Array storage in an FPGA, How is it handled?

asdf44 said:

By default assume it's stored in a lut flip flop.

However if your FPGA has block ram the tool may fit the array to ram if the way you implement and use it is compatible with that. See the output of your tool - it will print a message telling you when it infers a block ram.

Click to expand...

FPGAwarrior said:

I am using Quartus Prime. My demo board has a CPU, RAM and an FPGA. Where do I look to find this info? Also, I did not start my code from scratch. It is built on an existing code.

Click to expand...

ads-ee said:

Synthesis report will show how it implemented an array in the code.

Using existing code doesn't guarantee anything, even proper functionality. Unless a piece of code has a regression suite that I can re-run and check for coverage, I always assume that existing code I have to use doesn't necessarily function correctly.

The coding of an array to be implemented in RAM is usually called out in the vendors synthesis documentation. This usually involves registering the correct port, not having resets on the array, and having the read and write signals properly implemented.

Click to expand...

FPGAwarrior said:

Thanks. I saw in a report what appears to be an auto RAM system that utilizes a few hundred KB. Apparently, it does use part of the CPU RAM.

Click to expand...