I'm asking this question again, would a dual core E7500 bottleneck a 5870?

[toyota]

Settings: Demo(Ranch Small), 1600x1200 (75Hz), D3D10, Fixed Time Step(No),
Disable Artificial Intelligence(No), Full Screen, Anti-Aliasing(2x),
VSync(No), Overall Quality(Very High), Vegetation(Very High), Shading(Very High),
Terrain(Very High), Geometry(Very High), Post FX(High), Texture(Very High),
Shadow(Very High), Ambient(High), Hdr(Yes),
Bloom(Yes), Fire(Very High), Physics(Very High), RealTrees(Very High)

cpu @ 3.6 with 4 cores

Average Results
Average Framerate: 64.49
Max. Framerate: 93.12
Min. Framerate: 49.32

With cpu @ 2.0 and 4 cores

Average Results
Average Framerate: 61.57
Max. Framerate: 80.55
Min. Framerate: 47.83


cpu with 2 cores disabled @ 2.0

Average Results
Average Framerate: 41.59
Max. Framerate: 62.41
Min. Framerate: 29.51


Does far cry 2 like the quad?

it loves the quad BUT if you look at most benchmarks a highly overclocked dual core should match a slower quad at least with a moderate card.

also you need to remove those benchmarks because your 2.0 dual core results are too much like mine at 2.0 and apoppin already thinks mine are BS. lol

 

[RussianSensation]

This thread is too funny. At the end of the day, I presume everything is subjective:

1) One of my buddies thinks an N450 atom netbook is fast enough for his work (but he doesn't watch YouTube HD)
2) My other friend uses an E6400 @ 2.66ghz and says when he uses Virtualization, his CPU is pegged at 100% and he needs a quad...poor sap.
3) And then there is Aigo who feels that nothing less than a Core i7 980X @ 4.4ghz will do

But we all enjoy our toys. Happy medium, thanks for those FC2 benches. 29.51 fps mins vs. 47.83 fps mins. It's interesting if 5870 was losing this badly in any popular modern game to GTX480, all hell would break loose! But when it comes to CPU speed...mins dont' matter according to this thread because a gamer with 48 frames will enable 16AA on his GTX480 to match 29 frames minimum. ()

 

[apoppin]

Sounds like a plan!

You should include minimums as well as average frames:

Resident Evil 5 (cpu limited)
World in Conflict (both GPU and cpu limited)
GTA4 (cpu limited)
ArMA2 (cpu limited and gpu limited)
Warhammer 40000: Dawn of War 2 (cpu limited)
BF:BC2 (gpu limited)
Dirt2 (gpu limited)
Metro 2033 (gpu limited)
STALKER: Call of Pripyat (gpu limited)
Crysis (more gpu limited)

I am going to expect 48-50 frames minimum in Dirt 2 then on that X2 550 because I am not interested in 70 frames average on that DX11 5870 CF system with 8AA edge-detect if it dives to 30 frames minimum. That's not playable.
http://www.xbitlabs.com/articles/video/display/radeon-hd5830_11.html#sect1

Now you see why I am forced to run only 2AA on my 4890 in that racing game?

Look at Wings of Prey in the same link @ 1920x1200 8AA
4890 = 77 avg
5770 = 64 avg
^^ both appear playable, until you see 51 vs. 38 mins ()

i already did a massive comparison with GTX 280, 4870, 4870-X2 and 4870-X3 TriFire
- this is a rinse and repeat with a few less CPUs but with more powerful GPUs

i generally include minimums and maximums; not all games though and MOST of your games are already in my benchmark suite

 

[RussianSensation]

i generally include minimums and maximums; not all games though and MOST of your games are already in my benchmark suite

I am really anxious to see GTX480 vs. 5870 mins.

It seems from what I've read, with higher AA, GTX480 is able to sustain much better minimum framerates than a 5870. Would be interesting to see if this is true. :hmm:

 

[happy medium]

cpu with 2 cores disabled @ 2.0

Loop 1
Total Frames: 1949, Total Time: 51.01s
Average Framerate: 38.21
Max. Framerate: 62.20 (Frame:274, 5.90s)
Min. Framerate: 22.98 (Frame:1314, 34.02s)

Loop 2
Total Frames: 2175, Total Time: 51.02s
Average Framerate: 42.63
Max. Framerate: 66.11 (Frame:365, 6.81s)
Min. Framerate: 28.94 (Frame:1760, 41.27s)

Loop 3
Total Frames: 2240, Total Time: 51.01s
Average Framerate: 43.91
Max. Framerate: 64.06 (Frame:343, 6.50s)
Min. Framerate: 31.79 (Frame:1542, 34.59s)

Average Results
Average Framerate: 41.59
Max. Framerate: 62.41
Min. Framerate: 29.51

cpu with 2 cores disabled @ 3.6

Loop 1
Total Frames: 2804, Total Time: 51.00s
Average Framerate: 54.98
Max. Framerate: 72.75 (Frame:4, 0.07s)
Min. Framerate: 41.56 (Frame:1876, 34.41s)

Loop 2
Total Frames: 2826, Total Time: 51.01s
Average Framerate: 55.40
Max. Framerate: 72.13 (Frame:176, 2.69s)
Min. Framerate: 43.74 (Frame:1882, 34.10s)

Loop 3
Total Frames: 2811, Total Time: 51.00s
Average Framerate: 55.12
Max. Framerate: 71.09 (Frame:177, 2.77s)
Min. Framerate: 44.02 (Frame:1874, 34.18s)

Average Results
Average Framerate: 55.16
Max. Framerate: 71.39
Min. Framerate: 43.34

In this game looks like Toyota is right.:hmm:

 

[toyota]

cpu with 2 cores disabled @ 2.0

Average Results
Average Framerate: 41.59
Max. Framerate: 62.41
Min. Framerate: 29.51


cpu with 2 cores disabled @ 3.6

Average Results
Average Framerate: 55.16
Max. Framerate: 71.39
Min. Framerate: 43.34

In this game looks like Toyota is right.:hmm:

yep. I have tested the crap out of that game so I am very familiar with it at different cpu speeds. actually I have tons of benchmarks for Far Cry 2 and other games but I learned a while back that all I get is excuses from the same couple of people when I post them so I usually dont bother.

 

[apoppin]

I am really anxious to see GTX480 vs. 5870 mins.

It seems from what I've read, with higher AA, GTX480 is able to sustain much better minimum framerates than a 5870. Would be interesting to see if this is true. :hmm:

Well i have already posted minimums with GTX 480 vs HD 5870 in 14 games with i7 at 3.80 GHz. Yes it does.

What i am particularly interested in (next) is the respective performance hit with 8xMSAA; last gen the HD 4870 was better than the GTX 280; i think things may have changed this gen but i haven't done enough testing.

cpu with 2 cores disabled @ 2.0

What is this crap about disabling cores? it is not the same as using a dual core to compare to quad; you can have cache issues the way you are going about it.

 

[happy medium]

Well i have already posted minimums with GTX 480 vs HD 5870 in 14 games with i7 at 3.80 GHz. Yes it does.

What i am particularly interested in (next) is the respective performance hit with 8xMSAA; last gen the HD 4870 was better than the GTX 280; i think things may have changed this gen but i haven't done enough testing.


What is this crap about disabling cores? it is not the same as using a dual core to compare to quad; you can have cache issues the way you are going about it.

I would hope I had the same cache issues with 2 cores disabled at 3.6?

Appopin, When is this giant gtx 480/5870 overclocked acticle getting published?

 

[toyota]

I would hope I had the same cache issues with 2 cores disabled at 3.6?

Appopin, When is this giant gtx 480/5870 overclocked acticle getting published?

yeah I dont think he realized your were just showing a speed comparison. anyway thanks for showing that because it backs up my claims/benchmarks that a 2.0 Core 2 duo is a large bottleneck for some games. I believe they claimed like 5% difference or something silly for Far Cry 2 on their site with a Core 2 Duo at 2.0. lol

 

[apoppin]

I would hope I had the same cache issues with 2 cores disabled at 3.6?

Appopin, When is this giant gtx 480/5870 overclocked acticle getting published?

Next week; but this one is only GTX 480(825/1100) vs HD 5870 (975/1300)

Later on, i will compare performance of Core i7 to Dragon platform .. that is next month
- there are several reviews in-between

actually i better get busy and finish my ECS 890X CF mobo review .. it is late
- i write it while i am benching GTX 480 and working on that video comparison of noise/heat levels .. those are up tomorrow sometime for Tuesday

 

[BFG10K]

I have posted benchmarks plenty of time but all you do is come up with excuses. I showed a 40-50% difference in Far Cry 2 at 1920 with 2x AA and you told me I needed more AA.

http://www.firingsquad.com/hardware/intel_core_i7-980x_extreme_edition/page5.asp

At 1600x1200 with 4xAA, there’s absolutely no performance difference between a six-core i7 980X and a dual-core E8600.

http://www.overclockersclub.com/reviews/intel__core_i7_980x/8.htm

At 1920x1200 with 4xAA, the Athlon X2 is just 4 FPS slower than Intel’s hex-core monster.

http://www.neoseeker.com/Articles/Hardware/Reviews/intel_core_i7_980x/11.html

At 1920x1200, the E8400 is actually 1 FPS faster than the i7 980X.

http://www.guru3d.com/article/core-i7-980x-review/18

At 1600x1200 the i660 is the same speed as the i7 980X; at 1920x1080 (the resolution you used in the game) the dual-core i660 outruns the fastest processor available in consumer space today.

http://ixbtlabs.com/articles3/video/quadcore-p6.html

Having more than two cores makes basically zero difference here.

That’s five professional reviews that agree with my findings about Far Cry 2. It’s likely the problem here is your choice of benchmark (Ranch Small), which spends large portions looking at the sky, and spends large portions gliding through the air in an unrealistic fashion.

This isn’t rocket science. Even without benchmarks, anyone that has finished the game can load a few save-points at random and check the framerate in various parts of the game. I’ve actually done this at 2 GHz, 3 GHz, and on an i5 750, and concluded the processor makes little to no difference at the settings I play the game at (1920x1200, 16xAF, 2xTrMS). But dropping the resolution/AA makes the performance shoot up proportionally, indicating the GPU is bottlenecking things. Have you run such tests?

all I was doing was putting it on realistic settings for a gtx260 in that game and the other ones I tested.

So in other words you admit a faster GPU would allow higher settings in your games, ergo admitting you’re GPU bound?

for example in Red Faction Guerrilla the min framerate was cut in half by running my cpu at 2.0 while using all high settings at 1920. in Ghostbusters, the game was almost unplayable at times and the min framerate was basically cut in half in that game too while my cpu was 2.0 at 1920 highest settings.

Please link to professional benchmarks to back your claims. Also for you 1920 is actually 1080p, so you might want to make that clearer, otherwise you'll give people the wrong idea you're running a higher resolution.

Also a minimum means squat unless you have a plot putting it into context. Unless you can demonstrate a lower framerate for a meaningful period of time, a single data point means nothing. We’ve been over this repeatedly before.

Even Ryan Smith (AT’s GPU reviewer) was saying he can’t get minimums to show reliable results. I suppose you expect us to believe that you can?

an i7 gets nearly twice the frame rate in Bad Company as a decent dual core at 1680 so pretty sure even at a more realistic 1920 the gap would still be pretty wide. not saying that the dual core isnt playbale but the gap is still pretty large in that game.

Again, show us third party benchmarks to back your claims. I already linked two reviews last time that show BC is massively GPU limited. CPU load was demonstrated (it didn’t even max out two cores), and it also demonstrated a 5970 getting a substantial performance gain from a 5870 at 1280x1024. That’s not CPU limited.

Come on, show us benchmarks of the five games you keep repeating.

What’s more, you don’t even know if the OP plays any of those games in any of the threads you post in. Yet that doesn’t stop you from telling them to buy slower GPUs.

You’ve got three games so far (Arma 2, GTA 4 and RE5) which I’ll concede are CPU limited to a meaningful degree. So for those three games you’ll tell people to buy slower GPUs, when it can be demonstrated that 300 (if not 3000) other games will show vastly more benefit from a faster GPU?

 

[3DVagabond]

Anyone tried shutting down unneeded processes to see how much of a difference that makes? I imagine it would matter more an a dual core. I'm curious to see if a dual core might perform appreciably better if wasteful processes were shut down?

 

[Dark4ng3l]

Anyone tried shutting down unneeded processes to see how much of a difference that makes? I imagine it would matter more an a dual core. I'm curious to see if a dual core might perform appreciably better if wasteful processes were shut down?

Well it depends what you mean by unneeded processes. Most of these benchmarks are done on fresh installs so they probably don't have much superfluous crap running in the background.

 

[AzN]

60 frames min in Unreal Tournament 3, and 50 frames min in Dirt2. Every game is different because 45 frames may be smooth in one game but going from 60 to 38 frames in a racing game pisses me off because I can feel jerkiness! :awe:

Like I said you have high expectations.

I'm fine with 30fps min for racing games and above 30fps for most FPS games.

 

[toyota]

http://www.firingsquad.com/hardware/intel_core_i7-980x_extreme_edition/page5.asp

At 1600x1200 with 4xAA, there’s absolutely no performance difference between a six-core i7 980X and a dual-core E8600.

http://www.overclockersclub.com/reviews/intel__core_i7_980x/8.htm

At 1920x1200 with 4xAA, the Athlon X2 is just 4 FPS slower than Intel’s hex-core monster.

http://www.neoseeker.com/Articles/Hardware/Reviews/intel_core_i7_980x/11.html

At 1920x1200, the E8400 is actually 1 FPS faster than the i7 980X.

http://www.guru3d.com/article/core-i7-980x-review/18

At 1600x1200 the i660 is the same speed as the i7 980X; at 1920x1080 (the resolution you used in the game) the dual-core i660 outruns the fastest processor available in consumer space today.

http://ixbtlabs.com/articles3/video/quadcore-p6.html

Having more than two cores makes basically zero difference here.

That’s five professional reviews that agree with my findings about Far Cry 2. It’s likely the problem here is your choice of benchmark (Ranch Small), which spends large portions looking at the sky, and spends large portions gliding through the air in an unrealistic fashion.

This isn’t rocket science. Even without benchmarks, anyone that has finished the game can load a few save-points at random and check the framerate in various parts of the game. I’ve actually done this at 2 GHz, 3 GHz, and on an i5 750, and concluded the processor makes little to no difference at the settings I play the game at (1920x1200, 16xAF, 2xTrMS). But dropping the resolution/AA makes the performance shoot up proportionally, indicating the GPU is bottlenecking things. Have you run such tests?


So in other words you admit a faster GPU would allow higher settings in your games, ergo admitting you’re GPU bound?


Please link to professional benchmarks to back your claims. Also for you 1920 is actually 1080p so you might want to make that clearer, otherwise you'll give people the idea you're running a higher resolution.
wrong
Also a minimum means squat unless you have a plot putting into context. Unless you can demonstrate a lower framerate for a meaningful period of time, a single data points means nothing. We’ve been over this repeatedly before.

Even Ryan Smith (AT’s GPU reviewer) was saying he can’t get minimums to show reliable results. I suppose you expect us to believe that you can?


Again, show us third party benchmarks to back your claims. I already linked two reviews last time that show BC is massively GPU limited. CPU load was demonstrated (it didn’t even max out two cores), and it also demonstrated a 5970 getting a substantial performance gain from a 5870 at 1280x1024. That’s not CPU limited.

Come on, show us benchmarks of the five games you keep repeating.

What’s more, you don’t even know if the OP plays any of those games in any of the threads you post in. Yet that doesn’t stop you from telling them to buy slower GPUs.

You’ve got three games so far (Arma 2, GTA 4 and RE5) which I’ll concede are CPU limited to a meaningful degree. So for those three games you’ll tell people to buy slower GPUs, when it can be demonstrated that 300 (if not 3000) other games will show vastly more benefit from a faster GPU?

you might want to reread some of your links there buddy. you are calling many cpus by the wrong name such as an Athlon X2 when there is not even one there. also in those tests all those cpus are pretty equal anyway so with 4x AA on they probably would be close. the slowest thing on Firingsquads test is an 8600 which is about 60% faster than a Core 2 at 2.0 which you claimed there was no bottleneck. and yes I have fired up the game with fraps on and the difference is most certainly there with my cpu lowered to 2.0. hell I even took a screenshot in another thread showing the fps difference IN GAME.

as for BC 2 they already showed the faster cpus getting nearly TWICE the framerate at 1680 as the mid range dual core cpus so dont give me BS on that. a cpu doesnt have to be fully utilized at 100% to mean it would be faster on stronger cpu. there are plenty of games that utilzes faster cpus or quads just fine even though they may only be using 60-70% of a slower cpu.

yes genius I was slightly gpu bottlenecked in some of my benchmarks. that was the POINT to show you that even though I could put the settings on the edge of being gpu limited upping the speed on the cpu still helped because 2.0 is too slow. your crazy logic is to keep adding AA until the gpu becomes the bottleneck no matter what.

 

[toyota]

Like I said you have high expectations.

I'm fine with 30fps min for racing games and above 30fps for most FPS games.

yeah about 35fps and above is fine in most cases but I do hate wildly fluctuating framerates. I would rather have a steady 35-45fps then to have 30fps one second an 60 the next. also I do end up turning vsync on in most of my games so the higher I can average the better. I usually stay above 30 fps in BC 2 with my cpu at 3.6 and running high settings with 2x AA at 1920. HBAO takes a pretty big hit but I have played some with it and its not terrible in single player.

 

[BFG10K]

you might want to reread some of your links there buddy. you are calling many cpus by the wrong name such as an Athlon X2 when there is not even one there.

What on Earth are you babbling about?

http://www.overclockersclub.com/reviews/intel__core_i7_980x/8.htm

See the dark red column? The label clearly says “Athlon II X2 240e”. Would you like me to post a screenshot and circle it for you?

That processor is far slower than an E6850 @ 3 GHz (20-30%, if not more), and yet it’s only 4 FPS slower than the six-core i7 980X. Even better, that particular review uses a GTX260, which is the same as your GPU. Clearly you’re doing something you’re not telling us (are you even using HQ 16xAF for instance?)

What an absolute waste of money sinking money into the processor for Far Cry 2 when it’s obvious the GPU is the primary bottleneck at any reasonable resolution/AA/detail level.

also in those tests all those cpus are pretty equal anyway so with 4x AA on they probably would be close. the slowest thing on Firingsquads test is an 8600

What are you talking about? Are you even reading the benchmarks I linked to?

which is about 60% faster than a Core 2 at 2.0 which you claimed there was no bottleneck.

I said the GPU is the biggest bottleneck by far. I also said overall for 17 games, there was little to no difference from the processor on average.

For the specific case of Far Cry 2, a 50% CPU underclock went from 40.14 FPS to 38.25 FPS, which is a 4.71% performance loss. Meanwhile the GPU dropped down to 31 FPS from the same 50% underclock, which is a 22.77% performance loss.

So, the GPU affected performance by a factor of almost five. For Far Cry 2, I’d rather have an E6580 underclocked to 2 GHz paired with a GTX480, while you can have an i7 980X with six cores running at 4 GHz if you like, along with your GTX260.

It doesn’t take a genius to figure out that I’ll get far better performance and image quality than you will.

and yes I have fired up the game with fraps on and the difference is most certainly there with my cpu lowered to 2.0. hell I even took a screenshot in another thread showing the fps difference IN GAME.

Where? Show us the screenshots in Far Cry 2. Run the game at my settings and show us. I have five professional reviews demonstrating CPU flat-lining when the game is run at similar detail levels to mine, hence backing my claims.

as for BC 2 they already showed the faster cpus getting nearly TWICE the framerate at 1680 as the mid range dual core cpus so dont give me BS on that. a cpu doesnt have to be fully utilized at 100% to mean it would be faster on stronger cpu. there are plenty of games that utilzes faster cpus or quads just fine even though they may only be using 60-70% of a slower cpu.

You keep repeating statements but fail to provide anything to back them up. I’ll link this again to you for the last time:

http://www.techspot.com/article/255-battlefield-bad-company2-performance/page7.html

Even at 1680x1050 and 1920x1200, there’s no major performance difference between 2.22 GHz and 2.96 GHz. At 2560x1600, there’s absolutely no performance difference between 3.7 GHz and 2.22 GHz.

And this is with 5870 Crossfire, with just 2xAA. Use 4xAA on a single 5870 and you’ll be hitting that flat-line much sooner. On your GTX260 you’ll probably be flat-lining at 1680x1050 with 4xAA, or maybe even at 1280x1024.

http://www.xbitlabs.com/articles/video/display/battlefield-bad-company-2_5.html#sect1

Even at just 1280x1024 with 4xAA, a 5970 improves the average framerate by ~45% and the minimum framerate by ~55% over a single 5870. I’m not sure how anyone can claim the game is CPU limited based on that result.

yes genius I was slightly gpu bottlenecked in some of my benchmarks. that was the POINT to show you that even though I could put the settings on the edge of being gpu limited upping the speed on the cpu still helped because 2.0 is too slow.

Actually you haven’t really shown anything. The FC2 results you posted are not in line with the five other reviews I posted that back my claims, so you’ll forgive me if I don’t take your results seriously. Again, post up third party benchmarks for the other games you keep repeating and I’ll look at them.

your crazy logic is to keep adding AA until the gpu becomes the bottleneck no matter what.

My “crazy logic” is to increase the detail levels to ensure the highest possible IQ while being playable. That’s not “crazy logic”, that’s taking the best possible advantage of a product I paid money for (the GPU). A natural consequence of this is that the CPU becomes far less relevant that the GPU.

I also notice you’ve yet again dodged the issue of showing benchmarks for the rest of the games you keep parroting every time a thread like this comes along. I've conceded Arma 2, GTA4 and RE5, so let's see you prove the rest. If the sole basis of your argument rests on those three games, then you don't really have a leg to stand on.

 

[toyota]

What on Earth are you babbling about?

http://www.overclockersclub.com/reviews/intel__core_i7_980x/8.htm

See the dark red column? The label clearly says “Athlon II X2 240e”. Would you like me to post a screenshot and circle it for you?

That processor is far slower than an E6850 @ 3 GHz (20-30%, if not more), and yet it’s only 4 FPS slower than the six-core i7 980X. Even better, that particular review uses a GTX260, which is the same as your GPU. Clearly you’re doing something you’re not telling us (are you even using HQ 16xAF for instance?)

What an absolute waste of money sinking money into the processor for Far Cry 2 when it’s obvious the GPU is the primary bottleneck at any reasonable resolution/AA/detail level.


What are you talking about? Are you even reading the benchmarks I linked to?


I said the GPU is the biggest bottleneck by far. I also said overall for 17 games, there was little to no difference from the processor on average.

For the specific case of Far Cry 2, a 50% CPU underclock went from 40.14 FPS to 38.25 FPS, which is a 4.71% performance loss. Meanwhile the GPU dropped down to 31 FPS from the same 50% underclock, which is a 22.77% performance loss.

So, the GPU affected performance by a factor of almost five. For Far Cry 2, I’d rather have an E6580 underclocked to 2 GHz paired with a GTX480, while you can have an i7 980X with six cores running at 4 GHz if you like, along with your GTX260.

It doesn’t take a genius to figure out that I’ll get far better performance and image quality than you will.


Where? Show us the screenshots in Far Cry 2. Run the game at my settings and show us. I have five professional reviews demonstrating CPU flat-lining when the game is run at similar detail levels to mine, hence backing my claims.


You keep repeating statements but fail to provide anything to back them up. I’ll link this again to you for the last time:

http://www.techspot.com/article/255-battlefield-bad-company2-performance/page7.html

Even at 1680x1050 and 1920x1200, there’s no major performance difference between 2.22 GHz and 2.96 GHz. At 2560x1600, there’s absolutely no performance difference between 3.7 GHz and 2.22 GHz.

And this is with 5870 Crossfire, with just 2xAA. Use 4xAA on a single 5870 and you’ll be hitting that flat-line much sooner. On your GTX260 you’ll probably be flat-lining at 1680x1050 with 4xAA, or maybe even at 1280x1024.

http://www.xbitlabs.com/articles/video/display/battlefield-bad-company-2_5.html#sect1

Even at just 1280x1024 with 4xAA, a 5970 improves the average framerate by ~45% and the minimum framerate by ~55% over a single 5870. I’m not sure how anyone can claim the game is CPU limited based on that result.


Actually you haven’t really shown anything. The FC2 results you posted are not in line with the five other reviews I posted that back my claims, so you’ll forgive me if I don’t take your results seriously. Again, post up third party benchmarks for the other games you keep repeating and I’ll look at them.


My “crazy logic” is to increase the detail levels to ensure the highest possible IQ while being playable. That’s not “crazy logic”, that’s taking the best possible advantage of a product I paid money for (the GPU). A natural consequence of this is that the CPU becomes far less relevant that the GPU.

I also notice you’ve yet again dodged the issue of showing benchmarks for the rest of the games you keep parroting every time a thread like this comes along. I've conceded Arma 2, GTA4 and RE5, so let's see you prove the rest. If the sole basis of your argument rests on those three games, then you don't really have a leg to stand on.

you pick the parts out of reviews that you want to use for you argument and so do I. again look at the legionhardware cpu BC 2 results.

also again your 2.0 Core duo claims are rubbish for Far Cry 2 and especially more cpu intensive games. happy medium and I both showed you that and there was a 40-50% difference and not just the benchmark for me either.

but yeah you are right I will just turn on tons of AA and make it more gpu limited because my cpu is too damn slow to do anything else at 2.0 in Far Cry 2 or more cpu intensive games.


EDIT: okay so I also tested Red Faction Guerrilla and Ghostbusters a while back at this 2.0 you keep saying isnt a bottleneck for even a gtx285 or even better. there NO enemies on the screen, just a route that I repeated which included a little destruction using the same amount of explosives each time in Red Faction Guerrilla and shooting up the room in a little in Ghostbusters and then some running around. this is an average I got after three runs at each cpu speed for each game. I think I would pick my Core 2 duo at 3.8 over playing with it at 2.0. I could hardly even walk when the shit hit the fan at that cpu speed. could I use a faster gpu? sure but moving the gpu speeds around with my cpu at 2.0 had no impact on framerate in Ghosttbusters, Far Cry 2, or Red Faction Guerrilla. that certainly means I am cpu limited with it at 2.0 in these games.


Ghostbusters 1920x1080 highest settings

E8500 @ 2.0

Min, Max, Avg
11, 45, 20.326

E8500 @ 3.8

Min, Max, Avg
21, 71, 33.451


Red Faction Guerrilla 1920x1080 highest settings

E8500 @ 2.0
Min, Max, Avg
6, 38, 25.813

E8500 @ 3.8
Min, Max, Avg
15, 40, 30.227


if that would have been a 5870 in there the difference would have been massive and really showed how slow the Core 2 at 2.0 was. thats right even making the games as gpu limited as possible for the gtx260 and I still had a large difference in min framerate and even playability in those two games during action. I would probably get similar results in Bad Company 2 if I tested at 2.0 but I am too lazy to do that tonight.

 

[BFG10K]

you pick the parts out of reviews that you want to use for you argument and so do I. again look at the legionhardware cpu BC 2 results.

These ones?

http://www.legionhardware.com/articles_pages/battlefield_bad_company_2_tuning_guide,7.html

1680x1050 with no AA or AF? LMAO. Do you think someone with an i7 system is going to be gaming at such settings?

Also they only needed an i3 540 to hit 72 FPS, which is pretty much the bottleneck they were getting on 5870 Crossfire at 2560x1600 with 2xAA (~74 FPS) at the other site. For a single 5870 the settings and/or processor needed will be even lower.

This is an elementary inference to anyone that understands bottlenecking.

also again your 2.0 Core duo claims are rubbish for Far Cry 2 and especially more cpu intensive games. happy medium and I both showed you that and there was a 40-50% difference and not just the benchmark for me either.

Um, his setting of 1600x1200 was even lower than your 1080p setting.

Once again you continue type without providing anything to back your claims. You simply can’t explain away the five independent reviews that I posted that back my claims.

You’ve also failed to acknowledge the Athlon II X2 240e (probably not much faster than an E6580 @ 2 GHz) used in the tests with the same graphics card as you, and it still backs my claims.

You’ve also failed to provide benchmarks outside of the three titles we’ve already agreed are CPU bound.

but yeah you are right I will just turn on tons of AA and make it more gpu limited because my cpu is too damn slow to do anything else at 2.0

You’ve got nothing, so just admit it. Maybe in the future you should post “if you play these three games then get a faster CPU, otherwise get a faster GPU”.

That’ll avoid anyone misunderstanding what you’re trying to say.

EDIT: okay so I also tested Red Faction Guerrilla and Ghostbusters a while back at this 2.0 you keep saying isnt a bottleneck for even a gtx285 or even better.

Please link to third party benchmarks of those two games, thanks.

 

[Tempered81]

Breaking it down to the basics, you should match your gpu to your platform (cpu,ram,mobo)

You're running like the 3rd or 4th fastest out of hundreds of video cards on the 40th fastest CPU. So, not a good match. You want a 3.5ghz Quad, 3.5g i5/i7, 3.5g deneb, or 4.0g 6mb Penryn at the least. Out of these you have about 40 SKU's to choose from.

 

[toyota]

These ones?

http://www.legionhardware.com/articles_pages/battlefield_bad_company_2_tuning_guide,7.html

1680x1050 with no AA or AF? LMAO. Do you think someone with an i7 system is going to be gaming at such settings?

Also they only needed an i3 540 to hit 72 FPS, which is pretty much the bottleneck they were getting on 5870 Crossfire at 2560x1600 with 2xAA (~74 FPS) at the other site. For a single 5870 the settings and/or processor needed will be even lower.

This is an elementary inference to anyone that understands bottlenecking.


Um, his setting of 1600x1200 was even lower than your 1080p setting.

Once again you continue type without providing anything to back your claims. You simply can’t explain away the five independent reviews that I posted that back my claims.

You’ve also failed to acknowledge the Athlon II X2 240e (probably not much faster than an E6580 @ 2 GHz) used in the tests with the same graphics card as you, and it still backs my claims.

You’ve also failed to provide benchmarks outside of the three titles we’ve already agreed are CPU bound.


You’ve got nothing, so just admit it. Maybe in the future you should post “if you play these three games then get a faster CPU, otherwise get a faster GPU”.

That’ll avoid anyone misunderstanding what you’re trying to say.

Please link to third party benchmarks of those two games, thanks.

again you twist crap for your argument. I didnt say ONE damn thing about any cpu on that site being unplayable now did I? I said that the cpu looks like a limitation at those settings and that a high end cpu was getting nearly twice the framerate. and do you magically think that turning on 4x AA is going to close down a 80-90% gap? lol. the OP is at 1600x1200 which is more pixels but again not enough to remotely close that gap.

so he tested Far Cry 2 at 1600x1200 and I tested at 1920x1080? those are fairly close as for as absolute pixels. the point was it was nothing like your laughable claim of 4.7 % now was it? and again with my cpu at 2.0 in Far Cry 2 changing the gpu clocks made ZERO difference because I was 100% cpu bottleneck at that low speed.

those 2 games I just showed you were not even mentioned by you but I guess you still can only count to 3?

 

[toyota]

Breaking it down to the basics, you should match your gpu to your platform (cpu,ram,mobo)

You're running like the 3rd or 4th fastest out of hundreds of video cards on the 40th fastest CPU. So, not a good match. You want a 3.5ghz Quad, 3.5g i5/i7, 3.5g deneb, or 4.0g 6mb Penryn at the least. Out of these you have about 40 SKU's to choose from.

thank you. nothing wrong with just getting the 5850 in his case and overclocking it a bit. that will provide him basically the same results with his cpu at 1600 anyway. makes more sense than saying oh look I cant average more than 50 fps so I will spend 100 more bucks to get the same framerates but maybe up the AA.

 

[BFG10K]

again you twist crap for your argument. I didnt say ONE damn thing about any cpu on that site being unplayable now did I?

I never said you did, so stop trying to deflect the issue.

I said that the cpu looks like a limitation at those settings and that a high end cpu was getting nearly twice the framerate.

Yeah, at 1680x1050 with no AA and no AF. Do you expect someone that drops $1K onto an i7 980X is going to be gaming at such a setting?

and do you magically think that turning on 4x AA is going to close down a 80-90% gap?

Since I’m not convinced you’re even looking at the pictures, I’ll inline the image for you:

See? No performance difference between 2.22 GHz and 3.70 GHz.

And again, that’s on a 5870 Crossfire with just 2xAA; your single GTX260 @ 4xAA is going to hit that bottleneck far quicker, probably at just 1680x1050.

Are you actually looking at the benchmarks and understanding them?

the OP is at 1600x1200 which is more pixels but again not enough to remotely close that gap.

I’m saying his setting is as low as yours. That’s my point.

the point was it was nothing like your laughable claim of 4.7 % now was it?

But this was:

http://www.techspot.com/article/255-battlefield-bad-company2-performance/page7.html

53 FPS vs 49 FPS only shows a ~7.5% performance loss between i7 980X and the Athlon II X2 240e. Clearly such a small performance hit means the GPU is the bottleneck by almost 100%, with the same GTX260 you used in your tests.

This backs the findings of my test, and is in stark contrast to you running around and claiming you were 100% CPU limited and your GPU clocks made no difference to FC2.

 

[AzN]

You keep repeating statements but fail to provide anything to back them up. I’ll link this again to you for the last time:

http://www.techspot.com/article/255-battlefield-bad-company2-performance/page7.html

Even at 1680x1050 and 1920x1200, there’s no major performance difference between 2.22 GHz and 2.96 GHz. At 2560x1600, there’s absolutely no performance difference between 3.7 GHz and 2.22 GHz.

And this is with 5870 Crossfire, with just 2xAA. Use 4xAA on a single 5870 and you’ll be hitting that flat-line much sooner. On your GTX260 you’ll probably be flat-lining at 1680x1050 with 4xAA, or maybe even at 1280x1024.

http://www.xbitlabs.com/articles/video/display/battlefield-bad-company-2_5.html#sect1

Even at just 1280x1024 with 4xAA, a 5970 improves the average framerate by ~45% and the minimum framerate by ~55% over a single 5870. I’m not sure how anyone can claim the game is CPU limited based on that result.

It seems to me you hand picked particular articles in your favor. Techspot article downclocks and i7 to demonstrate CPU bottleneck which doesn't relate to what Toyota was talking about. Xbit article doesn't tell much about CPU except that it uses an i7 with many GPU's

http://www.legionhardware.com/articles_pages/battlefield_bad_company_2_tuning_guide,3.html

Here we have GTX 260 with dual core and quad core demonstrating what Toyota was proposing. I don't know about 100% better performance but it does give substantial performance boost going from dual core to quad core. I do however agree that BC2 is just as GPU intensive.

 

[AzN]

Since I’m not convinced you’re even looking at the pictures, I’ll inline the image for you:

See? No performance difference between 2.22 GHz and 3.70 GHz.

And again, that’s on a 5870 Crossfire with just 2xAA; your single GTX260 @ 4xAA is going to hit that bottleneck far quicker, probably at just 1680x1050.

Are you actually looking at the benchmarks and understanding them?

It could just be architecture. BC2 might not respond well with clock speeds at certain threshold and that's an awful high resolution.