After looking at the IGP options for the Intel platform, our attention today switches to AMD. Both AMD and NVIDIA have released IGP chipsets this year and we put them to the test to find out the ultimate one - AMD 780G/740G vs. GeForce 8300/8100 mGPU. Read on to find out which chipset and board to get for your next AMD HTPC!
Hi, you're right that the FX-62 is actually a 2.8GHz chip, but we've downclocked the CPU.
The reason is that we were previously using another CPU of lower frequency in an earlier IGP article (http://www.hardwarezone.com/articles...494&cid=6&pg=3), and to ensure results are comparable internally, we've downclocked this CPU which is of the same architecture of what we require.
Hope that clarifies matters
Quote:
Originally Posted by Lithotech
In the Test Setup section, you list the AMD Athlon 64 FX-62 as being clocked to 2.5GHz.
The AMD Athlon 64 FX-62 is actually supposed to be 2.8GHz:
Thanks for the reply! Yes, that makes sense to a certain extent. I realise how difficult it is to keep an even playing field sometimes, and how important it is to try to achieve comparable results when technology keeps moving forward.
Obviously, you are aware that the below example is still skewed? The X2-5000 clocked at 2.6GHz not 2.5GHz and has half the L2 cache that the FX-62.
Any reason the FX-62 was not clocked to 2.6GHz? If I am not completely mistaken, the memory divider would work differently as well, with a 13x CPU multiplier the memory would run slower at certain dividers.
Thanks!
Quote:
Originally Posted by Dr.Vijay
Hi, you're right that the FX-62 is actually a 2.8GHz chip, but we've downclocked the CPU.
The reason is that we were previously using another CPU of lower frequency in an earlier IGP article (http://www.hardwarezone.com/articles...494&cid=6&pg=3), and to ensure results are comparable internally, we've downclocked this CPU which is of the same architecture of what we require.
Hope that clarifies matters
__________________
___________________________________
You don't learn by doing things right.
Hi Lithotech, you're really sharp
Yeah, keeping a similar comparison is tough balance these days to make sense of all the results.
Regarding the Athlon X2 clocks and cache, you're right the X2-5000 is clocked higher at 2.6GHz but has half the cache of the FX-62. Reason we've clocked the FX-62 to 2.5GHz is because our internal testing has generally been consistent that a higher cache based Athlon X2 is equivalent to another X2 with half the cache size when the clocks speeds of the former is 100MHz lower.
That's the reason why we've run the FX-62 @ 2.5GHz to simulate an X2-5000 that normally runs at 2.6GHz. I think we made a mention of this reasoning in one of our old articles, but I just can't recall which one is it.
Hope the above explanation clears up matters!
Quote:
Originally Posted by Lithotech
Thanks for the reply! Yes, that makes sense to a certain extent. I realise how difficult it is to keep an even playing field sometimes, and how important it is to try to achieve comparable results when technology keeps moving forward.
Obviously, you are aware that the below example is still skewed? The X2-5000 clocked at 2.6GHz not 2.5GHz and has half the L2 cache that the FX-62.
Any reason the FX-62 was not clocked to 2.6GHz? If I am not completely mistaken, the memory divider would work differently as well, with a 13x CPU multiplier the memory would run slower at certain dividers.
Thanks!
__________________
Vijay Anand
Editor (www.hardwarezone.com) and Forum Administrator
Hardware Zone Pte Ltd
Very interesting, indeed! It makes sense, 100MHz isn't a lot, but then the larger cache doesn't make a huge difference in many tests.
I am now wondering how it calculates the other way, for example: I have an X2-5000 Black Edition, and if I were to clock it to 2.9GHz would it come close to matching the FX-62 at 2.8GHz? It's experiments like this that make me drool!
Quote:
Originally Posted by Dr.Vijay
Hi Lithotech, you're really sharp
Yeah, keeping a similar comparison is tough balance these days to make sense of all the results.
Regarding the Athlon X2 clocks and cache, you're right the X2-5000 is clocked higher at 2.6GHz but has half the cache of the FX-62. Reason we've clocked the FX-62 to 2.5GHz is because our internal testing has generally been consistent that a higher cache based Athlon X2 is equivalent to another X2 with half the cache size when the clocks speeds of the former is 100MHz lower.
That's the reason why we've run the FX-62 @ 2.5GHz to simulate an X2-5000 that normally runs at 2.6GHz. I think we made a mention of this reasoning in one of our old articles, but I just can't recall which one is it.
Hope the above explanation clears up matters!
__________________
___________________________________
You don't learn by doing things right.
Haha, I bet if you did clock it that high, it would probably within striking distance fo the FX-62, heh
Quote:
Originally Posted by Lithotech
/blush
Awww, shucks, I just find this all interesting.
Very interesting, indeed! It makes sense, 100MHz isn't a lot, but then the larger cache doesn't make a huge difference in many tests.
I am now wondering how it calculates the other way, for example: I have an X2-5000 Black Edition, and if I were to clock it to 2.9GHz would it come close to matching the FX-62 at 2.8GHz? It's experiments like this that make me drool!
__________________
Vijay Anand
Editor (www.hardwarezone.com) and Forum Administrator
Hardware Zone Pte Ltd
Channels
TECH GUIDES
