Other People Are Reading

    If I Change Web Hosts Can I Keep the Same Email?     

As of the time of publication, servers come in two main varieties: 64- and 32-bit -- the 64-bit type is the newer technology. The number of bits describes the size of the registers, data paths and other components in the server’s central processing unit (CPU). The fundamental advantages of a 64-bit server are its larger memory capacity and ability to process data in bigger pieces. 

Eranet Link:  Eranet Hot Plans

Memory Address Space
The number of bits in a CPU’s data addresses defines its maximum memory capacity. A 16-bit address, for example, can access up to 2^16 or 64K bytes of memory. Thirty-two-bit systems address up to 2^32 or 4GB of memory, which, for many users, is a generous amount. Sixty-four-bit servers have 18 exabytes (EB) of theoretical memory space -- or 4GB squared. Though 32-bit server designs have long provided ample support for large-scale software such as databases and Web servers, more demanding applications have placed a strain on the older technology. The larger memory address space of 64-bit servers provides capacity for future growth.

Data Movement
A server with 32-bit internal data paths moves memory 4 bytes at a time. A 64-bit server moves 8 bytes with the same effort. This is similar to doubling the lanes in a major highway: more lanes can handle more traffic. Having a 64-bit data path also helps energy efficiency, because the CPU moves twice as much data in a single operation. With increasingly large programs and databases, more efficient data movement is an advantage.

Software
Programers must develop different versions of software for 64-and 32-bit systems. Address sizes are crucially important to software. Though 64-bit systems can run older 32-bit software in a “compatibility mode,” this is less efficient than running software specifically written for a 64-bit system. However, a 32-bit server cannot run 64-bit software. Microsoft, Apple and Linux vendors such as Red Hat offer both 32- and 64-bit versions of their operating system software; the 32-bit versions cannot take advantage of 64-bit architecture.

Overhead
A 64-bit server’s larger addresses themselves make programs somewhat larger than their 32-bit counterparts. Every time a 64-bit program requests data, it needs twice as many bytes to specify the memory address. The larger memory capacities and faster internal data movement of 64-bit systems compensate for this.