Tom's Can't Overclock An Athlon FX!

[NEO3]

The "THG Tuning Test: Core 2 Extreme vs. Athlon 64 FX-62" is a shame for the site in terms of overclocking ability ( or luck with the sample).
They could only achieve +3Ghz, using a 12x254mhz.

http://www.tomshardware.com/2006/08/02/thg_tuning_test/page3.html

Compared to Intel's newcomer Core 2 Extreme, the overclocking potential of the Athlon 64 FX-62, which runs at a stock speed of 2.8 GHz, is pretty narrow. We couldn't achieve stable operation at more than 9 percent faster than stock speed, whether we used air or water cooling. This resulted in a top speed of 3.05 GHz. As soon as the clock multiplier was increased by 1, the motherboard won't start up any more. Let's be perfectly clear for the benefit of the hard-core overclockers among our readers: there is nothing more you can do at this point to squeeze any more speed out of this CPU. The step up to 3.2 GHz is simply too big, because of the 90 nm fabrication process used to build this processor.

Our friends at HardwareZone have some info for Tom's:

http://www.hardwarezone.com/articles/view.php?cid=2&id=2010

Since our Athlon 64 FX-62 and Core 2 Extreme X6800 processors are both multiplier unlocked, we are able to overclock them to these speeds through a simple multiplier increase.

I know that C2DE is faster than the FX. Its not the point here.

Im questioning Tom's ability to know what a false statement is:

Let's be perfectly clear for the benefit of the hard-core overclockers among our readers: there is nothing more you can do at this point to squeeze any more speed out of this CPU.

 

[SidVicious]

3GHz out of an FX-62 is a really poor OC, disappointingly, the THG team was quick to blame it on the CPU.

I saw no mention of dropping the HT multiplier down in order to keep it within specs (4x or 3x usually does the trick) and a Vcore over 1.5V would have been much more appropriate since we are dealing with a 90nm CPU.

IMHO, this was a crude and amateurish attempt at overclocking.

I can do 2.9GHz on an old Manchester core that was fabbed on the 43rd week of 2005, provided that I stacked all the odds in my favor with watercooling and a rather high Vcore I find it hard to believe that 3.05GHz was the best that FX-62 was really capable of.

I am not questioning the superiority of the C2D, I am questioning the THG team "Savoir Faire".

 

[SidVicious]

Even if you take into account that the 90nm SOI process ought to hit a wall at some point, given decent cooling, an high-end motherboard, premium RAM and a team that actually know what they are doing, I would have had expected at least 3.2GHz out of an FX-62...

Since the hardware they used was spec'ed out, it leaves us with a single option :?

 

[accord99]

Im questioning Tom's ability to know what a false statement is:

Let's be perfectly clear for the benefit of the hard-core overclockers among our readers: there is nothing more you can do at this point to squeeze any more speed out of this CPU.

The only thing is Hardwarezone only overclocked their FX-62 to 3GHz.

 

[jimytheassassin]

fx-62 vs 6700 I looked around and was curious also.. but I haven't seen anyone post much over 3.1ghz. That's not a huge margin of difference. But is what is is. Any review or article has to be taken with a grain of salt. Past performance may not predict future results

 

[mpjesse]

Dude... they were only using STOCK COOLING for both systems. The overclocking scenario was supposed to represent "easiest methods" for overclocking. Additionally they wanted to keep both CPU's in a similar temperature threshold.

And like someone said earlier, you ain't gonna get all that far beyond 3.0Ghz w/ a FX-62. Even with performance cooling...

 

[nobly]

Im questioning Tom's ability to know what a false statement is:

Let's be perfectly clear for the benefit of the hard-core overclockers among our readers: there is nothing more you can do at this point to squeeze any more speed out of this CPU.

The only thing is Hardwarezone only overclocked their FX-62 to 3GHz.
accord99 is right. The FX-62 is clocked at 3.013Ghz...
HardwareZone just increased the multiplier 15x200, but THG increased the memory speed...

If you have other proof that the FX-62 can go over that on air, lets see it.

 

[icbluscrn]

Dude... they were only using STOCK COOLING for both systems. The overclocking scenario was supposed to represent "easiest methods" for overclocking. Additionally they wanted to keep both CPU's in a similar temperature threshold.

And like someone said earlier, you ain't gonna get all that far beyond 3.0Ghz w/ a FX-62. Even with performance cooling...

accord99 is right. The FX-62 is clocked at 3.013Ghz...
HardwareZone just increased the multiplier 15x200, but THG increased the memory speed...

If you have other proof that the FX-62 can go over that on air, lets see it.

agreed



Neo 3 its a shame you didnt read toms article, or maybe you did and did not understand it
Yep, I dont know what they put in the water in brazil

 

[spud]

Yea it does not make much sense. I got 2.5ghz out of my older winchester core 3000+ on a budget board with genaric ram. Something just does not seem to add up at all. If i can get 700mhz more out of an older cpu why can they not get more than about 200 with a much better one? :roll:

Perhaps 3.0 area is the clocking limit of the current generation of K8's, unless you use some exotic cooling.

 

[spud]

Dude... they were only using STOCK COOLING for both systems. The overclocking scenario was supposed to represent "easiest methods" for overclocking. Additionally they wanted to keep both CPU's in a similar temperature threshold.

And like someone said earlier, you ain't gonna get all that far beyond 3.0Ghz w/ a FX-62. Even with performance cooling...

Word.

 

[m25]

Heh, you should have written ' TOM'S "CANT" OVERCLOCK AN ATHLON FX!'...
TH is a great site but it looks like they aren't totally inpartial, their titles show it:
>Conroe enters and it's immediately Core2 knocks out Athlon64, picture is an AMD marked Zidane being showed the red card
>The FX60 beats every contender and it's just AMD Athlon FX-60's Dual-Core Assault... not mentioned against, who, what, and with what results. The old fashioned picture shows a quiet FX-60 floating lost in space :roll:
>The benchmarks don't show anything interesting but it's worth a Woodcrest preyn on Opetron (Woodcrest for Opetron's head and things like this)
>The whole sempron line is THE star performer in value CPUs, the benchmarks show it a blade near A64s but the title reads A Dissatifying Compromise With AMD's 64 bit Sempron 3400+
Going back we find articles with the "SUPER P4" breaking the "poor" athlon XP, even if the edge (if there was one) was pretty slim...

Hope they're reading this & correct themselves, it really has nothing to do with a serious authority.

 

[jimytheassassin]

I'm sorry.. I read it and I missed where they said stock cooling? But in either case it does say they could not reach above 3.05 with air, OR water.. Now..the point remains..what would 100mhz do..not much, but here's a link pointing to an fx-62 at 3.1 aircooled and how they did it in detail. Oh and by the way.. TomsHardware is not gospel

 

[xBlueBoron]

Isn't it known that the FX is not a good overclocker, they are already pushed to the limits (Similar to the overlocking of 4800s and 5000s I belive). Which is why some older processors could be overclocked better, like opterons and such, they came out of the fab underclocked to increase yields. I bet the sample Tom's has is a good representation of all FX chips, most likely that is all you get.

 

[1Tanker]

You have to wonder why THG is known to be an intel paid site by many IT journalists and hardware sites. Now you know where all the bias comes from.

If that's truly how you feel, then go ah heck-up some other forums....maybe AMDZone,
and don't bother coming back here....You won't be missed.

 

[nobly]

I'm sorry.. I read it and I missed where they said stock cooling? But in either case it does say they could not reach above 3.05 with air, OR water.. Now..the point remains..what would 100mhz do..not much, but here's a link pointing to an fx-62 at 3.1 aircooled and how they did it in detail. Oh and by the way.. TomsHardware is not gospel

Its on page 1...

In this battle, performance is the only thing that counts. We'll explore the upper bounds of thermal stability for both of these top-of-the-line CPUs with standard retail components, the Intel versions of which have just hit store shelves.

http://www.tomshardware.com/2006/08/02/thg_tuning_test/

Yeah, THG isn't gospel, but neither is the title of this thread... Just because THG said they couldn't make it over 3.05 doesn't translate into the broad statement "No one will ever reach over 3.05".
Besides, an extra 50Mhz (3.1Ghz-3.05Ghz) isn't enough to catch Conroe.

 

[m25]

If they're not obscuring the truth, at least, they play with peoples' interests and the FIRS IMPRESSION influence that a title has.
If you read this for instance, you won't find a single comparision with the CeleronD 351 mentioned at the beginning (because it would show all the crap of the celerons despite the "D" marking). The title does not certainly reality.
http://www.tomshardware.com/2005/08/01/a_dissatifying_compromise_with_amd/page2.html

 

[geralt]

There are reports of higher OC on air.
For example 3.15 GHz on stock cooler.

http://www.amd3d.com/overclocking/athlon_fx62_breaks_all_performance_records.html

Dude... they were only using STOCK COOLING for both systems. The overclocking scenario was supposed to represent "easiest methods" for overclocking. Additionally they wanted to keep both CPU's in a similar temperature threshold.

And like someone said earlier, you ain't gonna get all that far beyond 3.0Ghz w/ a FX-62. Even with performance cooling...

 

[mpjesse]

You have to wonder why THG is known to be an intel paid site by many IT journalists and hardware sites. Now you know where all the bias comes from.

Prove it. If it's a known fact, it shouldn't be too difficult, right?

Yeah... that's what I though. No evidence.

So do us all a favor and STFU with your lame conspiracy theories.

Thanks,

-mpjesse

 

[NEO3]

Now...where is that "Report" option...
Done. :lol:

 

[mpjesse]

I'm sorry.. I read it and I missed where they said stock cooling? But in either case it does say they could not reach above 3.05 with air, OR water.. Now..the point remains..what would 100mhz do..not much, but here's a link pointing to an fx-62 at 3.1 aircooled and how they did it in detail. Oh and by the way.. TomsHardware is not gospel

Where did I say THG was gospel? I'm not ignorant enough to read only one hardware site. I regularly visit Anandtech, Xbit, and Trustedreviews to keep perspective. In fact, I probably spend more time reading anandtech and xbit than i do THG.

 

[jimytheassassin]

In this battle, performance is the only thing that counts. We'll explore the upper bounds of thermal stability for both of these top-of-the-line CPUs with standard retail components, the Intel versions of which have just hit store shelves.

http://www.tomshardware.com/2006/08/02/thg_tuning_test/

Yeah, THG isn't gospel, but neither is the title of this thread... Just because THG said they couldn't make it over 3.05 doesn't translate into the broad statement "No one will ever reach over 3.05".
Besides, an extra 50Mhz (3.1Ghz-3.05Ghz) isn't enough to catch Conroe.

Right..that's what i was saying..thanks i guess..it was an example for mpjesse in response.. though this topic is a non–topic

 

[jimytheassassin]

Hey mpjesse, don't get too flustered. It wasn't an attack, just my opinion to the general populous and more to the original poster of this thread. It's good you read other reviews.. more people should.

 

[shinigamiX]

The only person here who deserves a report and a ban is you, loser.

 

[NEO3]

:lol:
I hope the "Report" option on the msg's will work. What a bunch of "bad mouths" plain rude/ignorant comments.

Dont you get that Tom's could have done it better?
Is that SO hard?

tsc tsc tsc...
:roll:

 

[mpjesse]

Hey mpjesse, don't get too flustered. It wasn't an attack, just my opinion to the general populous and more to the original poster of this thread. It's good you read other reviews.. more people should.

I'm not flustered. But I did think you were implying that I only read this site. So I took a little offense to my intelligence. But since you cleared all that up... it was my misunderstanding.

Thanks!