Sign in with
Sign up | Sign in
Your question

Q6600 system vs AMD Phenom system - Page 2

Last response: in CPUs
Share
July 1, 2008 1:22:28 PM

Amiga500 said:
Eh?

So AMD does have an advantage running server software on many cores using alot of memory...


but "is not better at multithreaded applications"?



I run CFD software on a Q6600, Phenoms would be much quicker. Sure, its not your normal run of the mill desktop workload, but there ya go.



Correct.
It has nothing to do with Mult-Threaded Applications.
And it has ZERO to do with the Phenoms or Q6600s which are not installed in such servers. :pfff: 



July 1, 2008 1:27:30 PM

Zorg said:
AkisTzortzis,

Close, your a little mixed up. I'm not going to go through your post point by point. At 400 x 9 your CPU will be running at 3.6G. That will take more knowledge then you presently have to get a stable cool OC. Stick with the parts list you have in your 1st post and follow the settings I gave you in my 1st post and you will be at a nice mild OC to 3G. I would buy the DDR2 800 so that later as you learn more you can OC the FSB/CPU further without being forced to OC the RAM beyond DDR2 667.

At the 3G/333 (1333) OC You can run the RAM at 1:1 667 or 5:6 800 (a setting of 2.4 in the BIOS of that mobo).

I will also recommend an aftermarket HS with a backplate again.


If you can't understand what I'm saying you might want to read the OC stickies at the top of the CPU OC forum. You probably should read them anyway.

PS never start a thread with "Q6600 system vs AMD Phenom system". It's a guaranteed flame war and you will get lost in the mix.


OMG! A Rational Post :>

When ordering your Q6600, try someplace that promises a G0 stepping, such as Frys.
While you are very likely to get a G0 most places, not point in not ensuring you get one.

I also agree that you likely don't want to aim for 3.6Ghz.
While it can be done, I never try to push systems close to their limit.

3.0Ghz will be an EZ OC.
You can likely hit 3.2-3.4 with moderate effort.

3.6-3.8 are considered very Good OCs, but you will not see a huge difference from 3.2-3.4
July 1, 2008 1:29:11 PM

zenmaster said:
Correct.
It has nothing to do with Mult-Threaded Applications.
And it has ZERO to do with the Phenoms or Q6600s which are not installed in such servers. :pfff: 

If I understand you right. You beleve that Intel does EVERYTHING better on desktops?
Related resources
July 1, 2008 1:33:41 PM

Zorg said:
Give AMD a break, with a little slack from Intel they could make a showing. Let's hope so anyway.


Well, I did say their GPU department is doing well.
Their motherboards are decent as well.

While I hope their CPU department gets better since it will be good for all of us, it does not change
the fact the author wrote an article based on baseless facts. And even if the baseless facts turn out to be true, they still do not support the conclusions.

July 1, 2008 2:02:25 PM

Get the 6600, read the hundreds of reviews on Newegg about it, these are people who have actually used it. (I like reading the bad reviews too as sometimes you can gather some information about possible problems you may have with your build.)
July 1, 2008 2:08:14 PM

zenmaster said:
Correct.
It has nothing to do with Mult-Threaded Applications.
And it has ZERO to do with the Phenoms or Q6600s which are not installed in such servers. :pfff: 


Sorry run that by me again.


The march demonstrates better scaling with threads on servers, I would get better performance from a Phenom than my Q6600...


but it has nothing to do with multi-thread apps? :??: 
July 1, 2008 2:30:14 PM

I think he means that servers are servers and desktops are desktops, yes both run the same kind of applications but at the same time you would not ask a desktop PC to be a full time server running those applications (and if you did you would know what your getting into...)

Off topic analogy time!!! lol Its like when a trucker tells you his truck gets great gas millage and you ask how many miles to the gallon ? he says 8!!! omg... Cars are not trucks and trucks are not cars, different purposes.
July 1, 2008 2:44:09 PM

Amiga500 said:
Sorry run that by me again.


The march demonstrates better scaling with threads on servers, I would get better performance from a Phenom than my Q6600...


but it has nothing to do with multi-thread apps? :??: 


Very Simple - Opterons and XEONS are the Server Processors.
You will not find Phenoms or Q6600s in the 4-way servers that may show the difference.

Also, the Issue is memory band-width and not multi-processing.
And the memory band-width will only be seen on certain tasks.

We use Opertons for our VMWare ESX Servers for this reason.
We use XEONS for most other servers.

There are many links to many reviews that show the Phenom loses virtually every real-world benchmark test.
I don't recall ever seeing one the operton has won, but there could be one.

Please note the poster who makes the claim for the Phenom could not find any to support his claim either.
July 1, 2008 2:47:37 PM

kassler said:
If I understand you right. You beleve that Intel does EVERYTHING better on desktops?


Well, I think ATIs discrete GPUs beat Intel's Integrated Graphics.............
July 1, 2008 2:57:52 PM

zenmaster said:
Very Simple - Opterons and XEONS are the Server Processors.
You will not find Phenoms or Q6600s in the 4-way servers that may show the difference.

Is it possible to run a xeon on the desktop or Q6600 on a server? If you switch what is the difference?
a b à CPUs
July 1, 2008 2:58:28 PM

Q6600 @ 4.0ghz.
July 1, 2008 3:01:36 PM

zenmaster said:
OMG! A Rational Post :>

When ordering your Q6600, try someplace that promises a G0 stepping, such as Frys.
While you are very likely to get a G0 most places, not point in not ensuring you get one.


The processor he listed is the G0 (SLACR) version.

Quote:
CPU : Intel Core 2 Quad Pro Q6600 "Energy Efficient SLACR 95W Edition" 2.40GHz (1066FSB) - Retail ---- 124 pounds

a c 127 à CPUs
a b V Motherboard
a b À AMD
July 1, 2008 3:02:54 PM

Amiga500 said:
Sorry run that by me again.


The march demonstrates better scaling with threads on servers, I would get better performance from a Phenom than my Q6600...


but it has nothing to do with multi-thread apps? :??: 


The difference being is that kessler is trying to state that a Q6600 cannot multitask. In multi-threaded apps I have seen that Phenoms design benefits it. Although I have yet to see it benefit it in Source based games as that game engine does use multiple threads.

Multitasking and multithreaded are two different things. I do find it funny though that he does not own a Q6600 but based on the info he has read and his assumptions a Core 2 Quad cannot multitask. Yet I have one and on a daily bassis run multiple apps while gaming but thats not enough for him.

My main problem here is the OP had decided on a CPU and was asking for information on it and what was the best this and that for him to get. Then kessler has to come in and blab his BS. It was all fine and dandy till then.

To the OP, for the Q6600 all you need is PC6400 (DDR2-800MHz). You can get PC8500 (DDR2-1066) but its not needed. I would recommend trying to get either one (as the cost is very low for 2GB or 4GB) with at least 4-4-4-12 as the latencies. I have 4GB of the Corsair XMS2 PC8500 that is running at 1333MHz (1:1 with the CPU FSB and the normal timings were 5-5-5-15 but are running at 4-4-4-12 and 12 hours Prime95 stable. I highly suggest Corsair.

And good luck with your build.
July 1, 2008 3:12:51 PM

kassler said:
Is it possible to run a xeon on the desktop or Q6600 on a server? If you switch what is the difference?


Certain Xeons are basically just core2 clones and are made to run on socket 775 (Xeon E3110 for example) but most run on server boards using LGA771.

You can't drop a Q6600 in a server board for several reasons but the fact that they use a different socket should be a good first indicator.


The difference is that they are different chips tweaked to perform certain functions based on their expected use.

Please stop with the multitasking nonsense when you have nothing to back it up.





July 1, 2008 3:29:48 PM

Thanks, what is the gain of running the FSB at 333 MHz but the RAM at 400 MHz ? Is there a gain?
a c 127 à CPUs
a b V Motherboard
a b À AMD
July 1, 2008 3:38:04 PM

akistzortzis said:
Thanks, what is the gain of running the FSB at 333 MHz but the RAM at 400 MHz ? Is there a gain?


Yea. Some people say that running the memory at a faster FSB than the CPU increases the memory bandwidth and allows for a bit more performance. But the only down side is it is not as stable as a 1:1 ratio.
July 1, 2008 3:49:40 PM

zenmaster said:
Very Simple - Opterons and XEONS are the Server Processors.
You will not find Phenoms or Q6600s in the 4-way servers that may show the difference.

Also, the Issue is memory band-width and not multi-processing.
And the memory band-width will only be seen on certain tasks.

We use Opertons for our VMWare ESX Servers for this reason.
We use XEONS for most other servers.

There are many links to many reviews that show the Phenom loses virtually every real-world benchmark test.
I don't recall ever seeing one the operton has won, but there could be one.

Please note the poster who makes the claim for the Phenom could not find any to support his claim either.



The difference is evident in a lot more than 4 way servers.

Memory bandwidth and balancing latency is an issue for parallel processing (and no, you don't have to have a Xeon or Opteron to run MPI or PVM stuff).




The FSB and memory architecture can be a severe limitation on the performance of any Intel quad.

For instance, CFX has a speedup factor of 1.38 for a 2 thread run on CPU 0 and CPU 1 on a Xeon 3.0 1666 MHz FSB.

If that was CPU 0 and CPU 4 (i.e. no contention for cache space or data until FSB) the speedup is 1.98.




Now when I have a 4-way run on my Q6600 I a speedup of only around 1.6 (contention for cache and bus bandwidth) - a phenom would scale by a factor over 3.
July 1, 2008 4:14:39 PM

Depends on the Application to see the amount of scaling.
Some apps scale nearly 100%.
Others do not.

There are MANY reviews that show that increasing FSB and other Memory settings that increase bandwidth have little to no effect on performance.


Again, Please Link Reviews from a Well Trusted Website that shows the Intel Q6600 performing poorly on multi-threaded apps compared to the Phenom.

There are already links in this thread that show that the Intel does not have any issues.

There are also many OCing tests that show the Q6600 doing increasingly well as it overclocks in mult-threaded apps. If there was a memory wall that would not be the case.

The Q6600 will not hit a memory wall.
The Phenom can't compete, unless you can't the fact I've seen the Phenom 9600 for under $100 as part of some combos.

July 1, 2008 4:42:41 PM

zenmaster said:
There are MANY reviews that show that increasing FSB and other Memory settings that increase bandwidth have little to no effect on performance.


Why not try to read about how they really work?
You can 100 diet programs that say they work but then don’t… If you know how your body works then it is easy for you to know if the diet program works.
When you have learned how the processor works then you will be able to get some information out of the tests.

July 1, 2008 4:47:19 PM

kassler said:
When you have learned how the processor works then you will be able to get some information out of the tests.


This coming from the genius who wants to drop a Q6600 in a server board.
a c 127 à CPUs
a b V Motherboard
a b À AMD
July 1, 2008 4:54:46 PM

uguv said:
This coming from the genius who wants to drop a Q6600 in a server board.


:lol:  :lol:  :lol:  :lol: 

And he doesn't even own a Q6600. He just makes claims based off of info I am sure he got off of AMD fanboy sites hence why he probably doesn't post any links.
a c 113 à CPUs
a b V Motherboard
a b À AMD
July 1, 2008 4:56:27 PM

zenmaster said:
~~~
Again, Please Link Reviews from a Well Trusted Website that shows the Intel Q6600 performing poorly on multi-threaded apps compared to the Phenom. ~~~

The Phenom can't compete, unless you can't the fact I've seen the Phenom 9600 for under $100 as part of some combos.


Mainconcept H.264 Encoder
24 sec HDTV 1920x1080 mpeg2 (mpeg2 to H.264)


Phenom 9700 = 63 sec
Phenom 9600 = 65 sec
Phenom 9500 = 68 sec
q6600 = 69 sec
e8500 = 85 sec
e8400 = 90 sec

Tom's CPU Charts good enough for you?
July 1, 2008 5:00:25 PM

uguv said:
This coming from the genius who wants to drop a Q6600 in a server board.

hmmm... is your computer runned by a socket?

if you skip the socket, what differs the xeon from processors used in the desktops?
July 1, 2008 5:03:08 PM

jimmysmitty said:
And he doesn't even own a Q6600. He just makes claims based off of info I am sure he got off of AMD fanboy sites hence why he probably doesn't post any links.


are you also a "link"-man?
a c 127 à CPUs
a b V Motherboard
a b À AMD
July 1, 2008 5:06:21 PM

Wisecracker said:
Mainconcept H.264 Encoder
24 sec HDTV 1920x1080 mpeg2 (mpeg2 to H.264)


Phenom 9700 = 63 sec
Phenom 9600 = 65 sec
Phenom 9500 = 68 sec
q6600 = 69 sec
e8500 = 85 sec
e8400 = 90 sec

Tom's CPU Charts good enough for you?


How is this poorly when in the same scenario where AMD loses by only a few seconds is considered good by AMD fans? No one has said that Phenom does not shine. But the claims he is making are absurd. Its either "Intel CPUs cannot multitask" or "Intel CPUs FSB will hit a wall in games and normal multitasking".

He just needs to stop being a BSer.
July 1, 2008 5:20:35 PM

jimmysmitty said:
"Intel CPUs FSB will hit a wall in games and normal multitasking".

Just to clearify, tetris is a game that will run very smooth on intel.
If the game isn't that demanding then there is no problem.

check this test: http://www.hothardware.com/Articles/AMD_Phenom_X4_9350e...
They are running the game in 800x600...
why not also run it in 1920x1200 when they are on it
a c 127 à CPUs
a b V Motherboard
a b À AMD
July 1, 2008 5:20:58 PM

kassler said:
hmmm... is your computer runned by a socket?

if you skip the socket, what differs the xeon from processors used in the desktops?


There are some technical differences in the chip itself. There are 4 less pins for a few reasons as well. But I do not fully know these differences.

kassler said:
are you also a "link"-man?


No I am not. But it would be nice if you quit spouting your personal opinions and provide some facts. Here is a good analogy for ya. I could spout on and on with my opinion I formed after reading info from a "mysterious" source and sit there like it is fact. I wont provide any proof or, in this case, "links" with proof but I will continue to act like what I say is true even if I have no experience or facts to back it up. Sound familiar? It should. Its you.

You know what the difference is? As I said before I have experience and a lot of benchmarks to prove that the Q6600 is a great perfromer and easily beats Phenom in most normal user apps and games. Its also nice to know that I can easily Overclock my CPU without the need for either a new mobo cuz the one I got was messed up in some way or without having to buy a higher end mobo.
a c 127 à CPUs
a b V Motherboard
a b À AMD
July 1, 2008 5:28:02 PM

kassler said:
Just to clearify, tetris is a game that will run very smooth on intel.
If the game isn't that demanding then there is no problem.

check this test: http://www.hothardware.com/Articles/AMD_Phenom_X4_9350e...
They are running the game in 800x600...
why not also run it in 1920x1200 when they are on it


How about because the majority of people do not run at 1920x1200? Steam takes a sruvey of all the resolutions (and hardware) used by its gamers. The most common resolution was 1280x1024. 1920x1200 came in at the last with very very few people. Either way the higher resolutions tend to rely more on the GPU than on the CPU.
July 1, 2008 5:28:14 PM

Wisecracker said:
Mainconcept H.264 Encoder
24 sec HDTV 1920x1080 mpeg2 (mpeg2 to H.264)


Phenom 9700 = 63 sec
Phenom 9600 = 65 sec
Phenom 9500 = 68 sec
q6600 = 69 sec
e8500 = 85 sec
e8400 = 90 sec

Tom's CPU Charts good enough for you?

That's a nice find, but does it really show the Q6600 not being able to run multi-threaded applications any worse than the Phenom? You are talking about Intel's oldest Quad CPU, here. What is the score for the QX9650 or Q9450? Oh, Tom's Charts don't have the Q9450. But the QX9650 ran it in 51sec. And the QX6700, which is pretty much a Q6700 with unlocked multipler, ran it in 63sec, and it is the 3rd oldest Intel Quad CPU.

I am with jimmysmitty on this one. It isn't a poor showing by Intel, but it does show how AMD's latest and greatest Quad CPUs barely beats Intel's oldest Quad CPU.

kassler said:
are you also a "link"-man?

Seems like he is more of a "show me the facts" type person, than "I will believe whatever you say" type person.
July 1, 2008 5:30:20 PM

kassler said:
are you also a "link"-man?



I'm not sure... are you an "unfounded BS" man?
July 1, 2008 5:31:55 PM

kassler said:
Just to clearify, tetris is a game that will run very smooth on intel.
If the game isn't that demanding then there is no problem.

check this test: http://www.hothardware.com/Articles/AMD_Phenom_X4_9350e...
They are running the game in 800x600...
why not also run it in 1920x1200 when they are on it


I guess looking at charts prevents you from reading:
Quote:
For our next set of tests, we moved on to some in-game benchmarking with Crysis and F.E.A.R. When testing processors with Crysis or F.E.A.R., we drop the resolution to 800x600, and reduce all of the in-game graphical options to their minimum values to isolate CPU and memory performance as much as possible.
July 1, 2008 5:46:59 PM

NMDante said:
I guess looking at charts prevents you from reading:
Quote:
For our next set of tests, we moved on to some in-game benchmarking with Crysis and F.E.A.R. When testing processors with Crysis or F.E.A.R., we drop the resolution to 800x600, and reduce all of the in-game graphical options to their minimum values to isolate CPU and memory performance as much as possible.

But wouldn't it be interesting to check I/O performance? That will be a much more important factor running games. Just by telling that they need to lower the resolution to 800x600 to notice differences is like saying: Processor speed doesn’t matter
July 1, 2008 5:52:40 PM

akistzortzis said:
I think a moderate 775 motherboard + Q6600 is cheaper than a similarly moderate AM2 motherboard + Phenom, and both systems use same 1066 RAM. For example:

AMD system:

motherboard : Gigabyte GA-MA770-DS3 AMD 770 (Socket AM2) PCI-Express DDR2 Motherboard ----------- 55 pounds
CPU : AMD Phenom X4 Quad Core 9750 2.40GHz (Socket AM2) - Retail ------------------------------------------- 141 pounds

total: 196 pounds

Intel system:

motherboard : Gigabyte GA-EP35-DS3 Intel P35 (Socket 775) PCI-Express DDR2 Motherboard ------------- 67 pounds
CPU : Intel Core 2 Quad Pro Q6600 "Energy Efficient SLACR 95W Edition" 2.40GHz (1066FSB) - Retail ---- 124 pounds

total: 191 pounds


******************************


It appears the Intel system is slightly cheaper than the AMD system but moreover, at the stock frequency of 2.4 GHz the Intel quad is quicker than the Phenom (and it overclocks much better).


Am I right on the above or am I missing something?



You are correct.
July 1, 2008 6:01:28 PM

kassler said:
But wouldn't it be interesting to check I/O performance? That will be a much more important factor running games. Just by telling that they need to lower the resolution to 800x600 to notice differences is like saying: Processor speed doesn’t matter


Clearly, you have 0 clue what you're talking about.
July 1, 2008 6:01:40 PM

kassler said:
But wouldn't it be interesting to check I/O performance? That will be a much more important factor running games. Just by telling that they need to lower the resolution to 800x600 to notice differences is like saying: Processor speed doesn’t matter


Then make your own set of benchmarks, or question the site you linked to. They gave an explanation for why they ran the test at that resolution, to "isolate CPU and memory performance as much as possible" since it was a CPU review, and not a system review.

You can keep running in circles all you like, about FSB and I/O performance. The bottom line is simple: Intel CPUs are not limited by anything you claim, and if you want review sites to test at resolutions that only about 13% of gaming population plays at, then go talk to the review sites and complain.

Even after giving you the reason for why the site YOU LINKED ran the test at a certain resolution, you still have to backpedal and twist it to an I/O performance question. Simply unbelievable.

edit -
Oh, and also from your linked review is this lovely tidbit:
Quote:
In comparison to Intel's offerings, the Phenom X4 9950 is about on par with the Core 2 Quad Q6600, trading victories depending on the test.

Hmmm...seems like the ancient architecture and FSB still keeps it on par.
a c 113 à CPUs
a b V Motherboard
a b À AMD
July 1, 2008 6:04:04 PM

Hey, Guys,

I don't like getting involved in the flame bait but when a FUDmeister claims domination in multithreading and parallelism by Intel (which is contrary to the facts) you guys are welcome to obfuscate it any way you like. Let's review ...

Quote:
zenmaster wrote :

~~~
Again, Please Link Reviews from a Well Trusted Website that shows the Intel Q6600 performing poorly on multi-threaded apps compared to the Phenom. ~~~

The Phenom can't compete, unless you can't the fact I've seen the Phenom 9600 for under $100 as part of some combos.


My response:

Wisecracker said:
Mainconcept H.264 Encoder
24 sec HDTV 1920x1080 mpeg2 (mpeg2 to H.264)


Phenom 9700 = 63 sec
Phenom 9600 = 65 sec
Phenom 9500 = 68 sec
q6600 = 69 sec
e8500 = 85 sec
e8400 = 90 sec

Tom's CPU Charts good enough for you?





Quote:
How is this poorly when in the same scenario where AMD loses by only a few seconds is considered good by AMD fans? No one has said that Phenom does not shine.


You need to work on your reading comprehension and reasoning. Please try to keep up.
a c 127 à CPUs
a b V Motherboard
a b À AMD
July 1, 2008 6:05:48 PM

Um you are the one who replied to a question asking to show a Q6600 performing poorly to a Phenom. You failed to show this.
July 1, 2008 6:06:37 PM

NMDante said:
The bottom line is simple: Intel CPUs are not limited by anything you claim, and if you want review sites to test at resolutions that only about 13% of gaming population plays at, then go talk to the review sites and complain.


How do you know that?
Do you have any tests that can verify that I/O isn't important on desktop running demanding games and/or high resolutioins?
July 1, 2008 6:10:27 PM

kassler said:
How do you know that?
Do you have any tests that can verify that I/O isn't important on desktop running demanding games and/or high resolutioins?


Do you have tests that can verify that the I/O causes any issues at high resolutions or demanding apps?
You keep asking for us to find proof to disprove you, yet you come with nothing to prove your own claims.

When you can show us where the I/O comes to play, then show us, until then...you are full of it.
a c 127 à CPUs
a b V Motherboard
a b À AMD
July 1, 2008 6:10:40 PM

kassler said:
How do you know that?
Do you have any tests that can verify that I/O isn't important on desktop running demanding games and/or high resolutioins?


Do you have any test to provie your point? Dude go to Steams website (you know from VALVe) and look at their info that take account of all their players hardware and such. Now these people are probably the most common gamers as they play a wide array of games. You will see that resolutions higher than 1680x1050 are rarely used. Helly even 1680x1050 is a rare one.
July 1, 2008 6:11:37 PM

kassler said:
But wouldn't it be interesting to check I/O performance? That will be a much more important factor running games. Just by telling that they need to lower the resolution to 800x600 to notice differences is like saying: Processor speed doesn’t matter



No.

It is more like saying "We are removing the video bottleneck so we can accurately show you which of these two processors runs this application faster."

I am sure you get that though right?
July 1, 2008 6:16:02 PM

Wisecracker said:
Hey, Guys,

I don't like getting involved in the flame bait but when a FUDmeister claims domination in multithreading and parallelism by Intel (which is contrary to the facts) you guys are welcome to obfuscate it any way you like. Let's review ...

Quote:
zenmaster wrote :

~~~
Again, Please Link Reviews from a Well Trusted Website that shows the Intel Q6600 performing poorly on multi-threaded apps compared to the Phenom. ~~~

The Phenom can't compete, unless you can't the fact I've seen the Phenom 9600 for under $100 as part of some combos.


Quote:
How is this poorly when in the same scenario where AMD loses by only a few seconds is considered good by AMD fans? No one has said that Phenom does not shine.


You need to work on your reading comprehension and reasoning. Please try to keep up.


Not reading comprehension, but I continued past your highlighted red.
Quote:
that shows the Intel Q6600 performing poorly on multi-threaded apps compared to the Phenom.

So, your chart showed the Q6600 being 6 seconds slower than the fastest Phenom (until the 9950).
We simply asked, how exactly is that showing AMD's competitive stance vs. a year and half old Quad core?
Add the Q6700, which is just as old as the Q6600, and the 6 sec lead, disappears.
Not saying the Phenom isn't a good CPU, but as far as competitive, yes - against Intel's older CPUs. Against the newer 45nm Quads...?

a c 127 à CPUs
a b V Motherboard
a b À AMD
July 1, 2008 6:17:07 PM

yomamafor1 said:
Actually you made the claim first, so the burden of proof is on you.

But, of course, let's play nice.

http://techreport.com/articles.x/14424/7

Do you notice any "I/O limitation" you yapped about?


Hehe....I love that you found this. Interesting in the multitasking The 9850BE (with a 100MHz clock advantage) only beats the Q6600 by 3 seconds. Yet a QX6850 (or a Q6600 @ 3GHz) beats the Phenom by more than 50 seconds.

Now wasn't this what you were talking about Kessler? Intel can't multitask better due to the FSB being limited?
July 1, 2008 6:20:08 PM

kassler said:
Before I saw thist test: http://www.overclockersclub.com/reviews/intel_q9450/
I just suspected it. But after I did read that test even I was confused that i could have that big impact.


Suspect what? That desktop programs are almost independent of I/O performance?

Again, you provided no proof to your assertion.
July 1, 2008 6:53:13 PM

yomamafor1 said:
Suspect what? That desktop programs are almost independent of I/O performance?

Again, you provided no proof to your assertion.


check the game tests
July 1, 2008 6:56:51 PM

yomamafor1 said:
Suspect what? That desktop programs are almost independent of I/O performance?

Again, you provided no proof to your assertion.


dI3 1n+el N0o8S 4mD I5 +3H l1GH+

w0Rd, PlAY@.
July 1, 2008 7:55:59 PM

kassler said:
But wouldn't it be interesting to check I/O performance? That will be a much more important factor running games. Just by telling that they need to lower the resolution to 800x600 to notice differences is like saying: Processor speed doesn’t matter


yomamafor1 said:
Clearly, you have 0 clue what you're talking about.



I agree with YoMamma.


[meltdown]
Here's an idea, a more relevant benchmark to gaming performance IS A MOTHER F*CK*NG GAMING BENCHMARK, NOT AN I\O BENCHMARK.


You can't get off artificial memory bandwidth benchmarks because those are among the few that AMD can beat Intel in. Too bad memory benchmarks don't mean crap when it comes to the performance of desktop applications (ie GAMES, etc...) running on DEKSTOP MACHINES.


YARGHHHGHSHAHAHA!

[/meltdown]
a b à CPUs
a b V Motherboard
July 1, 2008 8:18:46 PM

kassler said:
Before I saw thist test: http://www.overclockersclub.com/reviews/intel_q9450/
I just suspected it. But after I did read that test even I was confused that i could have that big impact.

If you look at it, it shows exactly why you do CPU benchmarks at low res - at higher resolutions, everything tends to pretty much even out as far as the CPU is concerned, because it is limited by the graphics card. The one thing that puzzles me there are the Company of Heroes numbers. I'd have to look more into that, as those do not make sense at all.

The other interesting thing to note: in all tests except Company of Heroes, whenever there is a significant difference between the Phenom and the Q9450, it is always in the Q9450's favor, sometimes hugely so.
!