A relativistic collision between two particles is generally going to lose energy as some of the energy of the two particles goes into creating secondaries (e.g., the useful $p+p\to p+p+\pi^0$ reaction). So anything starting out with "what if they collide..." ought to fail on that principle alone.
Instead, what you want is something to accelerate a particle from some thermal speed (think Maxwell-Boltzmann) up to the ultra-relativistic speed (e.g., $v\approx c$). There is a class of neutron stars with very strong magnetic fields, called magnetars, that have magnetic fields of $\sim10^{10}$ T, which is quite powerful. From the Lorentz force, $\mathbf{F}\propto\mathbf{v}\times\mathbf{B}$, we can see that a particle can be accelerated due to the magnetic fields; hence this part of your question is quite reasonable: objects with very large magnetic fields can (and probably do) accelerate particles.
The question is then whether the neutron stars (including magnetars) are (dominant?) sources of (near) GZK-limit cosmic rays. To this question, I don't have an answer. When I was researching cosmic rays 8$^+$ years ago, the expectations were that Active Galactic Nuclei were the primary contender for the production of GZK-limit particles. Having taken a non-academic job, I've lost track of where the research is going since then.