The Fury X vs. the GTX 980 Ti Revisited with 36 Games including Watch Dogs 2

[happy medium]

http://www.babeltechreviews.com/

I'll comment later, I'm busy.

Our chart provides a lot of information across 36 games and 2 synthetics. What we can take away from the results generally is that the GTX 1070 is still the fastest single GPU video card of the three cards with the GTX 980 Ti in second place.

Overall, the GTX 980 Ti is still significantly faster than the Fury X in the majority of our games although the Fury X has been able to gain a little ground – 7 additional benches out of the 75 we originally tested are now in the Fury X’ favor – although it is still bested by the GTX 980 Ti overall. We don’t see the GTX 980 Ti losing any ground in the older games to the GTX 1070, although the newer card pulls further ahead in some of the newest games.

We continue to see good optimizations being made for the GTX 980 Ti although they appear smaller than for the GTX 1070. And as AMD’s flagship, we see AMD’s driver team continue to optimize it, making good on their progress to manage its limited 4GB of vRAM rather well. The games where Fury X had issues at 4K – especially with Assassin’s Creed Syndicate and GTA V – are now playing much better now than they were 6 months ago.

 

[Carfax83]

Weird how they used the latest drivers for AMD, but not for NVidia. I mean, if they included Watch Dogs 2 they should have used the 376.09 drivers..

 

[pj-]

Weird how they used the latest drivers for AMD, but not for NVidia. I mean, if they included Watch Dogs 2 they should have used the 376.09 drivers..

Judging by the ridiculous amount of tests done, they began before 376.09 was released

 

[happy medium]

Weird how they used the latest drivers for AMD, but not for NVidia. I mean, if they included Watch Dogs 2 they should have used the 376.09 drivers..

quote:
Nvidia’s GeForce WHQL 375.86 drivers used for the GTX 980 Ti and the GTX 1070 cards except for the latest GeForce WHQL 376.06 driver for Watch Dogs 2. Older drivers compared 368.22 for the GTX 980 Ti and 368.19 for the GTX 1070 launch.

 

[Bacon1]

His numbers look off, but not surprising with the source.


GoW he gets much higher than Fury X for some reason for example

http://www.pcgameshardware.de/Gears-of-War-4-Spiel-55621/Specials/Performance-Test-Review-1209651/

http://www.gamegpu.com/action-/-fps-/-tps/gears-of-war-4-test-gpu

http://www.guru3d.com/articles_page..._graphics_performance_benchmark_review,6.html

Maybe its because he tweaks the driver settings?

http://www.babeltechreviews.com/fury-x-vs-gtx-980-ti-revisited-36-games-including-watch-dogs-2/2/

 

[Phynaz]

His numbers look off, but not surprising with the source.

Can you explain further what is wrong with the source, or are you just trolling?

 

[esquared]

OK guys, stop it now. I have had 3 reported posts on this thread already.

Let it go.


esquared
Anandtech Forum Director

 

[HurleyBird]

I don't like the methodology to be honest.

As a guy with a 4K screen I'm most interested in the 4K results, but when you max out every setting so that you're getting unplayable or at least poorly playable frametimes in nearly every instance, what's the point?

I'm not saying to go as far as copying HardOCP's "highest playable settings," (I prefer an apples-to-apples comparison anyway) but it would be nice if the settings at 4K resolution were at least somewhat realistic for what people are doing in the real world.

 

[poofyhairguy]

The Nvidia 980 Ti was one of the best flagship GPUs in years for sure.

 

[JoeRambo]

The Nvidia 980 Ti was one of the best flagship GPUs in years for sure.

Yeah, i am very happy with mine. And tests were done with reference edition as well, that was not known for stellar perf or overclocks. 15-20% of perf is hidden for 980TI ( I know Fury can OC some as well, but 5-10% maybe? )

 

[Sven_eng]

never mind

 

[Elfear]

Can you explain further what is wrong with the source, or are you just trolling?

He's referring to the owner and author of the linked review site who has had major issues with AMD in the past. Lots of conspiracy theories, anger over not being sent AMD review hardware, and his benchmark results didn't line up with most other reviews.

 

[kawi6rr]

He's referring to the owner and author of the linked review site who has had major issues with AMD in the past. Lots of conspiracy theories, anger over not being sent AMD review hardware, and his benchmark results didn't line up with most other reviews.

Can't take that serious considering the source, OP should have known better.

 

[tential]

Can't take that serious considering the source, OP should have known better.

Why not? I took it seriously. I mean I obviously know the source is what it is but it's just another reference point. You know the source leans towards Nvidia but the results look good for Fury X. Maybe I'm reading this chart completely wrong?
When I opened this thread, I actually got sad I didn't buy Fury X. For $275, with a site like this, Fury X still is competitive to a 980Ti?

I mean, we know the OP and the history (or many of us do), so that's how I viewed it.
-----------------
When I say I looked at the chart, I only looked at 4K results... Just to be clear since that's the only resolution I care about.

 

[nurturedhate]

Why not? I took it seriously. I mean I obviously know the source is what it is but it's just another reference point. You know the source leans towards Nvidia but the results look good for Fury X. Maybe I'm reading this chart completely wrong?
When I opened this thread, I actually got sad I didn't buy Fury X. For $275, with a site like this, Fury X still is competitive to a 980Ti?

I mean, we know the OP and the history (or many of us do), so that's how I viewed it.
-----------------
When I say I looked at the chart, I only looked at 4K results... Just to be clear since that's the only resolution I care about.

When the reference point has a known history of being rather flawed and deviates so far from the standard norm it would be wise to through out those results or at the very least view them from that context. If the goal is to accurately gauge the difference in performance between 980ti and Fury X then this source should not be used due to it's overwhelming history as other more reputable sources could have been used. If the goal is to simply state 980ti is for the most part superior to Fury X then we didn't need this particular data point to do that and any of the dozens of more reputable sites would have been adequate in conveying that knowledge. That leaves us wondering what the purpose of using a known flawed data point was to begin with. When trying to state a hypothesis or develop an argument it is the data that is questioned first, not the conclusion. If the data is bad the conclusion does not matter.

 

[IllogicalGlory]

The Nvidia 980 Ti was one of the best flagship GPUs in years for sure.

Can't argue with that. GM200 is a total monster, as long as it has good cooling and power delivery. Practically every aftermarket 980 Ti delivers a 15-20% performance improvement over reference on average. NV really shot themselves in the foot with the reference cooling and thermal limits. It's a completely different card when it's properly built. 30% faster than reference rather than about 8%.

Judging from the OC results on Titan XP with its power limit increased, I think GP102 is going to behave similarly.

Regarding the results, isn't this the guy who tests cards with the power limit increased? The results aren't surprising if so. The 980 Ti is overclocked.

 

[happy medium]

He's referring to the owner and author of the linked review site who has had major issues with AMD in the past. Lots of conspiracy theories, anger over not being sent AMD review hardware, and his benchmark results didn't line up with most other reviews.

Do you have a source ?
lets derail the thread some more and show his conspiracy theories, anger over not being sent hardware, and where his benchmarks don't line up with other reviews.
Show me his unfounded MAJOR issues with AMD.
SHow me some links to back up your accusations.
Show me some facts, links, real evidence not someones opinion.

 

[tajoh111]

He's referring to the owner and author of the linked review site who has had major issues with AMD in the past. Lots of conspiracy theories, anger over not being sent AMD review hardware, and his benchmark results didn't line up with most other reviews.

I think your mistaking him for kitguru. Kitguru was the one with a bit of beef.

 

[happy medium]

I think your mistaking him for kitguru. Kitguru was the one with a bit of beef.

That's not the problem. The problem is someone posting in the thread with nothing constructive to say and derailing the thread. Not one thing about the content of the review but attacking the reviewer. Sound familiar? How many other threads do we have with people attacking the messenger or the reviewer? Its a thread derailment and off topic.

Would it be ok for me to voice my opinion that elfear is a AMD lover and a Nvidia bashing piece of crap in a Fury X review thread after one of his posts?

No, it would be off topic, trolling ,and derailing to the thread.
Well I reported the post, I'm done here. Chances are the thread will be closed and flushed down the toilet anyway. rinse and repeat.

 

[tential]

The Nvidia 980 Ti was one of the best flagship GPUs in years for sure.

Imagine the 980Ti clocked at 1450mhz on a techpowerup price to performance ratio charge back in its prime.... The thing made NO SENSE. So good....
I know the 1080Ti won't do that. Why should it? Nvidia has no reason to give away so much performance. The 980Ti could have easily launched at a much higher pricepoint. If the 1080Ti is less than $750 I'll be surprised.
At 4K though, Fury X looks good. Oh well, it's too late for that level of performance for me. 1080Ti vs Vega come tomorrow please!

 

[casiofx]

Imagine the 980Ti clocked at 1450mhz on a techpowerup price to performance ratio charge back in its prime.... The thing made NO SENSE. So good....
I know the 1080Ti won't do that. Why should it? Nvidia has no reason to give away so much performance. The 980Ti could have easily launched at a much higher pricepoint. If the 1080Ti is less than $750 I'll be surprised.
At 4K though, Fury X looks good. Oh well, it's too late for that level of performance for me. 1080Ti vs Vega come tomorrow please!

I would guess titan xp is the ceiling for 1080ti's performance

 

[iiiankiii]

No doubt about it. The Fury X is trash. I was totally disappoint with the Fury Line up. The GTX 980TI is a fine card. I just have to wonder how well it be by the end of 2017 when DX12 becomes more common place among AAA titles. The lack of asynchronous compute will only get worst as Nvidia moves away from specifically optimizing for Maxwell. That's why even though it's a bit stronger than the GTX1070 right now, I would not get a GTX980TI over a GTX1070.

 

[Bacon1]

That's not the problem. The problem is someone posting in the thread with nothing constructive to say and derailing the thread. Not one thing about the content of the review but attacking the reviewer. Sound familiar? How many other threads do we have with people attacking the messenger or the reviewer? Its a thread derailment and off topic.

I was discussing the data if you were referring to my post. I was pointing out that his testing methodolgy is flawed and his numbers are farther off than any other reviewer.

As for the quality of that site, its been discussed multiple times even directly with you responding to the posts...

https://forums.anandtech.com/thread...rx-480-120w-249.2478605/page-75#post-38393205

and

https://forums.anandtech.com/threads/closed.2468583/page-6#post-38138722

and other places as well

 

[monkeydelmagico]

Imagine the 980Ti clocked at 1450mhz on a techpowerup price to performance ratio charge back in its prime.... The thing made NO SENSE. So good....

Yep, and when 1070 was released they marked down all the 980ti's. I picked up a new Zotac Amp! Extreme for around $400.-. It came out of the box with a 25% OC and I've pushed it higher to around 1500mhz. At the time you couldn't get a decent 1070 for anything less than $460.-.

So far so good, but it remains to be seen if the 980ti will represent a good value decision

 

[happy medium]

Yep, and when 1070 was released they marked down all the 980ti's. I picked up a new Zotac Amp! Extreme for around $400.-. It came out of the box with a 25% OC and I've pushed it higher to around 1500mhz. At the time you couldn't get a decent 1070 for anything less than $460.-.

So far so good, but it remains to be seen if the 980ti will represent a good value decision

I was going to grab a used one for ~$320 but for me 250 watts and more WHEN I overclock the snott out of it, is a little to much heat.

1070's are in the $370 range now and are an excellent buy and way less heat. I've also notice in the more recent benchmark the 1070 is pulling away.