Intel Small Matrix Library (optimized for Pentium 3, free, probably what you are looking for)
Intel Integrated Performance Primitives ($)
AMD Core Math Library (free)

There are countless others, but I'm guessing the first one is what you want.

I personally use the boost UBLAS library with the boost-bindings library to use it with LAPACK and ATLAS. I am doing more scientific research type work and probably would not use it for game type things. I have heard some people mention the configurable math library and the site definitely sounds like it has a lot to offer. As far as I can tell it uses expression templates much in the way the boost matrix/vector library does. It is also under the Boost Software License so it is free for commercial and non-commercial use. The only possible downside I can see is that it is fairly new and under development.

Here is the link to the library: CML

http://vxl.sourceforge.net/
Specifically: vnl (numerics): Numerical containers and algorithms. e.g. matrices, vectors, decompositions, optimisers.

Intel's Small Matrix Library:

http://developer.intel.com/design/pentiumiii/sml/src.zip

BTW - I did some research on the topic some time ago, but could not find anything that would fit my needs (thinking about it now maybe the Small Template Library would have been a good choice if I knew it existed)

Post with more details on the results of my research:

http://www.gamedev.net/community/forums/topic.asp?topic_id=440356

cmldev and gmtl were very nice, maybe what you are looking for, here is the list I found:

gmtl, nice template library for math functions only http://ggt.sourceforge.net/html/index.html

vmmlib, nice template library from http://sourceforge.net/projects/vmmlib/

tinymath, http://sourceforge.net/projects/tinymath/

glmath, http://sourceforge.net/projects/glmath/

gamemath library proposals: http://game-math3d.svn.sourceforge.net/viewvc/game-math3d/proposals/

mathlib3d for C++ and java: http://sourceforge.net/projects/mathlib3d/

mtl3d optimized SIMD: http://fvmtl3d.svn.sourceforge.net/viewvc/fvmtl3d/trunk/

geometric tools: http://www.geometrictools.com/LibFoundation/Interpolation/Interpolation.html

cmldev, from gamedev member: http://cmldev.net


Note that I found that for the compilers I tried (VC and Intel) using a plain C++ library with SSE optimizations yielded some very good results.