== vs >= : area wise

[suquid29]

Hi
I was wondering what is the different, area wise, between >= and ==.

Lets say I want to start switch state in FSM (verilog, asic) with the condition cntr_t >= 4'h8.
(the counting is with a bit faster clock, so in the edge I'm checking it, it sometimes will be 8 and sometimes 9; there is sync).

Isn't it better to check only one bit with == ? (e.g cntr_t[3]==1'b1)

In general, how can I estimate area of different implementation of verilog code?
Is there some online tool?

I have synthesis tools, but its a big core ...

Thanks!

 

[morris_mano]

'==' can be implemented with an XOR gate between each corresponding bit of two values we are comparing. If we are comparing 3-bit values . Then we can do that by 3 (2 input XOR) and 1 (3 input NOR). If the output of the NOR is 1 means, those two 3-bit values are equal else not equal.

'>=' implementation need more gate than the implementation for '==' I described above. As the implementation output needs to be true for both '==' and '>' condition.

 

[FvM]

A synthesis tool can be expected to remove redundant bits from a comparison. In so far both expressions (cntr_t >= 4'h8) and (cntr_t[3] == 1'b1) should end up in the same gate level logic. I would prefer the comparison for it's better readability.