E8600 beats out Q9550? Help quickly!!! Need to send back.

[Guest]

Drastically... drastic is like 10 fps... not .4 fps....

anyway whats your resolution... if its high t won't matter


yes keep the q9550

and @ habitat87

your an idiot... i have a C0 stepping q9550 ... not the newer and better E0 revision one... and i am at 3.85 ghz at 1.25 volts.... ON AIR...

So yeah its possible... yeah OF F*CKING COURSE your not going to be able to OC as far... its got 2 more cores... He had valid points and now you look like the idiot... i'd like to see your f*cked up rebuttel of this...

 

[Spitfire7]

Drastically... drastic is like 10 fps... not .4 fps....

anyway whats your resolution... if its high t won't matter


yes keep the q9550



ok not drastically, but still lower frames for what should be increase by OCing right? I have my res at 1680x1050. Is my GPU bottlenecking?

 

[jerreece]

Drastically... drastic is like 10 fps... not .4 fps....

anyway whats your resolution... if its high t won't matter


yes keep the q9550



ok not drastically, but still lower frames for what should be increase by OCing right? I have my res at 1680x1050. Is my GPU bottlenecking?

The change in your FPS you showed in your post is really no change at all. It's so minor. If you ran that benchmark 3 times you'd probably get slightly different results each time, much like the results you got.

So basically, your CPU was not bottle necking your GPU in the first place. That's why you did not receive any gains in FPS from OC'ing, cause your CPU was already keeping up with the GPU.

 

[Zorg]

Right a half a frame, let it go.

By the way, and more importantly, what are your Prime95 small FFT temps?

 

[Zorg]

@ habitat87

your an idiot... i have a C0 stepping q9550 ... not the newer and better E0 revision one... and i am at 3.85 ghz at 1.25 volts.... ON AIR...

So yeah its possible... yeah OF F*CKING COURSE your not going to be able to OC as far... its got 2 more cores... He had valid points and now you look like the idiot... i'd like to see your f*cked up rebuttel of this...

A little stressed?

My Q6600 OCs to 4.6G, and that's with no HS at all. :lol:

Let it go, it's not worth it.

 

[Spitfire7]

Right a half a frame, let it go.

By the way, and more importantly, what are your temps?

My temps for 3.4Ghz are under load at 47 degrees and idle at 42. They pretty much stay like that each time. Is that good, bad, or great?

 

[Spitfire7]

For idle it's a bit high but load is good. 5 degree temp differences between idle and load? Ummmm... Anybody want to comment on that also?

Habitat87, I know you are being sarcastic, but to a noob like me, what did you mean by "want to comment on that"? Were you saying 5 degree difference is a good thing or a bad?

 

[Spitfire7]

ok I have reduced idle temps instantly. My CPU halt state was disabled so I enabled it which brought my volts from 1.3v constantly down to 1.1v at idle. Also, my CPU multiplier drops to 6.0x rather then always staying at 8.5x at idle. Its better now. Dropped by about 5 degrees instantly after running all day. I should notice more of a drop after my computer is off for a while or over night.

 

[jerreece]

LOL! WTF! It's all drawing from the same cache. Only the previous gen chips had this feature, and currently AMD I believe. Also, it will say x2 or x4 with the actually sizes split accordingly. How the hell did you come up with that? Yeah, if your using TWO processors in the quad it's still using 12 MEGS of cache.

My statement was based on my knowledge of the Q6600 chipset. Which is basically 2 x Dual-Core chips. Each dual-core chip has 4MB of cache. However, the total of 8MB cache cannot be shared between individual cores.

http://www.sharkyextreme.com/hardware/cpu/article.php/3652811

Although the Kentsfield does include two 4MB Core 2 processors, it also means that many Core 2 features are not fully realized. Core 2 processors feature either 2MB or 4MB of Advanced Smart Cache, which is shared between the two cores. This allows the Core 2 processor a great deal of flexibility, and it could allocate all 4MB of L2 cache to a single core in gaming scenarios, while dynamically shifting the L2 cache between cores in a multi-threading environment. This type of flexibility is not present in the Kentsfield core, there is no facility for sharing of the 8MB of total L2 cache, and instead it acts like dual Core 2 processors, each sharing 4MB of L2 cache. This also affects the entire processor, with absolutely no shared resources between the dual processor dies, resulting in duplication between the two distinct processors.

There are still some advantages to the architecture. The entire L2 cache is not shared, but there is still a total of 8MB, with each Core 2 unit able to dynamically allocate its own 4MB share. The Core 2's L2 bus is fully 256-bit, and the L1 cache is 32KB instruction/32KB data caches per core, and each has 8-way associativity. The Core 2 includes support for Intel SpeedStep technology, and so too does the quad core Kentsfield. This is a huge advantage, as with four cores under the same roof, power management is a major concern. The Kentsfield also supports features like Execute Disable Bit, Intel 64 Technology, SSE4, and Intel Virtualization Technology, among others.

Now that's for the Kentsfield, including my Q6600. The Q9550 is a Yorkfield chip. From what I can find so far, the Yorkfield (which are Penryn) chips are no different, just have larger cache and a faster clock cycle. So far it appears the architecture is the same as the Kentsfield, only a die shrink (65nm to 45nm). Both are still 2x Dual-core chips. And each dual-core splits its own cache, but cannot share its cache with the other on board dual-core.

I'd look more but I'm at work right now. If you can find something to show that my earlier post was wrong I'd be interested to see it and correct my knowledge. 😉

 

[Guest]

Habitat... Do you want to see it as a fact? I can and I will... but i won't as its a waste of time and energy proving to you that I can...

I have a q9550 on a P5Q Deluxe with a FSB of 453 GTL voltages set to auto (haven't messed with them... could definitely get more out of the processor though) 1.25 voltcore 1.36 northbridge

And a Core Contact Freezer Pro Cooler... load temp is 70 c in a 30 c ambient room....

Why would i lie about it? And also why would i regret going from a e6420 to a q9550... I noticed a MASSIVE increase in performance... not only is my computer noticeably faster in games... it doesn't get all choppy when doing lots of multi-tasking

Seriously habitat... your an idiot

 

[jerreece]

No, you got it wrong, I actually wanted to hear what someone had to say. Seriously, a 5 degree difference between idle and load. You ever hear of that before? That's why I asked what cooler you had also.

My question would be, are we talking 100% load on all 4 cores? Or load as in.... opened Firefox and downloaded something. Realistically, if you only go from 42C to 47C when going from idle to 100% load, you've got an extremely impressive cooling system on that CPU.

 

[jerreece]

However, the total of 8MB cache cannot be shared between individual cores.

Right, I'm pretty darn sure that's what I said in the first place. So what on earth are you arguing about?

 

[Zorg]

My temps for 3.4Ghz are under load at 47 degrees and idle at 42. They pretty much stay like that each time. Is that good, bad, or great?

I am loving this feud... it gets the blood flowing.

Are these Prime95 numbers with Core Temp monitoring? I think not.


Listen, I don't care if you cook your CPU. I will help you if you care.

Give me the data I asked for, or not.

 

[Zorg]

ok I have reduced idle temps instantly. My CPU halt state was disabled so I enabled it which brought my volts from 1.3v constantly down to 1.1v at idle. Also, my CPU multiplier drops to 6.0x rather then always staying at 8.5x at idle. Its better now. Dropped by about 5 degrees instantly after running all day. I should notice more of a drop after my computer is off for a while or over night.

I was enjoying the fight and I just saw that you were searching and learning.

Very good.

So Where are my temps?

 

[Zorg]

My fault.,

Run Prime95 small FFTs and Core Temp, then posts your temps.

 

[jerreece]

Here, let me help you. Look at the way they show the cache on these three cpu's.

http://www.newegg.com/Product/Product.aspx?Item=N82E16819115017
http://www.newegg.com/Product/Product.aspx?Item=N82E16819115041
http://www.newegg.com/Product/Product.aspx?Item=N82E16819103244

What's your point? The Q6600 on NewEgg.com shows 2x4MB cache. The Q9550 simply says "12MB", and the AMD Phenom (which was never a part of our conversation) says L1 cache 4x128KB, L2 cache 4x512KB, L3 cache 2MB. Again, we're talking about Core 2 Duo vs Core 2 Quad in this thread. Or did you somehow miss that?

I'm not denying saying anything. You're arguing against nothing. Here you come into this thread blowing steam about how mighty you are, and yet you're arguing nothing.

The Intel Quad core processors, are TWO Core 2 Duo chips put together. If Core 2 Duo = 2 CPUS, then Core 2 Quad = 4 CPUS.

A Q9550 has 12MB L2 cache. Half of that (6MB) is allocated specifically to ONE of the Core 2 Duo chips (which is TWO processing units). The other half (6MB) is allocated specifically to the SECOND Core 2 Duo chips (which is also TWO processing units).

If current games are only programmed to take advantage of Dual-Core chips, and not quads (therefore use TWO processing units, and not FOUR), then that game will only be using 6MB of that L2 cache in the first place. The second half of the L2 cache (the extra 6MB of that 12MB total) cannot be used by the game you're playing.

Therefore, the difference between a 6MB L2 cache on an E8600 vs 12MB L2 cache on a Q9550 won't matter, in a game which is only capable of using two cores since the Intel Quad Core processors (not AMD mind you, since we never mentioned AMD) do not allow for the two Core 2 Duo chips within the Quad core to share L2 cache with one another.

Therefore, shut up with your pointless argument over nothing.

 

[jerreece]

However, the total of 8MB cache cannot be shared between individual cores.

I know this, that's a feature of that chip gen. The specs on the q6600 actually does say 4mbx2, and yes, they did split it between the two duals in the processor. As for why, I asked myself the same thing, I guess it was better that way.

And BTW, I love how you edited/changed this post AFTER I quoted it and replied to it.

 

[Zorg]

where are my temps?

 

[Homeboy2]

Bottom line is whether you buy a E8500 or Q6600, aint a dimes worth of difference

 

[Silverion77]

All i can say is wow....i leave for 5 hours and this happens...

The only thing i see about Habitat is:
1) hes a fool
2) Hes a troll that loves dual cores
3) He knos little of other's experiences. Habitat go to f*cking xtremesystems and look at their OCs for a Q9550. DO SOME DAMN RESEARCH
4) "LOL!" is his trademark line
5) i rly feel like hes just jealous of quad owners....i mean he thinks we all regret the quad. Why in hell would i regret a quad?? Cause i dont have a cheaper dual. So i can OC and get 3 fps?! O right we did prove that its 3 fps. Lets do another test. Lets run Norton Antivirus, fraps, and hmmmm what another hogging program...O i kno Extracting a file thats good...on a dual and a quad while gaming and lets see what happens...

 

[Zorg]

You guys are all responsible for hijacking this thread, about something that is not that important.

That gets in the way of my request for Prime temps.

 

[Zorg]

THG is messing around. They have the right.

 

[Zorg]

Temps?

 

[Zorg]

Remember several months ago when everyone was saying what habitat87 is saying now. So he is a little behind the curve. Let it go.

spitfire7, I can't answer your question unless you give me the Prime Temps.

Can you do that, or should I just move on?

Or you didn't need an answer?

 

[Zorg]

Well, things have changed a lot in the last few months now hasn't it? Actually, I pulled that from a recent thread. He mentioned that quads are in a bad position as of now, so that's got to be recent. The nail in the coffin is the i7 release.

I want see some links though. They can't talk all that **** and not give links.

What you are saying now has been beaten to death for at least six months. He!! I argued this before the Q6600 came out. It's a very very old argument and you are are on the wrong side at this point. It's so old that it's boring. Think of something new.

Although I have to admit, you did get some takers, and a bad rep in the process.

I just recommended a dual core very recently. It's not a war, quad is the way it's moving, it's not a big deal.

As you said, i7 is the future and we are all on board, after we see the OC results on mobos that are available and prices come down.


I am personally holding out for the Octo HTT, 16 virtual cores.

I'm guessing that the developers are going to accelerate the multicore optimization... or die.

And a little Ray Tracing anyone? That might actually get me excited again.