To answer properly your question, yes.
xmm registers were introduced by Intel with the
SSE instruction set (IS) in 1999 with the Pentium III CPU.
SSE stands for Streaming SIMD Extension and is a set of vector instructions.
xmm registers are 128bit wide and can hold 4
doubles, or 16
SSE can speed up signal processing applications (image processing, sound processing, compression, ...), encryption, and others quite dramatically when used properly.
On the other hand,
mm registers are part of the
MMX IS, another vector instruction set older than
SSE (1997 I suppose), and are 64bit wide.
Nowadays the vector instruction sets are becoming quite a fashion in a certain way (vector
CPUs were the standard for supercomputers back in the 70s & 80s - Cray's, ThinkingMachine's, ... computer were all vector based). In the past few years, Intel came up with many versions of
SSE and two new IS called
AVX2 (Advanced Vector Extension) with 256bit wide vectors implemented on SandyBridge/IvyBridge/Haswell, and
AVX-512 first implemented on the KNC (Knight's Corner) of the Xeon Phi processor & co-processor line.
I encourage you to check the Intel documentation & Wikipedia for more information.