Native (processing)
From Wikipedia, the free encyclopedia
NATIVE: Methodology for sharing code processing over a number of CPU cores
 |
This article is uncategorized.
Please help improve this article by adding it to one or more categories, so it may be associated with related articles (how?). Please remove this tag after categorizing, but not before. |
| Contents |
|---|
| 1 Introduction |
| 2 Multiple Cores |
| 3 Practical Applications |
[edit] Introduction
At some fundamental level, all computers have a 'core' which takes care of the primary processing. Instructions, originally written in some high level language, finally get broken down into machine code and executed.
[edit] Multiple Cores
Recently, there has been a trend to add more cores to a system. In theory, the more cores a system has available to execute instructions, the more work gets done. It is a far from perfect model though, and anyone who expects that 2 cores can do twice the work of 1 core is going to be disappointed. More cores are almost certainly better, but the amount of improvement will depend on many factors - including available bandwidth and code efficiency. In this context, NATIVE refers to whether one core is independent of its neighbours.
[edit] Practical Applications
Current (2006) Intel quad-core technology can be viewed as a pair of dual core processors that happen to have been built onto the same substrate. Proposed (2007) AMD quad-core technology consists of 4 seperate cores, which can operate with their own resources, independently of their neighbours. How much of an advantage NATIVE technology offers will be a hotly debated subject for some time to come. Is there a native advantage - or is it just marketing terminology within a hair-splitting exercise ? For more info, try balancing http://multicore.amd.com/us-en/quadcore/ with http://www.intel.com/products/processor/core2quad/index.htm
