Monday, August 10, 2009
My Web Links
Sunday, August 9, 2009
32 bit vs 64 bit
Processor registers are divided into several groups such as integer, floating-point ,control etc. In most processors only the integer registers can be used to adress data in memory. So the size of these registers limits the size of the directly addressable memory. The main difference between a 32 bit and 64 bit processors is the amount of directly addressable memory. A 32 bit processor can access a maximum memory of 4GB (2^32) whereas a 64 bit processor can access a maximum 2^64 Bytes( 16.8 million Terrabytes or 16 exabytes) of RAM.
A 64 bit processor has a databus that is 64 bits wide while a 32 bit processor has a databus of 32 bits wide. Thus a 64 bit databus can carry more information than a 32 bit data bus. So 64 bit processors are faster when it comes to large data processing involving huge memory space access. [In simle words, a databus is the cable that carries information to and from the processor and memory.]
Even though, theoretically a 64 bit processor can support 16.8 million terrabyte, most of the 64 bit microprocessors in the market has an artificial limit on the amount of memory they can address. AMD64 architecture has a 52 bit limit on the physical memory and currently supports only 48 bits.
The main disadvantage of 64 bit processor is that the same data occupies more space in memory. This increases the memory requirements of the processes and can have implication on efficient processor cache utilization.
Currently most of the softwares are compiled into 32 bit code. So it does not take advantage of 64 bit address space or wider 64 bit address registers.
Sun's 64-bit Java virtual machines are slower to start up than their 32-bit virtual machines because Sun has only implemented the "server" JITcompiler for 64-bit platforms. The "client" JIT compiler, which produces less efficient code but compiles much faster, is unavailable on 64-bit platforms.
Labels: 32 bit, 64 bit, Processors
Subscribe to Posts [Atom]