Bare bones? Probably $200-300,000. Most companies will spend a couple million. And I'm sure if you loaded it to the gills with RAM and disk you could spend $25,000,000 USD.
Sounds like a lot, but it's cheaper than re-writing all the software you're running on your current Z10/Z11/Z12.
Depends on the model honestly. You purchase the maximum capacity frame you would ever want to run on the z14, and then you purchase the capacity you actually want to run right now. IBM has 4 models listed in the specs, so you purchase one of those models, then you configure the z14 to run at a certain capacity. You are about dead on with it being $300k for a basic frame, but then you have to add the "maintenance" to run your stated capacity. And then you pay a monthly fee for whatever capacity your company actually runs.
That's what we do anyway. Companies can run full capacity frame and not do the sub-capacity pricing, but that's a lot more up front. Better, IMO, to pay the monthly rate and spread the cash payments out over 5 years.
I understand in big deployments IBM (e.g. Verizon cell phone billing) puts a few extra machines on site at no cost so that if one of your primaries go down you can start one of the backups and then they start to charge you for that one. I understand they do this with disks as well.
It almost certainly is not "no cost"-- typically that kind of stuff is rolled into a line item under "service agreement". And if the line item cost does say "zero" that's because of negotiation and sales engineer magic (possibly after a few $20 martini's). The cost is folded in SOMEWHERE. IBM ain't no non-profit!
IBM does suggest purchasing one production mainframe and one at CBU (Capacity Back Up) pricing if the company runs their own warm/hot/cold Business Continuity location. As great as these mainframes are at staying online, we have to account for storms and other threats to the data center, and need that redundancy somewhere. The CBU box will usually be much cheaper to run than the production box, but they are not free.
It's a hard question to answer. With this kind of gear IBM sells both the machine itself and compute capacity. You have a long term lease on the baseline workload, then lease some MIPS when you need it.
It's further clouded by virtual x86 and Linux workloads which have a different billing scale.
They also do stuff like provide "free" DR hardware that can only be activated in the event of an emergency.
