A new technique is presented for partial pole
placement of linear time-invariant systems. It is almost always possible to arbitrarily assign
$\min(n, \varphi)$ poles using this
new method. Here $n$ is the order of the system, and
\begin{equation}
\varphi \triangleq \max(m,\ell) + \lfloor \frac{\max(m,\ell)}{2} \rfloor +
\ldots +
\lfloor \frac{\max(m,\ell)}{\min(m,\ell)} \rfloor
\notag
\end{equation}
where $m$ and $\ell$ are the number of inputs and outputs, respectively, and
$\lfloor . \rfloor$ denotes the nearest integer
lower than or equal to (ie. floor($.$)). Only the normal procedures of linear algebra are required to implement the
technique. We note that $\varphi \geq m+\ell-1$, which has been a long-standing barrier for linear
algebra methods in the partial pole placement problem.