Core name

Microprocessor manufacturers are constantly improving their products by re-designing and tweaking all or some CPU internal parts. These modifications fall into the following three categories:

• Major micro-architecture overhaul, where all parts of the CPU are redesigned from scratch, often using new methods and techniques of increasing CPU performance. These changes are introduced as new CPU micro-architectures.
• Adding new and/or changing existing parts of the CPU core with the purpose of improving processor performance. These modifications are presented as new core generations.
• Minor tweaks to the core with the purpose of fixing existing errata, increasing microprocessor stability and/or reliability, or enhancing other CPU characteristics. These changes are classified as new core steppings.

Each new core generation adds new, or substantially changes one or more of existing CPU features, without changing the underlying micro-architecture the CPU is based upon. During development phase of each new core generation, the core is assigned a unique name, so called "codename". While this name is widely used in companies' roadmaps and other pre-launch documents, it is not mentioned in product documentation or advertisements aimed at general public. Codenames, or core names, are still used by hardware enthusiasts for historical reasons, and because they may be more descriptive than, say, family names. Some CPU families, like Celeron and Sempron, span not only multiple core generation, but also multiple micro-architectures. Using core name for processors from these and similar families gives much better understanding about CPU features and functionality than the family name.