Internal Structur of the GAC

The Purpose of GAC is to maintain relationship between a strongly named assembly and a subdirectory. Basically, the CLR has an internal function that take an assembly's name,version,culture,and public key token. This function then return the path of a subdirectory where the specified assembly's file can be found.
If you go to the command propmt and change to the %SystemRoot%\Assembly directory, you'll see the GAC actually spans multiple subdirectories. On my machine, I see the fo;;owing GAC-related directories:
c:\windows\Assembly\GAC
c:\windows\Assembly\GAC_MSIL
c:\windows\Assembly\GAC_32
c:\windows\Assembly\GAC_64

The c:\windows\Assembly\GAC directory contains the assemblies that were created for versions 1.0 and 1.1 of the CLR.These assemblies may contains entirely of managed code(IL), or the assemblies may contain managed and native x86 code.

The c:\windows\Assembly\GAC_MSIL directory contains the assemblies that were created for version 2.0 of the CLR.These assemblies consits entirely managed code.Assemblies in this subdirectory can run in a 32-bit or 64-bit address space therefore, they can rn on a 32-bit or 64-bit version of Windows.

The c:\windows\Assembly\GAC_32 directory contains the assemblies that were created for version 2.0 of the CLR.These assemblies contain managed code as well as native x86 code.Assemblies in this subdirectory are allowed to run only in a 32-bit address space and therefore, they can run on an x86 version of windows or using the Windows On Windows64
technology running on a 64-bit version of windows.

The c:\windows\Assembly\GAC_64 directory contains the assemblies that were created for version 2.0 of the CLR. These assemblies contain managed code as well as native x64 code. Assemblies in this subdirectory are allowed to run on in a 64-bit address space,and therefore, they can run only on on a 64-bit version of windows.