I am using Quartus Prime Pro edition V19.1.0 targeting an Arria-10 FPGA.
I'm trying to create a single precision floating point Multiply-Accumulate IP block, using the floating point functions in the IP catalog.
The function I am trying to create is: output = accumulator + (input_a * input_b)
 I am unsure whether I should use "Multiply Add" or "Multiply Accumulate"
The latter sounds right, but the 'acc' port is only 1 bit.
 In general, I can not find any documentation for this IP. The Floating-Point IP Cores User Guide (UG-01058) doesn't actually specify which input and output signals are which. Is there a document I have missed?