This is something we’re still working on. Operator splitting methods are in a very rudimentary state:
http://docs.juliadiffeq.org/latest/solvers/split_ode_solve.html
It’s one of the next big sets of algorithms we are working on though. Until we have those kinds of methods, I would recommend just using the SSP methods
with a sufficient limiter. The other types of methods which would do well are the IMEX schemes which we don’t have yet:
https://github.com/JuliaDiffEq/OrdinaryDiffEq.jl/issues/191
So yeah, probably not satisfactory but it’s what we got. You can also just try throwing it into Sundials’ CVODE_BDF
which might do well enough.