What is a disadvantage of using stochastic gradient descent $($SGD$)$ over batch gradient descent?

a$.$ SGD requires more computational resources and may not be suitable for large datasets

b$.$ SGD is less accurate in estimating the gradient

c$.$ SGD is less stable and may oscillate or fluctuate around the optimal solution

d$.$ SGD is slower to converge to the optimal solution