Q6600 isn't real quad?

[amdfangirl]

Anyone at all?

 

[NMDante]

I don't think it's MMM, 9-inch, Sexbomb, Sharikou(insert random string of characters) or any of those guys from a couple of years ago. They didn't write as much. This actually looks much more like BM's work.

BM in disguise? Maybe.
Same stubborn attitude. Won't accept benchmark results. Makes up excuses or reasons why everything is setup to make AMD look bad. Claiming this and that, without showing proof. Same "better than you" style. Hmmm....

Could be BM in disguise. Don't think it's 9inch, Shakirou, or MMM, myself.

 

[keithlm]

I guess you also haven't learned that if voltage required is out of the range of voltage supplied, CPU will become unstable.

I agree that it usually takes hours to find a sweet spot. However it doesn't mean by tweaking and adjusting, you can lower a Vcore down by 0.15V. Not happening. Especially on Phenom, where tweaking IMC speed is as important as CPU speed, and also as a major heat generating source, I really don't think you can simply "tweak and reduce".

So let's see, a 0.1V increase to achieve a 400Mhz overclock. So if your theory, which is similar to linear projection, applies (which it does not), how much voltage increase is needed to achieve 3.4Ghz? If voltage vs. clockspeed relationship is more like a exponential graph (which is more true), how much voltage will it likely need to achieve 3.4Ghz?

You are pretending that these chips work exactly like Intel chips. They don't. But did you know that while increasing frequencies people have seen "walls" where it doesn't matter WHAT voltage you give... it won't be stable until the frequency is increased enough to get past that "wall". This would mean that the voltage requirement for the CPU is changing in a non-linear pattern. So "my theory" does not ascribe to the a linear projection as you claim in your post. It is more of a hybrid step pattern mixed with an exponential pattern.


(I have no idea where you came up with a requirement of +0.1V to get an overclock of 400Mhz. Or where you talking about Intel?)

Currently on the chart that you linked you can see 600Mhz overclocks on stock voltages using a motherboard that has the sb700. AMD has claimed that the SB700 has the same stability fixes that will be included in the sb750 but NOT the overclocking and power optimizations. BUT we do not have enough data on how the voltages on these boards increase as the frequency increases.

Also on the chart you linked: looking at the older sb600 boards you can see +600Mhz using +0.025V. It appears that the overclock "limit" was not based on the voltage or the temperature on these older sb600 chipsets. We do not have enough data to know what the power requirements will be with the newer stable southbridge.

Things are going to get very interesting in the near future. But for today I'm ordering some new Mushkin Ascent XP2-8500 ram and a couple of Zalman XF-1000 for my 4580's. (I hate seeing 70C at load on the GPU. But even worse I hate the fan noise.)

(EDIT: I meant 600Mzh for the stock voltage on the sb700... not 700Mhz. 2.4 to 3.0Ghz.. on stock.)

 

[yomamafor1]

You are pretending that these chips work exactly like Intel chips. They don't. But did you know that while increasing frequencies people have seen "walls" where it doesn't matter WHAT voltage you give... it won't be stable until the frequency is increased enough to get past that "wall". This would mean that the voltage requirement for the CPU is changing in a non-linear pattern. So "my theory" does not ascribe to the a linear projection as you claim in your post. It is more of a hybrid step pattern mixed with an exponential pattern.

True, that your theory is based on exponential function, but with inverted axis. You claim that as voltage goes up, clockspeed goes up exponentially, which is clearly not the case.

As for the wall, you should really consider this. 3.4Ghz with the same voltage ok, but not 3.5Ghz? Is that really a wall, or just that more voltage must be put in to maintain that clockspeed? Do you have proof that its a wall?

(I have no idea where you came up with a requirement of +0.1V to get an overclock of 400Mhz. Or where you talking about Intel?)

No, that came straight out of the chart I linked from XS: 3000Mhz for 1.3V. The stock 9850 BE is 1.2V @ 2600Mhz

Currently on the chart that you linked you can see 700Mhz overclocks on stock voltages using a motherboard that has the sb700. AMD has claimed that the SB700 has the same stability fixes that will be included in the sb750 but NOT the overclocking and power optimizations. BUT we do not have enough data on how the voltages on these boards increase as the frequency increases.

Also on the chart you linked: looking at the older sb600 boards you can see +600Mhz using +0.025V. It appears that the overclock "limit" was not based on the voltage or the temperature on these older sb600 chipsets. We do not have enough data to know what the power requirements will be with the newer stable southbridge.

Things are going to get very interesting in the near future. But for today I'm ordering some new Mushkin Ascent XP2-8500 ram and a couple of Zalman XF-1000 for my 4580's. (I hate seeing 70C at load on the GPU. But even worse I hate the fan noise.)

I agree, we shall wait and see. Just from my personal experience, the statistical data out there, and common senses, I really don't see a Phenom reaching 3.6Ghz on 1.4V.

 

[keithlm]

True, that your theory is based on exponential function, but with inverted axis. You claim that as voltage goes up, clockspeed goes up exponentially, which is clearly not the case.

As for the wall, you should really consider this. 3.4Ghz with the same voltage ok, but not 3.5Ghz? Is that really a wall, or just that more voltage must be put in to maintain that clockspeed? Do you have proof that its a wall?

(I have no idea where you came up with a requirement of +0.1V to get an overclock of 400Mhz. Or where you talking about Intel?)

No, that came straight out of the chart I linked from XS: 3000Mhz for 1.3V. The stock 9850 BE is 1.2V @ 2600Mhz



I agree, we shall wait and see. Just from my personal experience, the statistical data out there, and common senses, I really don't see a Phenom reaching 3.6Ghz on 1.4V.

My "theory" says nothing about the clockspeed going up exponentially. I do say that the Voltage use will go up exponentialy and not linearly. But in a hybrid step pattern as clockspeed goes up.

No. I have no proof that 3.5Ghz is part of the wall... but I also have no proof that 3.4Ghz and 3.5Ghz couldn't be achieved with less than the max available voltage. But then other than the one leaked little bit of data we don't know anything at this point; and they postponed the SB750 release until the 2nd or 3rd week of July. We ALSO do not know for sure whether "kenofstephen" used the new AOD Extreme feature that AMD has bragged about. If he did NOT... then we could be in for a "double whammy" when they actually do release. Maybe "triple" if they have also fixed the power savings features. (And I think it is this last that has caused the delay.)

NOTE: The stock 9850 BE runs at 2.5Ghz and the voltage for most newer bios is 1.3V. So you meant to say "stock voltage for +500Mhz" instead of "+0.1V for 400Mhz".

 

[akistzortzis]

I bought a Gigabyte EP45 montherboard and an OEM Q6600 which is rated at 266MHz FSB and 2.4GHz operation. I raised the bus to 375MHz and it is now running happily at 3.3375GHz. I did not raise the voltage or anything else, runs very cool with the Zalman 9500 cooler to minimum at 46-50 C after 15+ minutes of 100% CPU on all 4 cores.

I thought I would have to raise the core voltage to get it to overclock so much, but I did not - unless the motherboard does it automatically ?

 

[fazers_on_stun]

discreet graphics.

discreet

English

Etymology
From Old French discret, from Latin discretus, from past participle of discernere.

Homophones
discrete (separable into parts)

Adjective
discreet (comparative more discreet, superlative most discreet)

Meaning: Respectful of privacy or secrecy; quiet; diplomatic.
With a discreet gesture, she reminded him to mind his manners.
John just doesn't understand that laughing at Mary all day is not very discreet.
Not drawing attention, anger or challenge; inconspicuous.

Usage notes
Do not confuse with discrete.

 

[TechnologyCoordinator]

discreet

English

Etymology
From Old French discret, from Latin discretus, from past participle of discernere.

Homophones
discrete (separable into parts)

Adjective
discreet (comparative more discreet, superlative most discreet)

Meaning: Respectful of privacy or secrecy; quiet; diplomatic.
With a discreet gesture, she reminded him to mind his manners.
John just doesn't understand that laughing at Mary all day is not very discreet.
Not drawing attention, anger or challenge; inconspicuous.

Usage notes
Do not confuse with discrete.

LOL, thanks bro.

 

[Shadow703793]

Why is this thread not dead yet???

 

[TechnologyCoordinator]

Why is this thread not dead yet???

People keep bumpin' it!

 

[fazers_on_stun]

BM in disguise? Maybe.
Same stubborn attitude. Won't accept benchmark results. Makes up excuses or reasons why everything is setup to make AMD look bad. Claiming this and that, without showing proof. Same "better than you" style. Hmmm....

Could be BM in disguise. Don't think it's 9inch, Shakirou, or MMM, myself.

IIRC BM's posts were much shorter, lots of incomplete sentences, more "punchy" Also his retorts were fairly witty occasionally, which is why I usually read through his posts. He also had a tendency to threaten people he disagreed with - you could call that "punchy" as well . He usually went on and on about how Intel wasn't charging enough for its processors, which didn't let AMD make a profit.

BTW, somebody using BM's alter-ID, Christian M. Howell, posted recently on Scientia's blog. No shiny silver suit avatar but you could check that out to observe BM's writing style.

 

[fazers_on_stun]

People keep bumpin' it!

Actually the most ridiculous thread that refused to die was http://www.tomshardware.com/forum/173814-28-very-computer-working by some Canuck named maypep_necro. Went for 54 pages based on his refusal to buy a power cord to test whether or not some ancient Pentium 66 computer that was given to him would work or not. He didn't want to spend $5 to find out, but instead wasted months asking whether or not the computer would work. People even offered to send/buy him a power cord if he would just shut up ...

Just think - only 34 more pages to equal that record!

 

[NMDante]

IIRC BM's posts were much shorter, lots of incomplete sentences, more "punchy" Also his retorts were fairly witty occasionally, which is why I usually read through his posts. He also had a tendency to threaten people he disagreed with - you could call that "punchy" as well . He usually went on and on about how Intel wasn't charging enough for its processors, which didn't let AMD make a profit.

BTW, somebody using BM's alter-ID, Christian M. Howell, posted recently on Scientia's blog. No shiny silver suit avatar but you could check that out to observe BM's writing style.

Ah, yes. How could I ever forget the whole Intel should charge $800 for the E6600, since it beat the older Pentium 4, and since Intel didn't, they were actually using predatory pricing. That was a fun.

Oh, and I do remember his crazy threats of going to NY so he could "stomp" someone. Yeah. Uh, huh.

I also remember when someone "outed" him as a fired MS programmer, and he kept claiming that he left because he didn't like MS. HA.

The "use Styrofoam to pack a CPU" debacle was classic.

 

[jimmysmitty]

Well when I first started on these threads I was a bit of a newbie to the THF community. BM was still here and I swear he was arrogent and annoying. Then he left and we had no one to laugh at. Till thunderman.

Ahhhh good times....

 

[yomamafor1]

BM was actually the dying breed of rabid fanpois during that time (well, until now that is). Anyone who's here long enough remember MMM ,sharikou, and 9-inch. You think BM was rabid? Try MMM. I remembered going through pages without a single meaningful post.

EDIT: Now come to think of it, you guys really got it easy... :lol: :lol:

 

[amdfangirl]

Ah, those days... when I was a lurker... good times... had short cropped hair then...

 

[TechnologyCoordinator]

BM was actually the dying breed of rabid fanpois during that time (well, until now that is). Anyone who's here long enough remember MMM ,sharikou, and 9-inch. You think BM was rabid? Try MMM. I remembered going through pages without a single meaningful post.

EDIT: Now come to think of it, you guys really got it easy... :lol: :lol:

LOL, yup, I remember them.

I think 9-inch is Thunderman. The style is similar; unleash offensive\idiotic payload and then sit back and watch without participating further.

 

[BadTrip]

mike was the best, i have never seen anyone get so mad over forum posts. off topic whatever happened to Action Man

 

[jimmysmitty]

I think 9-inch was on the verge of leaving when I came in. I remember a few post by him too. Crazyness.

 

[NMDante]

reported. see ya later.

 

[amdfangirl]

Guys, this is really annoying! All this spam!

*comes up with idea*

wait abit...

 

[kassler]

here is another review showing how AMD pulls ahead on higher graphics.
http://www.extremetech.com/article2/0,2845,2320536,00.asp

 

[NMDante]

Yeah, compare a dual core to the triple core.
You got us there.

And it sure as heck crushed the dual core pretty badly, huh? What was the average...about 1fps in low res and 3.75fps in high res? Whoo hoo!

 

[thunderman]

I have many years experience and expertise…I feel confident when I claim AMD is superior. Intel CPU’s are dinosaurs. Feel the smoothness of using AMD….you’ll never turn back.
Imagine a Phenom 9950 plus four 4870 GPU’s Quad Crossfire spider platform…….Nothing will keep up period. The spider platform is unique….Intel does not offer anything similar.

 

[keithlm]

Yes... how dare you compare the X3 8750 that sells for $175.00 to the E8400 that sells for $189.95.

Oh the nerve. Actually showing something working better than an Intel chip. And it even has a lower price. Oh the travesty.