft. Collins, Colorado back in the 1970s. It was a heady experience. We were designing and building early, pre-PC desktop computers and we owned the market back then. The division I worked for eventually migrated to 32-bit workstations, chased from the deskto...
Jul 30, 2018
As discussed in part 1 of this blog post, each instance of an Achronix Speedcore eFPGA in your ASIC or SoC design must be configured after the system powers up because Speedcore eFPGAs employ nonvolatile SRAM technology to store its configuration bits. The time required to pr...