Home‎ > ‎Articles‎ > ‎Processors‎ > ‎

Intel Core i7 2700K versus AMD FX 8150 spec by spec

First of all, is this a fair comparison?  I would say so.  They were both launched towards the end of 2011, their speeds are pretty close, they are both unlocked for overclocking, they have similar cache sizes and are being sold to the same market.  Both Intel and AMD are marketing these processors as their flagship desktop processors.  They both sport multiple cores, and can handle pretty much anything you can throw at them.  But there are some key differences, that can make one of them better than the other.  And if you are price sensitive, only one of them is the best choice for you.  Let's compare the basics first, shall we.

 www.knowbytes.com
knowbyte on YouTube
knowbytes on BlogSpot
 Intel
i7 2700K

AMD
FX 8150
Launched October 24, 2011
October 12, 2011
Speed 3500 MHz
 3600 MHz
Level 3 Cache (L3)
8 Mega Byte
8 Mega Byte
Price (week of Nov. 10, 2011)
US $370
US $240
As you can see, the speed difference is only approximately 3%. The Level 3 cache is exactly the same, BUT, and this is a significant but, 8 MB on the i7 2700K is being shared by 4 cores, but on the FX 8150 it is being shared by 8 cores.  So on the FX 8150 there is that much less cache per core.  To be exact, each of the 8 cores on the FX 8150 gets 1MB of L3 cache per core (if you want to look at it on a per core basis), where as, each core on the i7 2700K gets 2MB of L3 cache.  Does this effect performance?  Well it may not, because the AMD Bulldozer architecture is different from the Intel i7 (there is also L1 and L2 cache which we will talk about later), and therefore, may be more efficient in using cache, consequently needing only half as much.  Traditionally, AMD processors have sported around half the L3 cache of their equally performing Intel counterparts.  So if the past has anything to bear on this comparison then one may not be alarmed by what seems to be half the L3 cache per core on the AMD FX 8150 versus the Intel i7 2700K.  Also bare in mind that we haven't yet compared Level 1 and Level 2 cache sizes yet, which, depending on their size and how efficiently they are used in the architecture can make up for the difference in performance caused by a smaller Level 3 cache size.

The price, on the other hand, instantly kicks you in the privates with the glaring advantage on the FX 8150 side.  Benchmarks be damned! If you are price conscious, then getting 8 cores on the FX 8150 is definitely attractive versus getting half the cores on the i7 2700K for around $130 cheaper. 

And if you are looking at whatever benchmark you trust, you will surely need to take into account the glaring price difference of 54% and match that with whatever percentage difference you are getting in your favorite benchmark on your favorite processor cycle usage (gaming, video/audio editing/rendering, 3D animation...) and wonder if its worth saving or splurging.





www.knowbytes.com
knowbyte on YouTube
knowbytes on BlogSpot
 Intel
i7 2700K
 AMD
FX 8150
Architecture Sandy Bridge Bulldozer
Transistors
 995 Million
 approx. 2000 Million
Cores 4 8
Floating Point Unit
 4 4 (each is shared by 2 cores)
Manufacturing
Process
 32 nanometer
32 nanometer

So far we notice three key differences, twice the number of cores, consequently, twice the number of transistors and the same number of floating point units even though the AMD Bulldozer FX 8150 has twice the number of cores.  The fact that the FX 8150 has twice the number of transistors should come as no surprise.  After all, if you have twice the number of cores, you are going to need that many more transistors to accommodate the increase in complexity that the processor faces.  Further on we will also notice how this effects energy usage. 

However, the fact that it has the same number of floating point units as a 4 core processor is surprising.  How does it effect performance?  Well, the short of it is that floating point units are not really used as much for desktop environments.  So based on that, one would imagine floating point units sitting around twiddling their thumbs waiting for the cores to ask them to do something.  Apparently this is the case in desktop user environments. 

So sharing the floating point units is not a bad idea, but its also not a great idea, because it can effect performance negatively if (and that is a rare if) the floating point unit is needed by both the cores at the same time, and when that happens, well, you can only imagine one of the cores sitting there waiting for the other core to finish, and that means a waste of the cores cycles, and thus, almost not worth having the other core.  But as is the case, this is a rare occurrence, and AMD research and development was probably betting on that rarity when they decided to save on resources by sharing floating point units.



 


The videos and information provided on this website are for entertainment purposes only.  So while I may use reasonable efforts to include accurate and up-to-date information on my sites, I make no warranties or representations as to its accuracy, timeliness, or completeness. I may periodically add, change, or improve any of the information, products, services, programs, and technology described on my sites without notice. I or anyone related to me assume no liability or responsibility for any errors or omissions in the content of my websites.
Copyright Benz Kohlee