WTF 5 Ghz P4?

[MJO]

EDIT: Warning: This post contains a link to an article with the THG guys showing off.

Usually I do not read THG but I made an exception.

Toms Hardware has set a world record in P4 OC.
Hardware used:
P4 cpu (unknown model engineering sample)
Asus P4C800-E Deluxe
Corsair PC4400 ram.

OC Result: 5255,4 Mhz.

Here is a link to their record attempt.
As always I am a bit sceptical because it is coming from THG.
http://www20.tomshardware.com/cpu/20031230/index.html

Here are some images

EDIT: Forget the images, THG doesn't like remote linking.

[ginipig]

That's just plain nuts. I should know. I made a living as a teen selling nuts at a local ballpark.

[keto]

Ya, I just read the article plus downloaded & watched the video. I'm not a big fan of Tom's either, they continually make statements then don't back them up with data - for instance, they say they were rock stable at 4700 but don't give any examples or illustrations. They also make some statements about bandwidth that they don't prove. It was interesting enough to watch but just as a 'spectacle', not as something particularly enlightening.

[ginipig]

I haven't been in the overclocking business that long (heck, I haven't even seen a wc kit irl,) but what I do know is that Tom has questionable data. When I first started tinkering with computers (a few months ago,) I was looking for sites that dealt with the topic in layman's terms. Tom's site was the first that was recommended on google, but each successive site after that seemed to have different information.
Every passing moment gave me the impression that I shouldn't have even bothered with Tom's guide at all.

Thanks to Creep (don't know where he's at now,) I arrived here. JSOY, Tom's a nut :aol:

[Thrax]

Herr Pabst has one of the dirtiest names in the entire industry.

Trying to lock other websites out of attending public events, inconsistent data, being paid for by Intel, etcetera.

It's not good.

[Disvengeance]

I dont thinks its true. They could atleast have shown other benchmarks ( even though they said it was ''impossible due to the high clock speeds'' ) to prove that they did have it running at that speed. CPU-z doesnt exactly prove it.

[RWB]

If you give me that same equipment, I'd hit 5GHz and post a screenshot. One thinG I would have done first is make damned sure to post a damned benchmark or something AT LEAST.

Of course people here who are better than I am would be able to not only meet 5Ghz with their equipment, but maybe even pass it on and still be stable enough to post some benchmarks.

THG is just showing he has money, but no real skillz. He is still on my HOSTS list, so them links don't work ;) THG and [H]

[MJO]

No CPU-Z is no warrant for stability.
I believe I could OC my processor very high, take a screen shot and then setting it at stock again.
What would that prove?
Nothing.
Where are the benchmarks?

[Thrax]

[H] has credibility left; I will read them and their articles, then reference them with other sites to make sure.. But Tom is just crap.

[MJO]

They could atleast have shown other benchmarks ( even though they said it was ''impossible due to the high clock speeds'' )

Well if they cannot run any benchmark programs it is probably bacause they have oc'ed too far. That would indeed make benchmarking impossible.
And their record attempt invalid.

[Disvengeance]

yes, by that they probably ment that they overclocked so far and unstable that all the benchmarks crashed and they could not complete, so they worded it all fancy. :skeptic:

[fudgam]

They could have atleast run a benchmark if it was stable at 4.7.

[RWB]

So.. are people scared to use this smilie?

:bs:

LOL

[shwaip]

Honestly tho, just a processor clocked @ 5ghz is quite impressive.

(not that it's useful if it can only be stable for the 2 secs it takes to get the screenie, but hey, what do you expect)

[TheLostSwede]

That 5gig shot isn't there to prove that it is stable. It's just there to prove that he can with handpicked equipment. No point bothering about it. Nice clock though.

However, it isn't easy to run a bunch of benchmarks when you have to refill with LN2 all the time. I hate the site but this is his best effort so far though.
Only one thing that really bothers me. How the hell can he say that 4.7 was rock stable? It takes a week at least to determine if a cpu is rock stable or not. I seriously doubt that he had LN2 enough for a week.

[edcentric]

What is OP's current record, he runs real stability tests.
No benches makes it very suspect.
I couldn't figure out why THG didn't LN2 the entire mobo. The wouldn't have needed fancy nb cooling and the memory would have liked it too.

THG includes some nice tables and ref. info in their reviews, but their conclusions are often very slanted.