how to fork parallella board?

[roberto]

Hello, i have and idea for a board, an initially i considered to startr form a "bare metal" FPGA board... but why redo the wheel? Because Parallela is open, it is smart to use it as base for another project.

What i need is just a board with all but without E16/E64 chip, because i want to do else: i wish eliminate from the FPGA the part dedicated to controll/interface E16/E64 and use these logic gates to do other things (maybe it is enough the internal logic gates, maybe i will need another external fpga to take my project, I still don't know).

First of all I want to be sure that this way is totally legal: can I modify the "code" inside the FPGA? I mean, i dont want to touch the VERILOG(?) of ARM core, i just want to add some code of my own.
Have I to pay the fee to ARM for their IP? Can i use - if any - another ARM softcore instead of what is inside FPGA now? In FPGA there is a dual core, but still i don't know if all that power is needed by me.

If uder legal viewpoint all is ok, parallela can be a good "bare metal" to start then modify the fpga programming and if I get success, rebuild the phisical board deleting the ephipany core and add the hardware i want.

Well, I am a _TOTAL_ newbe in FPGA programming, use of tools and so on. So, how to do? where i have to start? Someone of the Team can redirect me to the right "point of start"?

Thanks.

[9600]

[JohnR]

Hi,

If you are interested only in the ARM chip, would it not make more sense to buy a board such as one of the Xilinx Zynq boards for instance, http://zedboard.org/product/picozed.

JohnR.

[roberto]

[roberto]

[9600]

Sounds like you need to start with the requirements and work your way up. It may be that a simple board with a small FPGA will suffice for your particular application. Or perhaps you need something slightly more capable so that you can host a soft processor core and maybe run an IP stack, real-time O/S or even full blown Linux. But requirements must come first...

In any case, there are far simpler low cost FPGA development platforms available with support materials for novice projects. One of these would probably be good for getting started. Then once you learn more about the tools, what is possible and at what cost, effort and with all the various design considerations, you can start to think about what your ideal design might look like.

Regards,

Andrew

[roberto]

Andrew,
i have 2 mailine of study:

A. to implement some AND-NOT-OR-XOR-SHIFT-etc network with FPGA, to replicate in hardware what i did already in software to compare the speed between a CPU and a FPGA, just to sutisfy my curiosity.More general, i wish to replicate in fpga some programs i did in softwareand see how much speed up, and use fpga as "coprocessor".

B. to implement a softcore able to run full GNU/Linux OS (GUI not needed, Command line is enough), with support to ehernet, usb, and a rs-232 for remote console.

I wish to start with point A to get experience, then move to poit B for the real project.
The good thing is to start immediately with a fpga able to serve both points, to avoid to change fpga after archived point A because that fpga is not powerfull enoug to sustain point B.

Do you have suggestion about model of fpga to start with? a demoboard?
To give you a meter of paragon, The Point B should be at least equivalent to a real ARMv7 ARM core running at 5-700Mhz.

THanks.

[9600]

[roberto]

Andrew,
you eliminate many wrong ideas i had over FPGA, you evitate me a future delusion because of my "to be a newbie". Despite I drived you Off Topic, your hits clear my view. thanks a lot.