It's disappointing that xam2iks didn't do a follow up on this because I think it had - has, in fact - the potential to be a most interesting conversation.
MarenC makes a good point; if one did not know better - and one does not - one MIGHT feel that part of the reason that IBM is trying to overcomplicate the living h3ll out of the product is to make it all but impossible to run - for practical purposes - on premises. Presumably they would then assume that this will generate no ill will, resentment or lack of trust on the part of their on prem customers, and that everyone will follow like a herd of sheep into their cloud offering. Of course if they do think that, then IMHO they have their collective head shoved so far into different types of "clouds" that they are unlikely to see 787s full of departing customers vectoring straight for that head. But hey, this is purely speculation on my part. I make no claim that I have correctly read their intentions.
On the other hand, being an on prem customer I have extensive first hand experience of the level of "concern" that IBM shows us. As my smile shall now show.
(Aside: I really haven't used SQL Server Reporting Services enough to pass comment on it, but IMHO SSAS (Analysis Services) just keeps getting better and better.)
But this leads to my key point; I think the question is being looked at bass-ackward. What can
replace TM1? On the server side, not much yet. Its in memory advantage is being eroded; for example a lot of things in the "Big Data" space in Google Cloud Services are mostly if not entirely running in memory. Ragged dimension consolidations are something that we take for granted in TM1 but in reality it's not guaranteed that you'll get that everywhere. Implementing business rules by rule calculations isn't unique, but they're (overall) extremely well implemented in TM1. The ability to collaborate on budgets or forecasts in real time is not unique, but also far better than average. Alternative hierarchies are an inspired idea, but we don't yet have decent front end software to make use of them. Overall, however, there really isn't a server tool that can be "anything you want or need it to be" the way the TM1 server is.
What could replace PA on the client side? A lump of charcoal, a roll of toilet paper, a smoke signal fire and a set of tom toms. Oh wait, tom toms don't fail when you upgrade your antivirus software and are easy to install. Scratch that, the IBM client software has no parallel. No, really, I mean that. Just not in a complementary fashion.
So what could replace TM1 or, more generally, PA2? Probably nothing if you compare it feature for feature, the same as most software. Every program is unique, more or less.
But - key point - nobody is looking to
buy software. They are looking to solve a business problem. And in THAT regard the question to be asking isn't "what can replace TM1" but rather "What could be used to do the following business functions?" There may be alternatives, or it may be that for the moment at least TM1 is the only one that will cover the bases. It may be that one other product can do it and do it cheaper. (Like that last part would be hard.) It may be that several products would be a better solution. The only way we can discuss that is to know what it is that you need to do. From
there, we walk backwards to find the right solution.