How does the 290X stack against the 970 at 1080p as of now?

[LokiFire]

You really don't want the 290X to be faster, do you? I showed you actual, recent benchmarks, you called BS. I told you that I tested both cards and you called BS. I'm too lazy and don't care enough to look for those results I had.

And what does the CPU have to do with it? Both cards tested with it so the results are as similar as possible. Also, if CPU was a factor here, it would be in favor of the GTX 970 because Radeon GPUs are known to work better with lower CPU overhead.

No i am just not biased.

[Morgan MLGman]

No i am just not biased.

Everything proves otherwise.

1. You denying logical arguments while throwing some unsupported claims or outdated benchmarks

2. I couldn't help myself but go into your profile, this is the first thing that pops up: "I am just not biased."

@don_svetlio you're gonna have a good laugh over here mate ^^

[LokiFire]

Everything proves otherwise.

1. You denying logical arguments while throwing some unsupported claims or outdated benchmarks

2. I couldn't help myself but go into your profile, this is the first thing that pops up: "I am just not biased."

@don_svetlio you're gonna have a good laugh over here mate ^^

It's a joke about my PC you dumbass LOL.

[Morgan MLGman]

It's a joke about my PC you dumbass LOL.

Then why exactly did you call me "biased"? I told you that I've had both cards. In fact this is my first-ever AMD GPU cause I've been using Nvidia since like 2002.

And my CPU was on such a good deal I couldn't pass it up.

[Kinda Bottlenecked]

About 10-5% for the better models. Really depends on the model of 290x and what its boost clock is. 

[ivan134]

Just so you know a 970 OC's too and last time i checked OC far further than AMD GPU's.

A 390x is an overclocked 290x so I'm not sure what you're talking about. Your average 970 OC is between 1450 and 1500 and at 1500 it doesn't even beat a 390/290 OC'ed to 1150, so what makes you think it can compete with a 290x OC'ed to 1150?

[Morgan MLGman]

A 390x is an overclocked 290x so I'm not sure what you're talking about. Your average 970 OC is between 1450 and 1500 and at 1500 it doesn't even beat a 390/290 OC'ed to 1150, so what makes you think it can compete with a 290x OC'ed to 1150?

If you read all his posts in this thread, you won't find too much reasonable thinking or logic so I guess your question is a rhetorical one

[xAcid9]

Just so you know a 970 OC's too and last time i checked OC far further than AMD GPU's.

1mhz in Maxwell is not equal to 1mhz in Hawaii, i thought your already know this since your 950mhz GTX580 obliterate 1300mhz GTX960.

[Kinda Bottlenecked]

Just so you know a 970 OC's too and last time i checked OC far further than AMD GPU's.

 

 

1550mhz 970

1140mhz 390 

 

Same performance. 

[don_svetlio]

290X is faster than a 970 as proven by @Morgan MLGman who used to own BOTH

[sushisharkjl]

They were not nerfed, NVIDIA just does not optimize for Kepler anymore just like they stopped with Fermi, i wish people would get their head out of their asses.

 

If discontinued hardware support is not a nerf, then what is?

[Monarch]

I added a newer graph, it's 2% now. It's faster, no doubt about it and it's getting faster cause of AMD improving their drivers. And read what I also edited, it's a reference one they're testing. I can tell you from my own, first-hand testing with both cards that the 290X is faster. In fact my 290X after good overclocking surpassed a reference 980 by 4%. (in userbenchmark)

 

You can't generalize like that. It depends on what they're rendering. Firestrike 1.1 is a good example. It has 2 graphics test, 1 has heavy tessellation where 970 has advantage, and the other features smoke simulation that uses compute shaders in which case 290X performs a bit better. Overall, they're about the same, depending on the clocks.

 

Since a lot of games have tessellation and are draw call heavy, I actually found 970 to be faster in those cases. Driver overhead has a significant impact in draw call heavy games so keep in mind that it's not just hardware you should be taking into account, but also software.

[Morgan MLGman]

You can't generalize like that. It depends on what they're rendering. Firestrike 1.1 is a good example. It has 2 graphics test, 1 has heavy tessellation where 970 has advantage, and the other features smoke simulation that uses compute shaders in which case 290X performs a bit better. Overall, they're about the same, depending on the clocks.

 

Since a lot of games have tessellation and are draw call heavy, I actually found 970 to be faster in those cases. Driver overhead has a significant impact in draw call heavy games so keep in mind that it's not just hardware you should be taking into account, but also software.

That is why techpowerup.com tested over 25 games and found the 290X faster on average. I didn't have as much time to test this many games but I also can confirm that it's faster

[Monarch]

That is why techpowerup.com tested over 25 games and found the 290X faster on average. I didn't have as much time to test this many games but I also can confirm that it's faster

 

Games are not equally demanding on hardware all the time, on all levels/maps, regardless of what's happening in the game. No. GTA V for example runs worse on 290X when there's a lot of objects on screen, which happens in some parts of the city. So if the benchmarking path you use to compare graphics cards doesn't include these areas, then yes, 290X might have an advantage.

It's like if you used graphics test 1 from Firestrike 1.1 to compare graphics cards. Since it has heavy tessellation, Maxwell GPUs would have an advantage. 

 

Also, average fps doesn't tell the whole story. One area on the map or one level could be CPU-bound and the other GPU-bound. 290X might achieve higher framerate in a GPU-bound scenario (if there's no tessellation), but 970 will have an advantage in a CPU-bound scenario due to Nvidia's lower overhead. So let's say the 290X achieves 70 fps in the GPU-bound scenario, while the 970 achieves 60 fps. But then in the CPU-bound scenario 970 maintains 60 fps while the 290X drops to 50 fps. Their average would be the same, yet 970 didn't drop below 60.

This is why min fps is important. But the way tech sites measure them is wrong. The way they do it is they take a certain amount of frames, look at their times, calculate framerate for each frame, and then they calculate the average frametime and convert it into framerate. That's why I like the way DigitalFoundry do their tests. They show you the frametime in real time, so you can see each frametime spike.

[EdInk]

Games are not equally demanding on hardware all the time, on all levels/maps, regardless of what's happening in the game. No. GTA V for example runs worse on 290X when there's a lot of objects on screen, which happens in some parts of the city. So if the benchmarking path you use to compare graphics cards doesn't include these areas, then yes, 290X might have an advantage.

It's like if you used graphics test 1 from Firestrike 1.1 to compare graphics cards. Since it has heavy tessellation, Maxwell GPUs would have an advantage. 

 

Also, average fps doesn't tell the whole story. One area on the map or one level could be CPU-bound and the other GPU-bound. 290X might achieve higher framerate in a GPU-bound scenario (if there's no tessellation), but 970 will have an advantage in a CPU-bound scenario due to Nvidia's lower overhead. So let's say the 290X achieves 70 fps in the GPU-bound scenario, while the 970 achieves 60 fps. But then in the CPU-bound scenario 970 maintains 60 fps while the 290X drops to 50 fps. Their average would be the same, yet 970 didn't drop below 60.

This is why min fps is important. But the way tech sites measure them is wrong. The way they do it is they take a certain amount of frames, look at their times, calculate framerate for each frame, and then they calculate the average frametime and convert it into framerate. That's why I like the way DigitalFoundry do their tests. They show you the frametime in real time, so you can see each frametime spike.

I guess your proposal here to to play x-amount of titles from start to finish to get the full picture well NEWS FLASH nobody has that kind of time .When higher numbers are maintained you get a higher average, simple. As long as data is collected over a reasonable amount of time.

 

If an average of 25 games show the 290X perform better (marginally or not) its meaning is simple... the 290X is faster. Why is that such a hard pill to swallow?

[LokiFire]

LOL

[Monarch]

I guess your proposal here to to play x-amount of titles from start to finish to get the full picture well NEWS FLASH nobody has that kind of time .When higher numbers are maintained you get a higher average, simple. As long as data is collected over a reasonable amount of time.

 

If an average of 25 games show the 290X perform better (marginally or not) its meaning is simple... the 290X is faster. Why is that such a hard pill to swallow?

 

Because in reality it isn't, despite what average framerate might tell you. I regretted buying the 290X. I still own it, but I use the 970 in my main PC. Average fps is just part of the story. Framerate and frametime consistency are much more important for the gaming experience.

[App4that]

Because in reality it isn't, despite what average framerate might tell you. I regretted buying the 290X. I still own it, but I use the 970 in my main PC. Average fps is just part of the story. Framerate and frametime consistency are much more important for the gaming experience.

No shit!

The only reason average framerate is pushed on the forum is to win arguments, not for playing video games. That means the people coming looking for advice lose. Like I lost when I came here. At some point this forum will be about PC gaming, not debate.

[Zi6iX]

They were not nerfed, NVIDIA just does not optimize for Kepler anymore just like they stopped with Fermi, i wish people would get their head out of their asses.

And that's a good thing? "Sorry that we don't optimize for our one-gen old gpu's" NVIDIA

[EdInk]

Because in reality it isn't, despite what average framerate might tell you. I regretted buying the 290X. I still own it, but I use the 970 in my main PC. Average fps is just part of the story. Framerate and frametime consistency are much more important for the gaming experience.

Frametime, what is it? 

60 Frames per Second (1000millisec)

 

1 Frame is executed in (1000/60) = 16.6666ms (frametime)

 

45 Frames per second = 1000/45 = 22.22222ms (frametime)

 

Lower FPS means more time processing a frame... the 290X was consistent at outputting more frame rates hence the higher AVERAGE (out of 25 titles cmon!).. what am I getting so wrong here?

[Enderman]

Frametime, what is it? 

60 Frames per Second (1000millisec)

 

1 Frame is executed in (1000/60) = 16.6666ms (frametime)

 

45 Frames per second = 1000/45 = 22.22222ms (frametime)

 

Lower FPS means more time processing a frame... the 290X was consistent at outputting more frame rates hence the higher AVERAGE (out of 25 titles cmon!).. what am I getting so wrong here?

http://www.pcper.com/reviews/Graphics-Cards/Frame-Rating-New-Graphics-Performance-Metric

[EdInk]

http://www.pcper.com/reviews/Graphics-Cards/Frame-Rating-New-Graphics-Performance-Metric

WOW! an article from JANUARY 2013 that may not have materialised... I'm ECSTATIC :rolleyes:.. we are in 2016 right?

[Enderman]

WOW! an article from 2013 that may not have materialised... I'm ECSTATIC :rolleyes:.. we are in 2016 right?

its actually true

linus talked about it a lot too

 

the only reason why its not actually used is because of the cost and how complex it is to measure

 

this is more than just "oh 60fps is 16.66ms per frame"

this is about "why am i getting solid 60fps but the game looks unnatural"

[EdInk]

its actually true

linus talked about it a lot too

 

the only reason why its not actually used is because of the cost and how complex it is to measure

 

this is more than just "oh 60fps is 16.66ms per frame"

this is about "why am i getting solid 60fps but the game looks unnatural"

There's a thread on LTT about lower quality images on newer GPU's...(Nvidia&AMD) Soooo? what's this 'unnatural' angle you are bringing to the table?

[ivan134]

Because in reality it isn't, despite what average framerate might tell you. I regretted buying the 290X. I still own it, but I use the 970 in my main PC. Average fps is just part of the story. Framerate and frametime consistency are much more important for the gaming experience.

Which 290x do you have?