Let's say the convex constrained quadratic problem reads
\begin{align} \min_{x \in \mathbb{R}^n, \ t \in \mathbb{R}} \ & \ t\\ \text{s.t.} \ & \ x^TA_ix \leq t + b_i \quad \forall i = 1,\ldots,I\\ \ & \ \| x - z \|_2^2 \leq \delta \ , \end{align} where symmetric (i.e., positive semidefinite) $\{A_i \in \mathbb{R}^{n \times n}\}$ and $b_i, \delta \in \mathbb{R}$ are known.
What is the worst-case computational complexity of such a problem, for instance, using interior-point methods?
One of the article that I am reading states that the complexity of convex QCQP is approximately $O(n^{4.5})$. But I am not sure whether the above problem would also have the similar complexity?