As promised, Microsoft has taken the lid off of their DirectX 12 reveal earlier today at GDC 2014. Get ready for some nice performance gains on machines with lower end CPUs, but not until December of 2015 at the earliest.

The good news is that most current, modern GPUs will be able to make use of DirectX 12. Support for DirectX 12, at least for Nvidia, will be on all "the DX11-class GPUs it (Nvidia) has shipped." This includes cards from the Fermi line, Kepler, Maxwell, and beyond.
From the MSDN DirectX 12 Blog:
If you’re a gamer, you know what 3DMark is – a great way to do game performance benchmarking on all your hardware and devices. This makes it an excellent choice for verifying the performance improvements that Direct3D 12 will bring to games. 3DMark on Direct3D 11 uses multi-threading extensively, however due to a combination of runtime and driver overhead, there is still significant idle time on each core. After porting the benchmark to use Direct3D 12, we see two major improvements – a 50% improvement in CPU utilization, and better distribution of work among threads.


From the Nvidia Blog:
GPU performance can be exploited three ways: drawing better pixels, more pixels and more objects. We have reaped much of what can be gained from pixels. DX12’s focus is on enabling a dramatic increase in visual richness through a significant decrease in API-related CPU overhead. Historically, drivers and OS software have managed memory, state, and synchronization on behalf of developers. However, inefficiencies result from the imperfect understanding of an application’s needs. DX12 gives the application the ability to directly manage resources and state, and perform necessary synchronization. As a result, developers of advanced applications can efficiently control the GPU, taking advantage of their intimate knowledge of the game’s behavior.

There is no word yet as to what operating system this will be available on. Let's just hope Microsoft doesn't pull a DirectX 10 move here and limit DirectX 12 to Windows 9.