I think the virologist will give you a better answer.  But from an ecology point of view, an different strain would be using "slightly" "different resources, or same resources at "different rate", +/or a slightly different infectious rate.  Some change in use of "resources", i.e. host.  Were as a different species of virus; usually uses a different host. This is not totally true as we know  virus  jump hosts, i.e. flue from birds, swine, etc... and of course love "mutate".  Also strains should have some variances in their genetics.   A 1st cousin, might have specific marker that you could "use" to establish it as different strain than the parent.  I'm not sure if virologist, consider the two very different AIDS virus "strains" or sub-species, they have been studied in depth, genetically, ability to infect and length of dormancy.  I'd read up on them for a good "standard" study.  Old but good papers.  A "stain" should have some slight change that is measurable: rate infection, length of infection, mortality rate, and any changes in hosts.

For most of the microbiologists: Viral strain - virus type having identical physiological characteristics: enzymatic activity, virulence, multiplication rate, ability to activate an immune response, etc. Two strains may have different biological properties and activities, but the genetic material of one strain is consistent / match the genetic material of the second strain.

Geneticist point of view: Virusses exist in quasispecies. That is, they exist concurrently as a pool of genetically variant individuals. "strain" to me is an operational definition that has value for the taxonomist. 

Thank you very much to everybody!