ndmem.md

• Research (investigate) and understand current mem conversion (?? days), will document on daily blogs.
• Research (investigate) and understand Verilog and VHDL Ndarray support (?? days). Is the NDarray support, will document in daily blogs.
• Write a MEP for the 2D/NDarray conversion support (?? days).
• Get feedback update MEP (??days).
• Investigate the converion/general tests structure (?? days), will provide documentation on daily blogs.
• Write tests for the converstion support (?? days), will document provide status on blogs and have a complete set of tests.
• Start implementing the 2D/NDarray conversion support (?? days), blog status
• Prepare pull-request for this new and awesome feature (?? days), blog status
• Create documentation: release (what's new) and manual (?? days)
• Summer over, proad of the work accomplised :)

Above, NDarray is used to indicate a list-of-signals, or rather a list-of, list-of, list-of, ..., list-of-signals. Not to be confused with the numpy.ndarray (but similar).

xmem.py

# a list-of-signals (1D list)
mem = [Signal(intbv(0, min=0, max=256)) for ii in range(8)]

# a 2D list
mem2d = [[Signal(intbv(0, min=0, max=256)) for ii in range(8)]
         for ij in range(8)]
   
# a 3d list      
mem3d = [[[Signal(intbv(0, min=0, max=256)) for ii in range(8)]
         for ij in range(8)]
         for ik in range(8)]
         
# probably would want some type of class to create
# larger arrays