Can i use a single ALM(ALM#0) to check 8 independent bit for zero-equality and move this result via COUT output to the next-bottom ALM(ALM#1). Then check another 8 bits in the ALM#1considering CIN from ALM#0 and move this result via COUT to the next bottom ALM(ALM#2) and so on.
Suchwise totally i want to reach utilization of 8 bits per ALM.
When i am checking 4 bits per ALM it works perfect. But it doesnt work with 8 bits. But according to the handbook ALM supports up to 8 independent inputs.
For example Xilinx's LUT has a 6 independent inputs, and i am free to use any truth table for this LUT and then move a result to the next LUT(via MUXCY/CARRY) and then to the next slice and so on.
What primitives/megafunctions should i use to implement 8 bits per ALM?
I am attaching a picture of needed implementation. But it works with 4 bits per ALM only.
Thanks in advance.