Difference between process and concurrent statements

Question

when I assign Signal F the complement of Signal B in a process like this:

architecture V1 of E is 
 signal B: bit; 
begin 
 process (A) begin 
  B<=A; 
  F<=not b; 
 end process; 
end architecture; 

then F will hold the old negated value of B in the first cycle.

But why doesn't the same hold for this example where I'm just using concurrent statements?

architecture V1 of E is 
 signal B : bit; 
begin 
  B<=A; 
  F<=not b; 
end architecture; 

When I'm thinking of the Hardware components F will receive the signal from B so in the first cycle F should also hold the old negated value of B?

Answer 1

Concurrent signal assignments are just shorthand for processes. Your second code is the same as:

process(A)
begin
  B <= A;
end process;

process(b)
begin
  F <= not b;
end process;

The reason why the behavior differs is that with your first code the process resumes only when signal A changes while with the second each process resumes each time the signal in its sensitivity list changes.