Comparing The Two Major PC Processor Architectures

Posted on May 28, 2012 by Nikhil P Naik

Processors are present in every single computer device available since they are the primary brain of the computer system.  Given the advancement and development of technologies, they have gotten much smaller and at the same time incredibly more powerful than ever before.  With the development of these technologies, speed has become a primary factor, since the need is always growing for power.  Unfortunately most people do not understand how processors work, but with some explanation how a computer works can be something anyone can get.

In order to get a good initial overview, looking at the most popular processors from Intel and AMD helps the most.  Intel for many years has concentrated on the technologies their processor uses to stay ahead of the competition.  Their idea behind processor development is to only use the amount of cores needed within a processor and make each of those cores as fast as possible.  AMD approaches things a bit differently since they focus on the middle range marketplace instead of the top end, and they develop more cores, each significantly slower than its competition.  When tests are done that take full advantage of each processor’s capabilities, they are actually not far apart from each other.  Unfortunately, AMD falls short in real world performance for a number of reasons.

Benefits Of Using A Single Core

Even though most applications thrive when using more cores to process, there are still simple tasks that have more of a benefit from just using a single core.  AMD copied Intel’s idea by creating a turbo boost feature that speeds up a core when necessary, but given that it is already significantly slower it does not help much.  The way the processor is designed also brings up some issues since modern versions of Windows simply do not support it yet.  Even though there are 8 cores, they are put together to share resources amongst pairs of cores.  The architecture is designed to handle the balance of resources efficiently, but Windows today simply cannot tell this is the case and uses every core as if it is completely separate from one another.  This causes an extremely inefficient means of processing and significantly hurts performance in nearly every application on the market.  The good news is that eventually, this will not be the case and everything will work as planned, but this method still is not as efficient as Intel’s methods.

Intel approaches the situation differently by only sharing level 2 cache, which works in its favor since there is a pool of information for each processor to pull from.  This means information gets to the processor and back to the Cisco compatible GLC-T module faster than in an AMD processor.  Additionally, Intel has been integrating more controllers in their processor to make the computer a simpler device, and even though high end Cisco compatible SFP modules are not integrated yet, technology is headed that way.

Tony Farinholt is a fiber optic technologies specialist writing mostly about things like the Cisco GLC-T module.

Image by Wimox

About Nikhil P Naik

Nikhil Naik has a Master's Degree in Information Systems, and is currently working as a Software Engineer at Microsoft. He also loves playing cricket, listening to music, and traveling. Twitter Handle - @buzz_nikhil.

Leave a Comment


This site uses Akismet to reduce spam. Learn how your comment data is processed.