What are non-Abelian plasmas? I know what a non-Abelian group is: a group whose operation does not obey commutativity.  Does this have anything to do with a non-Abelian plasma?  I'm not really looking at a specific paper, but I'll give an example here: http://arxiv.org/abs/1211.0343
 A: In Quantum Field Theory certain particles/fields are actually different `components' of the same field. Examples would be the electron and the neutrino or the three different colored quarks. Such a multiplet is often called a Yang-Mills multiplet as Yang and Mills originally thought the positron and neutron were two components of the same field (an idea that turned out to be wrong, but the mathematical approach was gold).
$$ \begin{pmatrix} e^{-}\\ \nu_e \end{pmatrix}, \begin{pmatrix} q_r\\ q_g \\q_b \end{pmatrix}$$
In quantum mechanics, when we have the Dirac equation, we would expect the wavefunction of the electron to be invariant under a phase rotation$\psi \rightarrow \psi e^{i\phi} $. These phase rotations are unitary complex numbers $e^{i\phi}$ and form the group $U(1)$. This property of the Dirac equation being invariant under a phase transformation is called global phase invariance. 
$$ i \hbar \gamma^\mu \partial_\mu \psi - mc \psi = 0$$ 
Note furthermore that the Dirac equation is NOT invariant under a phase shift that varies from point to point $\psi \rightarrow \psi e^{i\phi(x^\mu)}$. Using the product rule on the first term we find that inserting $\psi \rightarrow \psi e^{i\phi(x^\mu)}$ in the Dirac equation gives us no longer zero:
$$ \left[ i \hbar \gamma^\mu \partial_\mu - mc \right] \psi e^{i\phi(x^\mu)} =  e^{\phi(x^\mu)}\left[ i\hbar \gamma^\mu \partial_\mu \psi - \psi \hbar \gamma^\mu \partial_\mu \phi(x^\mu) - mc \psi \right] \neq 0$$ 
However, we can make the Dirac invariant under such a 'local' phase transformation by introducing some additional term that 'absorbs' the extra term we get. Note that this extra term $\partial_\mu \phi(x^\mu)$ is the gradient of a scalar field $\phi(x^\mu)$ and thus behaves like a vector field. We could thus hope to make the Dirac equation invariant under local phase transformations by inserting some sort of vector field. However, since we wish to be able to cancel an arbitrary phase transformation this should be a vector field which has a certain gauge freedom of the following form
$$i \hbar \gamma^\mu \partial_\mu \psi - \frac{e}{c}\gamma^\mu A_\mu - mc \psi = 0 $$
$$A_\mu \rightarrow A_\mu - \frac{\hbar c}{e} \partial_\mu \phi(x^\mu) $$
Notice how miraculous! We found that if we wish the electron wavefunction to be completely invarient under an arbitrary phase transformation $\psi(x^\mu)$ this is only possible when we introduce a field that behaves exactly like the EM-field! Any arbitrary $U(1)$ phase-transformation of the electron field can now be compensated by a corresponding gauge transformation of the EM-field! \ \
In Quantum Field Theory the $(e^-, \nu)$ and quark multiplets are not only invariant under phase transformations, but they are also invariant under certain groups of (unitary) matrices (with determinant 1) that mix the components of the fields in a specific way. 
The $(e^-, \nu)$ multiplet is (at high energies) also invariant under the non-Abelian group $SU(2)$ while the quark multiplet in also invariant under the non-Abelian group $SU(3)$. Hence the name non-Abelian gauge fields / non-Abelian Yang-Mills theory.
Just as we had to introduce a gauge field (the Electromagnetic field) to make the Dirac equation truely invariant under these phase transformations we need to introduce other gauge fields (the $W^\pm$ and $Z$ gauge fields in the case of the $SU(2)$ transformations of the $(e^-, \nu)$ to make this multiplet truely invariant under these $SU(2)$ transformations. The gluons (which are actually 8 different gauge fields) are the fields that make the quarks invariant under these $SU(3)$ transformations.
So to give a simple answer to your question: the quark-gluon plasma is typically a non-Abelian plasma (I hope you can put some more terms in context now). The paper you mentioned studies more general supersymmetric versions of these non-Abelian gauge theories with larger multiplets and larger gauge groups.
