I think delt0r and I were trying to be polite about the idea.
Confining a charged ion beam in a large aspect ratio ring is one thing. But once you introduce a second beam at a different velocity it will be unstable.
http://en.wikipedia.org/wiki/Two-stream_instability
Here is an animation – note the y axis is velocity, so you start with the beams being two lines at different y positions. Then through wave-particle interactions it goes unstable and you quickly get a mix in velocity space (this doesn’t require any large angle collisions).
http://www.youtube.com/watch?v=ZzJFvYVVWPw
This, with collisions, will settle down to a Maxwellian thermal equilibrium.
Also, if your two beams had a deltaE of 30keV this would be the approximate temperature they relax to. Given your ridiculously high estimate of density of 10^23m^-3 this would equate to a pressure of 4.7*10^8 Pa or 4700 atmospheres. I’ll leave you to figure out the magnetic field you need to contain that at a beta of less than 1.
Even if the pressure was highly anisotropic so the radial part was only a few percent of the total, the required toroidal field would be impossible to create on that scale.