Gears of War 4 Benchmark

[guachi]

Tests like this make me very happy I bought an 8350 for $100 just this year.

Also highlights something Russian Sensation is always carrying on about - namely that the i7 is a much better CPU choice than the i5.

Perhaps these new games using more CPU bode well for AMD's new Zen chips in games.

If the 8300 series is still competent, Zen will hopefully be competitive with the current i7.

 

[2is]

Tests like this make me very happy I bought an 8350 for $100 just this year.

Also highlights something Russian Sensation is always carrying on about - namely that the i7 is a much better CPU choice than the i5.

Perhaps these new games using more CPU bode well for AMD's new Zen chips in games.

If the 8300 series is still competent, Zen will hopefully be competitive with the current i7.

It'll be competitive when both are GPU bound. As has been the case for about the last decade.

 

[Bacon1]

It'll be competitive when both are GPU bound. As has been the case for about the last decade.

Both are GPU Bound. Or do you think people spent $1000 for a 5960x for 1080p?

 

[2is]

Both are GPU Bound. Or do you think people spent $1000 for a 5960x for 1080p?

So do you agree or disagree?

 

[Bacon1]

So do you agree or disagree?

I think both are GPU bound @ 1440p or larger. And both are GPU bound on anything under a Titan XP or maybe 1080. If you are spending $2k+ to play @ 1080p... you could have a much better IQ.

It can run @ ~125 fps as shown in the Titan XP graph, any card that can run faster than that is money wasted.

 

[AtenRa]

Has this been posted before ??

http://www.techspot.com/review/1263-gears-of-war-4-benchmarks/

Ultra Doesn't Always Mean Ultra
Something I noticed when testing many GPUs was the difference in image quality, despite the settings remaining the same. Forcing the settings to ultra, there were still distinct differences between the 3GB and 6GB versions of Nvidia's GeForce GTX 1060 for example.

The character resolution seemed lower at times on the 3GB model. It was almost like a different anti-aliasing method was being used. The characters would appear blurrier and feature lower textures. The environments would often feature lower quality textures as well, less lighting and lower quality shadows.

 

[Bacon1]

Has this been posted before ??

http://www.techspot.com/review/1263-gears-of-war-4-benchmarks/

Yeah thats why IQ checks are critical to testing.

Isn't there a "dynamic memory" type option in GoW4 though?

Maybe the dynamic resolution scaling was left enabled?

 

[RussianSensation]

good review of performance and game here.
http://www.babeltechreviews.com/gears-war-4-campaign-review-pc-performanceiq-analysis/

Stop linking this NV-biased source. We are talking about a guy who crapped all over one of the most legendary GPUs of all time - the HD 7970. He called it short-lived videocard and over time removed his entire article. This is pure gold:

"The most interesting thing about this launch is that it appears that Tahiti’s HD 7970 may be short-lived. AMD’s 28nm yields are known to not be great and the specifications of the new GPU indicate that something strange is going on. The chip itself is quite complex with 4.3 billion transistors but its core only runs at a very conservative 925MHz. Most telling is that its performance is not that great except in DX11 games and especially in heavily tessellated scenarios – only about 25% faster than a GTX 580 overall." ~ Reviewer credibility drops to 0.

On launch date, with "crappy" AMD drivers, 7970 OC ended up 48-80% faster than GTX580 OC.

And now this has been revealed:

"I run my benches a bit differently than other sites which use completely default settings for their HW. I have consistently benchmarked my games with the Power Draw/Limit and Temperature targets of all hardware (both AMD and Nvidia) set to their maximum which means my results tend to be a little higher for the high TDP cards." ~ More here

Translation - since Kepler era, he has run NV OC GPUs against AMD GPUs operating at stock because of how NV vs. AMD GPU boost works.

There is only 1 consistency about BTR or ABT GPU reviews after BFG left -- they are consistently biased, with settings manipulated to make sure NV cards win.

 

[Bacon1]

Yeah thats why IQ checks are critical to testing.

Isn't there a "dynamic memory" type option in GoW4 though?

Maybe the dynamic resolution scaling was left enabled?

After looking at the full sized images:

1060 3GB:

1060 6GB:

460 2GB:

460 4GB:

I'm going to say yes, they did have that setting enabled.

1) FPS Counter and Text at top of the image are both larger on the two AMD cards, which makes sense if it was scaling the game down (or they used a lower resolution all together).

2) Shadows are completely broken on the 3GB 1060 and 2GB 460, but even worse on the 1060 3GB.

Look at the house in the top left, it's just a black box placed over for the shadow on the 3GB 1060. Both the 1060 3GB and 2GB 460 also have lots of random black spots all over the ground.

 

[PhonakV30]

You mean Similar to Nvidia's MRS? I think It's just bug.

 

[SolMiester]

Stop linking this NV-biased source. We are talking about a guy who crapped all over one of the most legendary GPUs of all time - the HD 7970. He called it short-lived videocard and over time removed his entire article. This is pure gold:

"The most interesting thing about this launch is that it appears that Tahiti’s HD 7970 may be short-lived. AMD’s 28nm yields are known to not be great and the specifications of the new GPU indicate that something strange is going on. The chip itself is quite complex with 4.3 billion transistors but its core only runs at a very conservative 925MHz. Most telling is that its performance is not that great except in DX11 games and especially in heavily tessellated scenarios – only about 25% faster than a GTX 580 overall." ~ Reviewer credibility drops to 0.

On launch date, with "crappy" AMD drivers, 7970 OC ended up 48-80% faster than GTX580 OC.

And now this has been revealed:

"I run my benches a bit differently than other sites which use completely default settings for their HW. I have consistently benchmarked my games with the Power Draw/Limit and Temperature targets of all hardware (both AMD and Nvidia) set to their maximum which means my results tend to be a little higher for the high TDP cards." ~ More here

Translation - since Kepler era, he has run NV OC GPUs against AMD GPUs operating at stock because of how NV vs. AMD GPU boost works.

There is only 1 consistency about BTR or ABT GPU reviews after BFG left -- they are consistently biased, with settings manipulated to make sure NV cards win.

Utter BS yet again RS!..It appears your reading comprehension on the reason of a short lived GPU is lacking!

I just love the way you twist your opinions as fact to prove an argument. Dont forget your legendary GPU had broken CF drivers for the first 18mths.

 

[moonbogg]

Bought it. BOOYA. Will mess with it later. Downloading now.

 

[moonbogg]

3440x1440 all ultra except ambient occlusion on high. Min 60, Average 73. CPU render was right around 210. Game seems to run great. Nice!

 

[Carfax83]

3440x1440 all ultra except ambient occlusion on high. Min 60, Average 73. CPU render was right around 210. Game seems to run great. Nice!

Great. Just remember that the game looks much better once you pass the prologue.. By the time you get to Act III, you're going to think that Doom has nothing on Gears 4

 

[moonbogg]

I can't figure out how to "cover-slide-hit-F-attack-blah blah" and all that JASS. I'll get it. I am in the first levels of the first act. Prologue looked pretty good to me. I have good feelings about the game. Like it so far. No issues with it. Pretty shocking.

 

[2is]

The campaign has got a bit too much chaos for my tastes, but it does run well even in very heavy action. I played one MP match and that was more to my liking.

 

[vissarix]

Stop linking this NV-biased source. We are talking about a guy who crapped all over one of the most legendary GPUs of all time - the HD 7970. He called it short-lived videocard and over time removed his entire article. This is pure gold:

"The most interesting thing about this launch is that it appears that Tahiti’s HD 7970 may be short-lived. AMD’s 28nm yields are known to not be great and the specifications of the new GPU indicate that something strange is going on. The chip itself is quite complex with 4.3 billion transistors but its core only runs at a very conservative 925MHz. Most telling is that its performance is not that great except in DX11 games and especially in heavily tessellated scenarios – only about 25% faster than a GTX 580 overall." ~ Reviewer credibility drops to 0.

On launch date, with "crappy" AMD drivers, 7970 OC ended up 48-80% faster than GTX580 OC.

And now this has been revealed:

"I run my benches a bit differently than other sites which use completely default settings for their HW. I have consistently benchmarked my games with the Power Draw/Limit and Temperature targets of all hardware (both AMD and Nvidia) set to their maximum which means my results tend to be a little higher for the high TDP cards." ~ More here

Translation - since Kepler era, he has run NV OC GPUs against AMD GPUs operating at stock because of how NV vs. AMD GPU boost works.

There is only 1 consistency about BTR or ABT GPU reviews after BFG left -- they are consistently biased, with settings manipulated to make sure NV cards win.

The same legendary gpu that got owned by the cheaper smaller more power efficient gtx680?

 

[coercitiv]

The same legendary gpu that got owned by the cheaper smaller more power efficient gtx680?

When making such remarks it's usually a good idea to check what happened immediately after with the competition. It helps avoid setting yourself up for an easy comeback.

 

[Keysplayr]

When making such remarks it's usually a good idea to check what happened immediately after with the competition. It helps avoid setting yourself up for an easy comeback.

That still wouldn't make what he said untrue. A lot of things happened from then to now.

 

[Carfax83]

When making such remarks it's usually a good idea to check what happened immediately after with the competition. It helps avoid setting yourself up for an easy comeback.

Perhaps you should take your own advice. Immediately after? More like a year and a half later if anything.. Up till the time developers started using compute more heavily and games started being designed for the PS4 and Xbox One, Kepler was still doing very well against GCN. And by the time the writing was on the wall and the pendulum began to swing in the other direction, Maxwell was released and started a new cycle of NVidia domination.

 

[crisium]

So I don't get it, was Tahiti "short lived" because GK104 came out 2.5 months later and they competed neck and neck for 2.5 years? And how is that a bad thing for Tahiti's longevity? Outside of the initial undercutting GK104 actually cost more than Tahiti for most of this time frame anyhow.

Or is Tahiti in reality "long lived" because after 2.5 years of neck and neck we came to November 2014 and onward when Kepler became a coin flip and Tahiti won the war in the long term for gamers still using them from that point vs the competition?

I guess it depends on your partisan power level how much people want to fight on this hill. I was always under the impression ABT was banned from here and unreliable. But I guess they are valid since Tahiti can apparently be interpreted to have been short lived.

 

[coercitiv]

Perhaps you should take your own advice. Immediately after?

Can you honestly claim that Tahiti was a short-lived chip? Because that's what being debated, whether it was competitive for a significant period of time, not whether it came out on top.

Up till the time developers started using compute more heavily and games started being designed for the PS4 and Xbox One, Kepler was still doing very well against GCN.

Your own words describe a product that managed to stay relevant for a long period of time.

I'll end my OT here, since GoW4 doesn't deserve this.

 

[96Firebird]

Good job derailing another thread with whining about a review site, some people just never learn.

 

[2is]

So I don't get it, was Tahiti "short lived" because GK104 came out 2.5 months later and they competed neck and neck for 2.5 years? And how is that a bad thing for Tahiti's longevity? Outside of the initial undercutting GK104 actually cost more than Tahiti for most of this time frame anyhow.

Or is Tahiti in reality "long lived" because after 2.5 years of neck and neck we came to November 2014 and onward when Kepler became a coin flip and Tahiti won the war in the long term for gamers still using them from that point vs the competition?

I guess it depends on your partisan power level how much people want to fight on this hill. I was always under the impression ABT was banned from here and unreliable. But I guess they are valid since Tahiti can apparently be interpreted to have been short lived.

As an owner of a GTX 680 as well as a 7970, I can say with confidence Tahiti was not at all short lived. I had 2 680's. One I gave away to a friend, the other either sits in a closet collecting dust or gets tossed in a "spare parts build" if I'm bored while my 7970 has a home in my 2nd gaming machine which still does great at 1080p.

 

[PhonakV30]

quote deleted

lol? what is this link ? be ready for Mod!

Please do not repost spam links.
administrator allisolm