Central Processing Unit: Difference between revisions
→Architectural: Added image from Wikimedia Commons
(→2010s: Expanded) |
(→Architectural: Added image from Wikimedia Commons) |
||
| (2 intermediate revisions by the same user not shown) | |||
Line 1:
{{Useful Notes}}
[[File:Intel@intel7(10nmESF)@RaptorLake@RPL(8P+16E)@i9-13900K@ES DSC05138-DSC05138.jpg|thumb|A CPU die]]
The Central Processing Unit (CPU, or simply “the processor”) is essentially the “brain” of a computer.
Line 12 ⟶ 13:
All processors have a basic vocabulary known as an Instruction Set Architecture (ISA). Different processors sharing the same or similar architectures can run each others' software with little or no speed penalty. In contrast, running software written for a different architecture requires [[Emulation]], and imposes an enormous speed penalty. Examples of popular ISAs today are 80x86, POWER/PowerPC, ARM/XScale and MIPS.
A CPU is usually made faster by increasing its [[Clock Speed]], by increasing [[Powers of Two Minus One|the word length or "bit number"]], by designing it so that it can handle different stages of multiple instructions at one time (called “pipelining”), or by adding a small "cache" of high-speed memory to store frequently-used instructions and data. Higher clocked and more complex chips run hotter, which is typically dealt with by shrinking the manufacturing process of their components (called a “process shrink.”) The core component out of which CPU circuits are created, the [[wikipedia:MOSFET|MOSFET]]s, in CPUs of 2014 are as small as 14 nanometers<ref>That's 1.4 millionths of a centimeter</ref> across, source to drain.<ref>At least, according to this: [http://newsroom.intel.com/community/intel_newsroom/blog/2014/08/11/intel-discloses-newest-microarchitecture-and-14-nanometer-manufacturing-process-technical-details "Intel Discloses Newest Microarchitecture and 14 Nanometer Manufacturing Process Technical Details"]</ref>
=== A brief History ===
==== Architectural ====
[[File:Intel 4004 ad.jpg|thumb|Ad for the Intel 4004: a big deal in the history of CPUs]]
The first CPUs were not very centralized at all. Instead, they were comprised of multiple chips, and even basic electronics that did some logic were done with discrete chips that did [[Binary Logic]]. These systems were also only capable of integer math at a reasonable speed. In order to do complex math faster, specialized chips such as Floating Point Units and Digital Signal processors were used. When the original Intel 4004 was released, it was called the first "Integrated Microprocessor" because it contained both the logical part and arithmetic part in the same package. During the 80s, processor manufacturers were finding more and more chips to put into the CPU in order to keep system costs down. Eventually this accumulated to the the microcontroller, which contains a CPU, memory, and plenty of configurable I/O ports. The evolution of this is the System-on-a-Chip (SoC), which, for all intents and purposes, is a self contained computer with human-friendly user interfaces (such as a GPU to drive a display).
| |||