Companies split when two halves just have such vastly different objectives and futures that, at an organizational practical sense, it no longer makes sense for the same board/CEO/management to be running them together.
Splitting them up lets both halves select boards/CEOs/management that is best for them, and pursue strategies that are best separately. The "underperforming" segment may now perform better now that it's free to use AWS/Azure/Google cloud tools instead of just IBM's... it can enter into strategic alliances it couldn't before... it can merge with another company that wouldn't have made sense before.
As for the people who work there... they're still employed so nothing really changes day-to-day.
But the main point is that this frees the "worse part", if you still want to call it that, to do what is best for it. It may very well turn out to thrive and be a huge success. It's still a normal business like any other.
If it were truly dead weight it wouldn't be spun off -- it would be shut down and everyone would be laid off. The fact it's being spun off or split means it's expected to be a viable business on its own. Nobody can predict the future -- who knows, it might outperform the cloud part long-term.