4 card GTX 1080 ti set ups are doing NTLM and NetNTLM in the 200 and 100 GigaHashes/s range, respectively... this rig would double that.
For reference, our 3 card GTX 980s do 64 GH/s for NTLM and 35 GH/s for NetNTLMv1
If we built this rig, we'd blow our whole upgrades budget on it and would have to live with our elderly VMware server down here for another round.