We are expected to solve a conjecture of the title. Reference is Jean-Pierre Serre — Propriétés conjecturales des groupes de Galois motiviques et des représentations l-adiques. 
Precisely;

**Conjecture A**:  
k:an algebraic number field   
v:a non archimedean place of k   
E:a motive with good reduction at v 
For E, we have a corresponding unramified l-adic representation. The image of the decomposition group of that representation is topologically generated by a Frobenius element.  
The conjecture is that the characteristic polynomial of that Frobenius is ℚ-coefficients and if E is pure of weight i, its roots in ℂ has absolute value Nv^i/2. Here Nv is the number of elements of the residue field of kv.

We also have **Weil conjecture**;  
F:a finite field of q elements   
X:a smooth projective variety over F   
The characteristic polynomial of frobenius of its l-adic cohomology of degree i has absolute value q^i/2.

In addition we have two **standard conjectures** on algebraic cycles;   
**Lefschetz standard conjecture** asserts that an abstract analogue of the Λ-operator of Hodge theory is induced by an algebraic cycle on two products of smooth projective varieties over an algebraic closed field.   
**Hodge standard conjecture** asserts that there is an abstract version of the Hodge index theorem for ℚ-vector space of classes of algebraic cycles.

I have the following three questions.     

1.Conjecture A and Weil conjecture both concern the roots of weight of a characteristic polynomial. How are these two related?    

2.Weil conjecture is proved under the assumptions of Lefschetz and Hodge standard conjectures. Is the conjecture A also proved using them?

3.Is the conjecture A proved for abelian varieties over any number field? 

Thank you for your answers.

Edit:motives are defined by numerical equivalence of algebraic cycles, good reduction of motives are defined by the corresponding l-adic representation is unramified