Posted by: eschorn | May 12, 2009

ARM in FPGA

Only a few years ago, if you were to run across “ARM” and “FPGA” in the same sentence, you would invariably find the world “prototype” as well. Soft/embedded processors were only begining to ramp in production popularity and ARM was not participating in that space…yet.

All this changed on March 19th, 2007 (my birthday BTW) when ARM announced the Cortex-M1, the first tier-1 vendor independent processor specifically optimized for volume production on FPGA. The tremendous applicability of the ARM Architecture was expanded to include the huge number of FPGA-oriented design teams out there.

The Cortex-M1 offers FPGA vendor independence, IP reuse, flexibility, ecosystem, and the perfect entry to a future-proof roadmap.

The applicability of the ARM Architecture was expanded even further through additional business models as well. There are now three ways to get your hands on the Cortex-M1

  1. ARM and Actel have executed a license that enables free access, use, and production on Actel FPGA silicon.
  2. ARM, Arrow, and Altera have developed a low cost Cortex-M1 development kit for Cyclone III devices.
  3. An FPGA design team can license the RTL directly from ARM in the traditional model

I was excited by the reception from the FPGA Journal

ARM and Altera
Why You Should Care
by Kevin Morris, FPGA and Structured ASIC Journal

The ARM Cortex M1 is out to conquer the world.

Stealthily.

Cleverly.

Kindly.

It will probably be one of those rare conquests where the conquered smile and rejoice and say “Thank you! Thank you for conquering us!”

Electronic designers will flood the streets in celebration, now having access to exactly what they have needed all along. Even those who weren’t aware that they needed anything different will be excited and appreciative. The earth will simply be a better place.

OK, maybe it’s not as big as all that.

Agreed – it takes time for a new entrant to become established. I understand Actel has over 100 design wins, Arrow has been shipping Altera kits,  and we have done additional licenses, but I am certainly not ready to claim victory in any respect. The size of the remaining prize remains large.

BTW, space precludes me from describine Altera Excalibur and shipping ARM7, ARM9 on FPGA…watch this space!

Possibly related posts: (automatically generated)

Posted in ARM, FPGA

«
»

Responses

  1. A great addition to the armh website your insights are very informative and much appreciated
    Thanks in advance
    vic enright

    By: vic enright on May 13, 2009
    at 7:42 AM

    Reply

  2. Hi,

    ARM is an embedded processor. Embedded processors execute only dedicated programs. I don’t see why an FPGA should be programmed this way. I would use a tool to convert C code into gates (like http://www.c-to-verilog.com). Arguably this will be much more efficient then the ARM design (on FPGA).

    By: Adam on May 15, 2009
    at 6:15 AM

    Reply

    • Adam,

      I understand that current c-to-verilog technology is very limited. The URL you mention is near a FAQ that suggests:

      You may use most of the C language features except for a limited list of features including recursive functions, structs, pointers to functions and library function calls (printf, malloc, etc). These cannot be represented in hardware. You MAY define multiple functions as long as you declare them as ’static inline’. You MAY use any of the standard types (array, pointers, int) but may NOT use float/double types. You may not declare global variables and arrays which are local to the function. All of your arrays must be declared as parameters of your function. Only one-dimensional arrays are supported.

      A lot of C-code uses this stuff.

      Regards, Eric.

      By: eschorn on May 15, 2009
      at 6:36 AM

      Reply

  3. Hi Eric

    After reading your comments, knowing that ARM’s architecture will be at the core of IBM’ Common Platform for 32nm & below and reading the following paper from IBM

    http://download.boulder.ibm.com/ibmdl/pub/software/rational/web/whitepapers/SmarterProducts-whitepaper-051909.pdf

    it becomes quite apparent that when LTE becomes a reality worldwide users will have smart electrical devices with more intelligence, real-time instrumentation and interconnectivity.

    This then will mean that software becomes the most active ingredient in product innovation and differentation in these devices.

    So my question is this:

    At what point in time do you envision devices already deployed in the marketplace with integrated processors to be able to just download new software in pico-seconds, to become future-proof, as opposed to having to develop new ASP’s
    for each new application ?

    By: Garrett Lackey on May 29, 2009
    at 12:34 PM

    Reply

    • Thanks for the comment Garrett – that is a particularly tough question! I wrote earlier how everything with a screen and everything without a screen will ultimately become connected but do not think one particular technology will provide the sole unifying (connectivity) factor. I also suspect this is a trend rather than a discontinutity. My take is that first we will see bug fixes (already apparent), then we’ll see some minor functionality upgrades, then some paid upgrades…but I don’t see any silver bullets with respect to single silicon ASSPs covering huge application spaces. I think the forces of innovation are stronger than the forces of convergence, so life will remain incredibly interesting. You have hit upon some controversy ;) Regards, Eric.

      By: eschorn on June 2, 2009
      at 12:43 AM

      Reply


Leave a response






Your response:

Categories