Have HP (and or the article authors) forgotten
that HLRs used to be natural territory for the NonStop systems that are also widely deployed in the high value high availability financial sector? Systems that are already highly scalable in hardware and software, using software which HP will soon be ported onto x86 (though quite possibly not just any old x86).
The idea of running an HLR under a hypervisor might look good to the marketing types, and that would include Intel's Wind River subsidiary.
I'm not sure the reality would be quite as appealing to engineering and operations types (and even to management types with an interest in service levels), but then I'm an outside observer rather than an insider.
"Carriers, however, are wary in their NFV adoption, because they like the high availability associated with traditional telco kit."
The missing words: Tandem? NonStop? OpenCall?
http://www.availabilitydigest.com/private/0102/hp_hlr.pdf