Intel Q9450 vs Phenom 9850 - ATI HD3870 X2

[gosh]

Sorry i cant share your opinion. German reviewers are not more or less positive then US Sites. Maybe we count more on Price/Performance and we dont go with the attitude that a 5 % Performance penalty makes the different between the most sucking piece of hardware and the world best super duper silicon you can get.

Sorry, that was what I meant. I have only read some in german forums, and the discussion there is much more "in the reallity".

Finding one review using AMD with high performance video cards on high res is impossible. The test on overclockers is the only I have found. Other than that I have seen some results from forum members but there is alwayas difficult to know exactly but they have all pointed to the same direction.

The test that jack did showed the same
http://www.xtremesystems.org/forums/...5&postcount=33
http://www.xtremesystems.org/forums/...9&postcount=35
http://www.xtremesystems.org/forums/...1&postcount=62
http://www.xtremesystems.org/forums/...9&postcount=73

[JumpingJack]

http://www.overclockersclub.com/revi...l_q9450/12.htm
http://www.overclockersclub.com/revi...l_q9450/14.htm
http://www.overclockersclub.com/revi...el_q9450/8.htm

Yeah, I will try to get to this tonight... if not definitely tomorrow night.

[JumpingJack]

@Jack

Just wanted to remember you clarifying the numbers given from overclobersclub with your test systems

http://www.overclockersclub.com/revi...l_q9450/12.htm
http://www.overclockersclub.com/revi...l_q9450/14.htm
http://www.overclockersclub.com/revi...el_q9450/8.htm

Ok.... so I did WIC for tonight. This is not a good compare the following is different.

OS: I am using XP, he is using Vista Ultimate (does not specify 32 or 64 bit)
DX: I am using DX9, he is using DX10
Graphics card: I am using a 8800 GTX, he is using an 8800 GT -- drivers are most likely different (I am using Forceware 169.21 -- an older version)
Monitor: I am using a max of 1650x1080, so no 1900x1200 data (I can generate this, but need to swap monitors)
Phenom clock speed: I left mine at 2.5 GHz (default for 9850), he is using B2 2.3 GHz Phenom, doesn't really matter really he is comparing all his processors with games that are GPU bound anyway.

Also, if you are accustomed to my other screen shots... --> I put the taskbar to autohide so I could see the entire WIC screen in the 1280x1024 mode.

I do match his different resolutions (except 1900x1200), very high graphics settings 0x AA and 16X FA. After my first run on 2.5 Ghz / DDR2-800 for the Phenom, I also ran DDR2-1067 so you will see two sets of screen dumps for Phenom.


QX9650 @ 2.67 GHz (333x8) DDR2-1067 8800 GTX in order of resolution



Phenom 9850 @ 2.5 GHz (200x7.5) DDR2-800 8800 GTX



Phenom 9850 @ 2.5 Ghz (200x7.5) DDR2-1067 8800 GTX



QX9650 @ 2.67 GHz
1024x768. Max = 123 Ave = 50 Min = 21
1280x1024 Max = 104 Ave = 47 Min = 22
1650x1080 Max = 95 Ave = 46 Min = 21

Phenom @ 2.5 Ghz DDR2-800
1024x768. Max = 69 Ave = 27 Min = 10
1280x1024 Max = 69 Ave = 27 Min = 10 (this is odd, exactly the same)
1650x1080 Max =70 Ave = 28 Min = 10

Phenom @ 2.5 GHz DDR-1067
1024x768. Max = 67 Ave = 28 Min = 11
1280x1024 Max = 68 Ave = 27 Min = 9
1650x1080 Max =72 Ave = 29 Min = 9


Using DDR2-1067 vs DDR2-800 did not help the Phenom ... I will do a DDR2-800 run on the QX9650 tomorrow.

EDIT: If you are interested, here is WIC with all the GPU restrictive assets removed:

This is much much less of a fair compare.... all setting set to low, only the CPU affecting items were enabled or set to high, for example physics quality is set to high. A full disclosure of settings is available upon request.

QX9650 @ 2.5 GHz, DDR2-800 (in this case, all my baseline data is there)



Phenom 9850 @ 2.5 GHz, DDR2-800



Second Edit: He major fubar'ed his 9450 setup:

http://www.overclockersclub.com/revi...images/666.htm this probably would not affect his results, but I am surprised he made a stable system. He mixed PC-6400 memory with PC-8000 memory. Shockingly he seemed to have been able to get it to run at 463 Mhz speed, or DDR2-926 speed wow. Obviously he relaxed his timings from factory to hit this.

jack

[JumpingJack]

No I didn't. You know that AMD was very strong for those that built their own computers a couple of years back. They had one big market share for home users and maybe that processor was the big money maker for hardware sellers.
The impact that internet had then wasn’t that big as it is now. When anandtech presented numbers about Nehalem the response was enormous. More and more companies are using internet as their main marketing area. I don’t think that the Intel marketing section skips this area.
It isn’t hard to draw these conclusions if you think about it. Quad users are very low compared to other users. This market is also not comparable with other markets. It is hard to find markets that only have two contenders worldwide. If Intel and AMD hadn’t been US companies I can assure you that actions had been taken because we all need competition.

It is also interesting to see that the reviews made in English differ some with reviews in other languages. It seems that German sites are much more positive to AMD compared to US sites.

Actually, AMD still has a huge marketshare in retail. They have historically averaged 40 to as high as 80%.

I don't speak German, but does it surprise you that German sites will be more amenable to AMD products -- I mean if I were a German site reviewing AMD products I too would feel compelled to give positive impressions, if for anything, to help promote a friendly partner ... but data is data is data. So long as it is reproducible, the truth is the truth, so long as they give all the info to reproduce the information then they should be getting similar results.

EDIT: Internet sites also have a degree of accountability if they consistently fudge or over inflate the numbers or skew the information, people soon catch on and soon readership drops off, hits go low and advertising dollars/euro's will start to diminish.

The best way to judge for yourself is do what I did, simply build two systems side by side ... it is quite a nice way to learn more about the topic of computing. I bought 2 Phenoms myself, for many reasons, one reason is the architecture is still fascinating, and different. Comparing and contrasting and experimenting is one way to dig deeper into the rigs.

Now I agree -- the internet is a commercial space, and it would be foolish for companies not to use the internet for such purposes, but I seriously doubt Intel planted the Nehalem samples with Anand ... it is not good to put work in progress out. Intel stands to hurt itself more than anything letting Nehalem out in the wild through an Osbourne effect.

Though I cannot prove they didn't, nor can you prove that they did. This is a very Scientia like position to take

Similarly, one could argue the leaked Deneb data was also planted by AMD ... don't but I would make the same argument against that as well. I think what happened is some tier 2 supplier/MB maker/OEM floated an engineering sample under the table in both cases.

Who knows ... who cares .... but without concrete proof I would hold off accusing either AMD or Intel for using these people as puppets.

But again, go back before March 2006 and count the number of sites that used AMD to review video cards ... why did they do that? AMD certainly did not have the money compounded to pay them off, yet every last one used AMD CPUs to review video cards ...

[gosh]

JumpingJack: So you think that there isn’t any interest among readers to know how video cards perform using AMD processors?
One thing that you can be certain of next time nVidia or ATI is going to release a new card is that there will be at least 10+ reviews using intel processor to test the card. Good imagination among reviewers don't you think?

[JumpingJack]

JumpingJack: So you think that there isn’t any interest among readers to know how video cards perform using AMD processors?
One thing that you can be certain of next time nVidia or ATI is going to release a new card is that there will be at least 10+ reviews using intel processor to test the card. Good imagination among reviewers don't you think?

No, there are plenty of people who would like to see it no doubt, me being one of them.

But the reviewers are reviewing video cards, not processors. If they used a Phenom, most likely at best it would show the 4870 X2's tied with the nVidia 280 and 260 since the games will all bunch up to the limit of the CPU.

They use the Intel CPU not to advertise Intel CPUs, the use the fastest available platform so that the cards will demonstrate the performance differences without any CPU bottlenecking. Even with Intel CPUs, there is some bottlenecks being shown through the reviews.

Many sites have to use over-clocked Intel quads to make sure that the CPU does not rail to the same value. Even the nVidia 280 GTX can open up some bottlenecks.

I have two 4870 X2s on the way, I will provide you examples of this. At the same resolutions, settings, etc. The phenom paired with the 4870 X2s will most likely produce lower scores in some (likely most, if not all) the test runs compared to the Intel CPUs.

It isn't about imagination, it is about doing the right thing and reviewing the cards correctly. Just like before the C2D was launched, all the reviewers used AMD processors to review video cards, not because they liked AMD or wanted to promote AMD, but because the Athlon 64 just B!tch-slapped the Pentium 4 up, down and sidways at gaming... why would you use a P4 to run a high end game with a high end video card. It would be pointless... it is just that today, that role is reversed.

[Loser777]

I don't get what your argument is, gosh... Are you saying that bandwidth is the deciding factor in CPU performance? I doubt that a 386 would be wicked fast (by today's standards) even if we got it running on a HT 3.0 equivalent bus. Intel must be pretty foolish to sell its CPUs at such a high price when they don't perform as well as Phenom eh?

If you can show me an example of a Desktop PC being bottlenecked by FSB, your argument might have some merit.
On the opposite side, there are plenty of examples of games being bottlenecked by CPU... Try dropping in a 3000+ and a 4870X2 and run Crysis. You'll get the same FPS from 640x480 to 1900x1200. Your arguments about Intel's problems with FSB are looking like a bunch of garbage because you have no concrete proof.

AMD must be so unlucky because SOFTWARE IS PROGRAMMED TO FAVOR INTEL. Even with 100% LOAD on all cores, I fail to see one example of where Phenom is faster.

[JumpingJack]

Wooop.... 4870 X2 is gonna be here tomorrow:

[JumpingJack]

Follow up on WIC...

here is Phenom 9850 @ 3.0G DDR2-1067 with the Overclockers Club settings (i.e. very high settings 0X AA 16X AF), same deltas apply though (OS, graphics card etc)



Here is the QX9650 @ 3.0G, all else the same as the Phenom:



Phenom @ 3.0 GHz DDR-1067
1024x768. Max = 81 Ave = 33 Min = 12
1280x1024 Max = 72 Ave = 32 Min = 13
1650x1080 Max =75 Ave = 32 Min = 13

QX9650 @ 3.0 GHz DDR2-1067
1024x768. Max = 112 Ave = 47 Min = 22
1280x1024 Max = 101 Ave = 46 Min = 21
1650x1080 Max = 92 Ave = 45 Min = 23

So even on an 8800 GTX, OCC's settings would never let you see any difference between any CPUs... the phenom shows about 4 FPS improvement, so there is some CPU interactions going on.

[Hornet331]

wow... i didn't realize what a bottleneck a 2,5ghz phenom is for wic... from 1024x768 to 1650x1080 totaly cpu bound....

But its a bit strange to see that the phenom gets higher avarg fps the higher the resolution goes up, sure one fps is within the margin of error, but non the less its kinda odd to see.

Jack would you mind to run this test several times, e.g. five times or so, and give us the avarage of the runs? But only if it is not to much work. Thx in advance

[JumpingJack]

wow... i didn't realize what a bottleneck a 2,5ghz phenom is for wic... from 1024x768 to 1650x1080 totaly cpu bound....

But its a bit strange to see that the phenom gets higher avarg fps the higher the resolution goes up, sure one fps is within the margin of error, but non the less its kinda odd to see.

Jack would you mind to run this test several times, e.g. five times or so, and give us the avarage of the runs? But only if it is not to much work. Thx in advance

I have low-res stuff multiple times, it takes alot of work to generate screen shots so for a first pass, I will simply quote the numbers. I will do it later tonight.

Jack

[gosh]

About World in Conflict

http://www.yougamers.com/articles/45...terview-page6/

YouGamers: Let's talk about high end. Do you have a threaded engine for multicore processor support?

Westberg: Yes, we do. On the CPU side, Intel has been very supportive helping us out, because it's a big step moving to a threaded architecture. We've been working with them, and we scale quite well, but if you have a quad-core you won't run the game four times as fast, because it's really hard to reach that. Also, if you have a quad-core, each of those four cores is pretty fast, and we still have to scale down to this 2 GHz machine that's our low-end spec for everyone to be able to run the game.

So we have this 2GHz processor here, and then we have four CPU's that are twice as fast, so we have eight times more [processing power] over here. It's hard to scale all the way [across that]. It does scale, so a dual core runs faster than a single core, and a quad-core runs even faster.

What we do thread is the entire physics update on a separate thread, we thread our shadow volume updates, [...] particle updates and tree updates. And then the obvious things like everybody is probably doing, like sound, voice over IP and things like that. But the four I mentioned are the [threaded processes we render] frame to frame.

[JumpingJack]

About World in Conflict

http://www.yougamers.com/articles/45...terview-page6/

YouGamers: Let's talk about high end. Do you have a threaded engine for multicore processor support?

Westberg: Yes, we do. On the CPU side, Intel has been very supportive helping us out, because it's a big step moving to a threaded architecture. We've been working with them, and we scale quite well, but if you have a quad-core you won't run the game four times as fast, because it's really hard to reach that. Also, if you have a quad-core, each of those four cores is pretty fast, and we still have to scale down to this 2 GHz machine that's our low-end spec for everyone to be able to run the game.

So we have this 2GHz processor here, and then we have four CPU's that are twice as fast, so we have eight times more [processing power] over here. It's hard to scale all the way [across that]. It does scale, so a dual core runs faster than a single core, and a quad-core runs even faster.

What we do thread is the entire physics update on a separate thread, we thread our shadow volume updates, [...] particle updates and tree updates. And then the obvious things like everybody is probably doing, like sound, voice over IP and things like that. But the four I mentioned are the [threaded processes we render] frame to frame.

Well I will never break you of your paranoia but look at this last line... the thread their updates per physics, shadow volumes (fog), particles, and tree .... hmmmm, after the update... that get's sent to the GPU. Frame by frame, each thread works in parallel to complete a update to the frame which is then sent to the GPU ....

Thanks for the link...

Jack

[Boschwanza]

The additional 1920x Settings would be nice, thanks a lot i appreciate your work.

So far your scores are pretty close to what overclubersclub discovered , they are just getting earlier into gpu bound situation with the 8800GT then you.

[JumpingJack]

The additional 1920x Settings would be nice, thanks a lot i appreciate your work.

So far your scores are pretty close to what overclubersclub discovered , they are just getting earlier into gpu bound situation with the 8800GT then you.

I think you can generalize this to a yes. The most curious one is the Company of Hero's run -- I do not have opposing forces, if I can I will stop by the store and pick up a copy.

I will repeat what I do on the 4870 X2, UPS tracker is showing an afternoon delivery

jack

[Loser777]

gosh, you posts are just a bunch of excuses for why AMD fails to match or beat Intel.

[gosh]

gosh, you posts are just a bunch of excuses for why AMD fails to match or beat Intel.

Well I can tell you that even I was suprised when I saw results from World in Conflict that Jack did. I was starting to wounder if he had forgotten to disable the TLB bug. The difference was so big that it can’t even be explained by differences between how the processors work. You need to optimize for one processor to get that big difference so I looked it up. And found it, Intel has been helping them…

I know it’s hard for a non programmer to understand. But if one game scales to large threads and don’t share memory and don’t syncronize threads then it works well on intel. It’s like running separate single threaded applications that has some point where they joins work and then they go back to work again. This is also easier to do for programmers but it isn’t effective if you really want to use all the power that the processor has. I did show you a link bout the render split design before.

I wouldn’t be surprised if the did use intel’s compiler on the World in Conflict also.

http://yro.slashdot.org/comments.pl?...3&cid=13042922
http://aceshardware.freeforums.org/c...iler-t428.html

[JumpingJack]

Well I can tell you that even I was suprised when I saw results from World in Conflict that Jack did. I was starting to wounder if he had forgotten to disable the TLB bug.

I know it’s hard for a non programmer to unde

http://yro.slashdot.org/comments.pl?...3&cid=13042922
http://aceshardware.freeforums.org/c...iler-t428.html

... I am using B3 silicon, it does not have the TLB bug and turning the patch on or off via BIOS or AOD makes no difference.

Here is a comprehensive effect of the TLB patch on /off on a 2.3 GHz 9600 BE http://forum.xcpus.com/motherboard-c...om-menace.html

I haven't generated the charts yet, but for the all low res/low detail (CPU bound condition) at 1280x1024

pre Patch: Max = 139 Ave = 61 Min = 30
post Patch: Max = 111 Ave = 51 Min = 28

Unlike some of the silly runs here just to answer questions, the data in the article above is usually run 3 times for reproducibility, averaged, and reported as an average, the data I quote above is from the first run.

I can post screen shots if you like.

Computationally, the TLB patch is a huge hit -- especially on games, but considering that games page most of their memory and are so branchy, it is not surprising. However, I have also never observed a TLB bug manifest itself, it was an unfortunate PR disaster for AMD, but running a B2 (TLB bug) processor is perfectly stable over the 20 or so games I tested on it. I would have no reservations recommending someone take a B2 CPU if they so desired, the TLB bug was blown way way out of proportion.


Jack

[RunawayPrisoner]

Again with the programmer excuse. Seriously, pal, no disrespect intended, but knowing how to code doesn't mean anything at all. Just because you are a programmer does not mean you know more than others. And you are the one who suddenly come out with Phenom winning against even Q9450 while hundreds of others find Phenom not even measuring up against Q6600 in most cases.

If you are so knowledgeable, would you mind explaining why Intel processors do so much better than AMD processors with PCSX2? (In the same manner you have been "explaining" things, that is)

[highoctane]

Oh geez the coding again, anyways if it where a coding issue whats the point in buying a phenom rig for gaming considering everything is better optimized to run on Intel anyways if thats the case.

Hmmmm, so basically what we have is build Intel and get better performance in a desktop environment say 95% of the time due to universally optimized coding or buy Phenom for better bandwidth but is slower 95% of the time at the same task due to lack of optimized code.

[gosh]

Oh geez the coding again, anyways if it where a coding issue whats the point in buying a phenom rig for gaming considering everything is better optimized to run on Intel anyways if thats the case.

Because AMD will run all that Intel runs without problems too, and if you add applications doing normal work AMD will like it more.
You can also be certain of that Intel is doing what they can to change how developers create applications. Nehalem/Larrabee etc is thread monsters (more than phenom), you need to scale well in order to use all the power in those processors. If ray tracing will arrive then you probably also going to need one processor that is good at scaling to threads.

The memcpy Intel seems to do with their compiler is really ugly, it is one of the most common functions used in DirectX games (maybe game developers know this if they use intel compiler and write their own function)

[RunawayPrisoner]

Does not matter. Regardless, games were made for Intel processor, and they run better on Intel as they were written for Intel (you are implying this) so end of story.

You just literally admitted to Intel being better at handling game, sir.

[gosh]

You just literally admitted to Intel being better at handling game, sir.

What type of game?

[RunawayPrisoner]

Every type of game. Unless you are telling me you know precisely how each game works even without looking at their source code?

[gosh]

Every type of game. Unless you are telling me you know precisely how each game works even without looking at their source code?

Well it isn’t.
There are some links in this thread showing that AMD performs better on high resolutions.

Single threaded games you may be right, core 2 is designed for that