[BENCHMARKS] Need For Speed [2016] - AMD seems to have the lead again.

[Morgan MLGman]

So a Polish site (very reliable btw) did benchmarks of the upcoming Need For Speed game (which looks AMAZING btw) and the results are as in the three previous cases in a row in AAA titles performance, AMD has the lead.

What's weird is that for none of these games AMD released a Game-Ready driver (because apparently there was no need to), I'm not sure if Nvidia has one for NFS already though. They did.

Drivers used:

• AMD Radeon Crimson 16.3
• GeForce Game Ready GeForce 364.51 WHQL

Test bench consists of i7-4790K @4,5gHz, 16GB of 2400mhz DDR3 RAM, and a 1TB Samsung 850 Pro SSD

So here are the results:

1440p High settings:

 

1080p Ultra Settings:

 

Spoiler

 

1080p High Settings:

 

Spoiler

 

 

1080p Medium Settings:

 

 

Spoiler

 

 

 

 

What do you guys think? Did you expect those results? Let me know in the comments.

Link to the full article (it's in Polish though): http://www.purepc.pl/karty_graficzne/test_wydajnosci_need_for_speed_komputer_z_nitro_mile_widziany

 

If it's a repost, then sorry and please ignore it, search button didn't let me find any NFS benchmarks on the forum.

 

[Moonzy]

3 minutes ago, Morgan MLGman said:

GeForce Game Ready GeForce 364.51 WHQL

iBlameThis

jk, or am i?

[Morgan MLGman]

2 minutes ago, Moonzy said:

iBlameThis

jk, or am i?

I checked, turns out the 365.51 WHQL is indeed a Game-Ready driver for Need For Speed.

[byalexandr]

1 minute ago, huilun02 said:

AMD needs to get their shit together for their drivers. Nvidia always had better driver support.

??? They are excelling here... Nvidia is the one at fault.

 

 

YAY FOR AMD! SUCK MY ASS NVIDIA!

[beebskadoo]

1 minute ago, huilun02 said:

AMD needs to get their shit together for their drivers. Nvidia always had better driver support.

Have you seen Nvidias drivers lately?

[Moonzy]

2 minutes ago, Morgan MLGman said:

 

yea but its a sht driver lol

caused dozens nothing but trouble

 

dont mind me, just bashing the driver, not either companies

[beebskadoo]

Just now, Moonzy said:

yea but its a sht driver lol

caused dozens nothing but trouble

 

dont mind me, just bashing the driver, not either companies

Both companies suck equally

[Moonzy]

Just now, beebskadoo said:

Both companies suck equally

ikr, intel iris pro ftw!

#IrisProMasterRace

[beebskadoo]

Just now, Moonzy said:

ikr, intel iris pro ftw!

#IrisProMasterRace

pshhhh

Voodoo SLI OG master 1337 haxorrace

[Kobathor]

This is really interesting. nVidia needs to step up their driver optimization, even their game-ready drivers are failing them in newer games versus month-old drivers from AMD. I still know nVidia is going to sell more GPU's though, since their brand is so popular.. 

[Nena Trinity]

I paid more for better driver support yet Nvidia have been a let down whole year... (AMD next time maybe...)

[i_build_nanosuits]

lol that's bullshit i'm running the game with all the settings maxed out at 1440p and i get around 85 to 115FPS...mostly around 105FPS i would say

the game run awesomly well and the graphics are breath taking...but the controls are super arcade it's a bit of a downer...overall a good game but was expecting more...might need to play more as well though...cars customisation is amazing though you can do anytihnng.

[i_build_nanosuits]

havn't seen the game go as low as any of the numbers in those benchmarks and i played around 6 to 8 hrs of it so far...lowest i've seen is around 80FPS.

So i don't know if those numbers are pulled out of somebody's butt crack? or if AMD really has that low performance....but i'm telling you 60/68.9FPS at 2560x1440 on a GTX 980ti ain't happening...nowhere near...you'd have to supersample from 4K to get in that FPS range. i can make a video when i'm home tonight for you guys to see.

DO NOT BELIEVE EVERYTHING YOU READ ON THE INTERNET GUYS!!

[Morgan MLGman]

8 minutes ago, i_build_nanosuits said:

lol that's bullshit i'm running the game with all the settings maxed out at 1440p and i get around 85 to 115FPS...mostly around 105FPS i would say

the game run awesomly well and the graphics are breath taking...but the controls are super arcade it's a bit of a downer...overall a good game but was expecting more...might need to play more as well though...cars customisation is amazing though you can do anytihnng.

If an NFS doesn't have an advanced car customisation system then it's not worth playing to me With the exception of NFS: Shift because I liked this game a lot.
Hope it's at least a little similar in that regard to NFS: Underground 2 because it was the best game ever.

2 minutes ago, i_build_nanosuits said:

havn't seen the game go as low as any of the numbers in those benchmarks and i played around 6 to 8 hrs of it so far...lowest i've seen is around 80FPS.

So i don't know if those numbers are pulled out of somebody's butt crack? or if AMD really has that low performance....but i'm telling you 60/68.9FPS at 2560x1440 on a GTX 980ti ain't happening...nowhere near...you'd have to supersample from 4K to get in that FPS range. i can make a video when i'm home tonight for you guys to see.

DO NOT BELIEVE EVERYTHING YOU READ ON THE INTERNET GUYS!!

I don't know how they did the benchmarks exactly, but I know how they do it in general. One dude responsible for benchmarking takes the cards to his home where he's got a test bench and benchmarks them one after another for a day or two, depending on how many cards they test and then they post the scores on their website, it's a very popular tech-site here

But I guess if the testing was flawed in some way, all of the cards had their scores diminished, not only a 980Ti because that would just be dumb and looked stupid on the benchmark chart (they'd see that something was wrong instantly) so the pattern pretty much stays. It's more about how the cards corelate than about 'absolute' numbers of FPS ^^

[i_build_nanosuits]

18 minutes ago, Morgan MLGman said:

If an NFS doesn't have an advanced car customisation system then it's not worth playing to me With the exception of NFS: Shift because I liked this game a lot.
Hope it's at least a little similar in that regard to NFS: Underground 2 because it was the best game ever.

I don't know how they did the benchmarks exactly, but I know how they do it in general. One dude responsible for benchmarking takes the cards to his home where he's got a test bench and benchmarks them one after another for a day or two, depending on how many cards they test and then they post the scores on their website, it's a very popular tech-site here

But I guess if the testing was flawed in some way, all of the cards had their scores diminished, not only a 980Ti because that would just be dumb and looked stupid on the benchmark chart (they'd see that something was wrong instantly) so the pattern pretty much stays. It's more about how the cards corelate than about 'absolute' numbers of FPS ^^

ill shoot a quick video tonight like i did for arkham knight where i'll show my settings and the big OSD with all the stats...you'll see what i mean.

The game run beautifuly and it's the (IMHO) most photo-realistic scenery and effects so far in a video game...in some areas it just look so real and so good...the car sounds are fantastic too..but the controls man...you run into a steep turn at 275kmh/h and just hit the brakes 50 meters from the curve and you'll be good....drift this out and off you go...it's VERY arcade...but it run well and it looks and sound awesome...and like i said 60FPS @ 2560x1440 is not possible without supersampling (DSR)

The only setting i did not maxed out is motion blur...cause it's ugly and bad...so maybe that's the ONE setting that would kill performance that much? but i doubt it...

[Morgan MLGman]

1 minute ago, i_build_nanosuits said:

ill shoot a quick video tonight like i did for arkham knight where i'll show my settings and the big OSD with all the stats...you'll see what i mean.

The game run beautifuly and it's the (IMHO) most photo-realistic scenery and effects so far in a video game...in some areas it just look so real and so good...the car sounds are fantastic too..but the controls man...you run into a steep turn at 275kmh/h and just hit the brakes 50 meters from the curve and you'll be good....drift this out and off you go...it's VERY arcade...but it run well and it looks and sound awesome...and like i said 60FPS @ 2560x1440 is not possible without supersampling (DSR)

What do you mean by that last part?

And yeah, as I said, benchmarks might be flawed in some way so the results are lower than they are in reality, but if they were flawed, it's all of the cards and not only 980Ti that are impacted, looking forward to the video

[Monarch]

1 hour ago, Kobathor said:

This is really interesting. nVidia needs to step up their driver optimization, even their game-ready drivers are failing them in newer games versus month-old drivers from AMD. I still know nVidia is going to sell more GPU's though, since their brand is so popular.. 

 

40 minutes ago, Nena360 said:

I paid more for better driver support yet Nvidia have been a let down whole year... (AMD next time maybe...)

Nvidia's driver optimization is still better, it's just driver isn't the bottleneck in NFS, which makes optimization irrelevant and so better hardware is able to win.

 

28 minutes ago, i_build_nanosuits said:

havn't seen the game go as low as any of the numbers in those benchmarks and i played around 6 to 8 hrs of it so far...lowest i've seen is around 80FPS.

So i don't know if those numbers are pulled out of somebody's butt crack? or if AMD really has that low performance....but i'm telling you 60/68.9FPS at 2560x1440 on a GTX 980ti ain't happening...nowhere near...you'd have to supersample from 4K to get in that FPS range. i can make a video when i'm home tonight for you guys to see.

DO NOT BELIEVE EVERYTHING YOU READ ON THE INTERNET GUYS!!

It's because your 980 Ti runs at much higher core clock than 1164. Apparently they used reference cards.

 

 

 

[i_build_nanosuits]

Just now, Morgan MLGman said:

What do you mean by that last part?

And yeah, as I said, benchmarks might be flawed in some way so the results are lower than they are in reality, but if they were flawed, it's all of the cards and not only 980Ti that are impacted, looking forward to the video

The only setting i did not maxed out is motion blur...cause it's ugly and bad...so maybe that's the ONE setting that would kill performance that much? but i doubt it...

so in order for me to get in the 60 to 68FPS area like your benchmark suggest would be to downsample the game from a higher resolution (IE running it at 4K using nvidia DSR technology and downsampling it to 2560x1440)

---

[i_build_nanosuits]

1 minute ago, Monarch said:

 

Nvidia's driver optimization is still better, it's just driver isn't the bottleneck in NFS, which makes optimization irrelevant and so better hardware is able to win.

 

It's because your 980 Ti runs at much higher core clock than 1162. Apparently they used reference cards.

 

 

 

i run at 1430mhz...i doubt this would give me what? like a + 30 or +40FPS compaired to the benchmarks he posted...

[i_build_nanosuits]

3 minutes ago, Morgan MLGman said:

...but if they were flawed, it's all of the cards and not only 980Ti that are impacted, looking forward to the video

...or maybe AMD wrote them a check and they tested squadoosh?

i will do a video if nvidia experience and shadowplay work...i had my first issues with them this week now when i open nvidia experience it just search for new drivers and never found anything its just searching so i couldnt access anything...i just hope when i open it tonight it will be fine

[Monarch]

1 minute ago, i_build_nanosuits said:

i run at 1430mhz...i doubt this would give me what? like a + 30 or +40FPS compaired to the benchmarks he posted...

Well it depends on the area you're in as well, but 1164 vs 1430 MHz is a huge difference.

[i_build_nanosuits]

2 minutes ago, Monarch said:

Well it depends on the area you're in as well, but 1164 vs 1430 MHz is a huge difference.

i played at least 5 or 6 hours of the game so far...and never seen my FPS go bellow 70FPS...even 75.

Also there are times where the 980ti is not even maxed out...pointing a CPU bottleneck...so that makes it even less likely to see AMD cards perform better than nvidia in that title. The test he showed are either flawed or complete BS i'm 100% sure of that for owning the game and having played with it for the last 2 days.

[Morgan MLGman]

Just now, i_build_nanosuits said:

...or maybe AMD wrote them a check and they tested squadoosh?

i will do a video if nvidia experience and shadowplay work...i had my first issues with them this week now when i open nvidia experience it just search for new drivers and never found anything its just searching so i couldnt access anything...i just hope when i open it tonight it will be fine

Well, I dont' know. But wouldn't that make R9 Fury X be 30-40 FPS behind a 980Ti? Everyone knows it's much closer than that
 

5 minutes ago, i_build_nanosuits said:

The only setting i did not maxed out is motion blur...cause it's ugly and bad...so maybe that's the ONE setting that would kill performance that much? but i doubt it...

so in order for me to get in the 60 to 68FPS area like your benchmark suggest would be to downsample the game from a higher resolution (IE running it at 4K using nvidia DSR technology and downsampling it to 2560x1440)

---

Ah, okay, I understand now

The motion blur shouldn't change anything basically (besides making the game uglier), although there are settings that can cause that, like HairWorks in The Witcher 3 or that nvidia interactive smoke that caused your arkham knight FPS to drop dramatically, that shouldn't be the case in NFS though

[ivan134]

This was expected. Frostbite loves AMD Hardware. DICE and AMD are BFFs.

[Monarch]

2 minutes ago, i_build_nanosuits said:

i played at least 5 or 6 hours of the game so far...and never seen my FPS go bellow 70FPS...even 75.

So around 70 fps min vs 60 fps in that benchmark, that's around 15% difference. Sounds about right for 200+ MHz core clock difference. But it's dumb that they test reference cards. Who buys those? Most people buy non reference cards that have much higher clocks.