gtx 980 or fury for 1440p ?

[RussianSensation]

OOS tried to purchase it.

Ya they go really fast. You gotta be on top of your game

If you have a bunch of old PC parts that you don't mind selling and you don't have to have a DIY system, you can try this 850e ENVY 5820K system with a 980. Supposedly comes with CLC cooling and the mobo apparently allows some overclocking to 4.2ghz+. I am having a really hard time matching this deal with the OS with a DIY 5820K + 980 rig.

http://slickdeals.net/f/8045263-hp-...on-free-shipping-i7-4790-gtx980-for-875ar?v=1

 

[tential]

It's either a Bstock Evga at this point, or a GTX 980Ti or I just wait to the node drop.

Will they have Bstock GTX 908Tis ever?

 

[RussianSensation]

It's either a Bstock Evga at this point, or a GTX 980Ti or I just wait to the node drop.

Will they have Bstock GTX 908Tis ever?

They might have them in 2016 but then there will be Pascal

I think 970/390/290X + resale and reinvesting $180-200 saved now towards a 16nm GPU is still a good option as well. If he is gaming on a 7950, that's still a great upgrade that doesn't destroy his wallet and gives him plenty of $ towards his next upgrade. If the OP gamed that long on a 7950, he is OK turning down some settings and doesn't require 60 fps constant at all times. In that case, I can't stress enough how awesome this upgrading strategy is for him.

Looking back, did GTX580 outlast GTX470/570, or did HD7970Ghz outlast the much cheaper HD7950 or did GTX780Ti outlast 780/290? Nope. If you look at history of GPU upgrades and GPU obsolescence, as long as you get a card with 80% of the performance of the next tier cards, everything in that generation will become outdated at the same time. In that sense I do not see any value in $480-550 980/Fury over say a $300 390 given what next gen should bring and especially how much faster the $650 980Ti is.

 

[tential]

Given his frame rate requirement, I agree the R9 390 is definitely the best choice.

I hope AMD updates it with 4K VSR.... I would have crossfired them already if that feature hadn't been removed.

 

[3DVagabond]

I went from 7950 to 780 specifically for 1440. It really depends on what you play. There is a bit of difference between Fury and 980 but I don't think it's enough to justify the price.

Don't forget to look at 390x too

You need to look at the individual games. In many the difference is pretty substantial.
This is just the first six games in TPU's test suite. The 980 is really more the peer of the 390X and not even close to fury.

It's because of a couple of games like these that are so handicapped against AMD that the overall even looks close. I mean these games are so far out that even the 970 is faster than FuryX.

Again, look at the whole review, and unless it's these three games you are playing, the 980 isn't competitive with Fury. It's the 390X that's it's competition.

 

[Silverforce11]

You need to look at the individual games. In many the difference is pretty substantial.
This is just the first six games in TPU's test suite. The 980 is really more the peer of the 390X and not even close to fury.

Again, look at the whole review, and unless it's these three games you are playing, the 980 isn't competitive with Fury. It's the 390X that's it's competition.

And the other NV sponsored title: COD

If they had Dying Light as well, it would have been even in the summary chart.

 

[Carfax83]

Again, look at the whole review, and unless it's these three games you are playing, the 980 isn't competitive with Fury. It's the 390X that's it's competition.

Perhaps you should take your own words under advisement. It's more than just "three games."

And the 980 is definitely competitive with the Fury, which the review doesn't reflect as it's limited. For instance with a GTX 980, you can use MFAA. If the review had enabled that, it would have skewed the benchmarks in NVidia's favor since it boosts MSAA performance substantially..

Secondly, nobody is going to buy a reference clocked 980 when the 980 has such a high TDP ceiling.. You can have factory overclocked GTX 980s that boost as high as 1430MHz automatically..

Thirdly, TechPowerUp's reviews always need to be taken with a grain of salt, because their driver database is way behind the curve. They are using the 353.06 drivers. Yesterday, NVidia released 355.60 drivers..

All that and I still recommend the R9 Fury over the 980 at 1440p, because it has higher stock bandwidth courtesy of HBM, plus the potential for greater performance increases down the road since it's a newer architecture.

Although I believe the potential performance improvements via driver optimization will not be as high this time around, as most of the improvements in Hawaii over Kepler came from it's compute focused architecture, and the fact that more and more games are using compute shaders.

Maxwell addressed the compute performance that Kepler lacked, so I expect Maxwell to mature much better than Kepler did..

 

[3DVagabond]

Perhaps you should take your own words under advisement. It's more than just "three games."

And the 980 is definitely competitive with the Fury, which the review doesn't reflect as it's limited. For instance with a GTX 980, you can use MFAA. If the review had enabled that, it would have skewed the benchmarks in NVidia's favor since it boosts MSAA performance substantially..

Secondly, nobody is going to buy a reference clocked 980 when the 980 has such a high TDP ceiling.. You can have factory overclocked GTX 980s that boost as high as 1430MHz automatically..

Thirdly, TechPowerUp's reviews always need to be taken with a grain of salt, because their driver database is way behind the curve. They are using the 353.06 drivers. Yesterday, NVidia released 355.60 drivers..

All that and I still recommend the R9 Fury over the 980 at 1440p, because it has higher stock bandwidth courtesy of HBM, plus the potential for greater performance increases down the road since it's a newer architecture.

Although I believe the potential performance improvements via driver optimization will not be as high this time around, as most of the improvements in Hawaii over Kepler came from it's compute focused architecture, and the fact that more and more games are using compute shaders.

Maxwell addressed the compute performance that Kepler lacked, so I expect Maxwell to mature much better than Kepler did..

I was posting in response specifically to the post I quoted. Read it before you want to post rebuttal that has nothing to do with that.

 

[Carfax83]

I was posting in response specifically to the post I quoted..

Which changes nothing. Your post is still erroneous.. In that specific review, the 980 is competitive with the Fury in more than just three games..

 

[3DVagabond]

Which changes nothing. Your post is still erroneous.. In that specific review, the 980 is competitive with the Fury in more than just three games..

It's more competitive with the 390X.

 

[DustinBrowder]

Which changes nothing. Your post is still erroneous.. In that specific review, the 980 is competitive with the Fury in more than just three games..

No its NOT!!!

The Fury wins in 7 out of 11 games, the games it loses 3 of those have Nvidia gameworks code to obviously destroy AMD performance arbitrarily. I bet if you removed gameworks and the games had clean code they would have better performance under AMD cards.

The games 980 wins are old games anyways, who cares about those games anyways? AMD Fury wins in all of the newer and most popular games like Witcher, Batman, GTA, etc...

 

[Carfax83]

No its NOT!!!

The Fury wins in 7 out of 11 games, the games it loses 3 of those have Nvidia gameworks code to obviously destroy AMD performance arbitrarily. I bet if you removed gameworks and the games had clean code they would have better performance under AMD cards.

The games 980 wins are old games anyways, who cares about those games anyways? AMD Fury wins in all of the newer and most popular games like Witcher, Batman, GTA, etc...

I think you need to look up what the word "competitive" actually means.

Also, I see that you complain about Gameworks, but not GamingEvolved.

There are seven AMD sponsored games in that review, so I guess NVidia is automatically disadvantaged am I right?

You know what, why am I even bothering to have a conversation with you? D:

 

[Carfax83]

It's more competitive with the 390X.

It is more competitive with the 390x, but that doesn't mean it's not competitive with the R9 Fury. Being competitive isn't about winning all the benchmarks. It's about keeping up and not getting left behind.

The R9 Fury is able to pull ahead in most of the benchmarks, but not by very large percentages..

Also as I said, using a GTX 980 with reference clocks makes the GTX 980 look a lot worse than it actually is.. Similarly, not using MFAA and using older drivers also makes the GTX 980 look worse than it really is.

TPU's habit of using out of date drivers actually favors AMD, since AMD doesn't release drivers as often as NVidia..

 

[flopper]

Fury is overpriced right now, IMO. It's a great card and a better performer than the 980, but not worth the extra money over a 390x/980. The 980 tends to overclock really well and the 390x tends to overclock well also but will use more power.

.

390 OC a bit seems to me the wise choice overall for a gamer today.

 

[you2]

This is probably sound advice I will follow. I should have said when I made the post is that I will naturally wait for a good price (sale or whatever) hence the reason I asked if prices are the same which card. I plan to make the purchase between tomorrow and blackfriday. Yes I realize that in the worse case I will end up waiting 3 months but I can be patient since my current system is functional. Also I might end up replacing more than just the video card (the current hardware is 2500k based and the motherboard has a few nagging issues; then there is the side question of whether I want to upgrade to windows 10 (currently on 7). The 390 is tempting if it is fast enough for the 27inch monitor.

Anyway, I would still get a B-Stock EVGA 980 for $375. Another possibility is you grab a B-Stock GTX970 or get a GTX970/290/390 and just resell that card and upgrade in 2016/early 2017 to 16nm HBM2 GPU. I personally don't think either the 980 or the Fury are worth their prices. The only cards that make sense to me at regular prices are 290/290X/390 and 980Ti. Everything else is overpriced (I do think $375 for the eVGA 980 is a good deal but you gotta be quick!). :thumbsup:

If you are not in a rush to get a new card, just keep an eye out here:
http://www.evga.com/Products/ProductList.aspx?type=8

Another option is to apply $25 off $200 with Newegg's Visa Checkout but you have to wait until any gifts disappear from the promotion. Just find a 390 with a $20-30 rebate and wait until any promotional gift cards are cleared so you can apply the visa checkout. You could score yourself an R9 390 for $275 that way.

At $275-280, an after-market 390 is a smoking value for 1440P as a great stop-gap card between now and 16nm GPUs.

http://www.guru3d.com/articles_pages/powercolor_radeon_r9_390_pcs_8gb_review,1.html

You get the point. If it were my money, unless you can get a 980 for $370-390, just get a $280-300 390 as it's a huge upgrade over your 7950 and resell that 390 card in 1.5 years or so, while reinvesting the resale value and $180-200+ saved from NOT buying a 980/Fury towards a 16nm GPU that is bound to be way faster than a 980/Fury and have more features. Since you are happy with 40-60 fps, this is the best deal by far and gives you a ton of money left over for your next GPU upgrade.

Remember, look back at any generation you want and you will see 10-15% differences in performance never count towards playability in the future. What that means is if you have to pay $180-200 extra for just 15% more performance, you are wasting your money -- far better to get a card with 90% of the performance in this case and reinvest the savings towards a next gen card 1.5-2X faster. 390 is the sweet spot.

 

[3DVagabond]

It is more competitive with the 390x, but that doesn't mean it's not competitive with the R9 Fury. Being competitive isn't about winning all the benchmarks. It's about keeping up and not getting left behind.

The R9 Fury is able to pull ahead in most of the benchmarks, but not by very large percentages..

Also as I said, using a GTX 980 with reference clocks makes the GTX 980 look a lot worse than it actually is.. Similarly, not using MFAA and using older drivers also makes the GTX 980 look worse than it really is.

TPU's habit of using out of date drivers actually favors AMD, since AMD doesn't release drivers as often as NVidia..

You have to be purposely avoiding the point of my post just so you can go off on your rhetorical rant. I posted benches from the games that made up the overall performance chart. You on the other hand are talking about metrics that don't apply to the review and then trying to interject them. It would be like me complaining that W1zzard should have unlocked the Fury to a full Fury X. Completely irrelevant to the point.

 

[you2]

To be honest; for me it is about being fast enough not being fastest. If there is slight slow down to 35 or 40 frames a second I'm not going to worry about it and spend an extra $100 or $200. So showing a comparison of 100 fps vs 200 fps is meaningless to myself (i realize that is not the case here so please don't take this as being direct at you; just the general discussion).
-
As for the type of games I play - mostly rpg (witcher 3, DAI, D:OS, wasteland, ...) and a few fps (wolfenstein) and a few tbs like xcom (frame rate just doesn't matter here).
-
Even in the links below the 390/390x look not bad unless they have some other major flaw that should be pointed out. One concern I do have is card length; I have a short 7950. My case sucks as it doesn't support long cards very well due to drive cage position (of course I could always get a new case but I kind of like it). Also I ahve a 600 watt seasonic psu but I think all these cards as single card solution will work fine with that.

You need to look at the individual games. In many the difference is pretty substantial.
This is just the first six games in TPU's test suite. The 980 is really more the peer of the 390X and not even close to fury.

It's because of a couple of games like these that are so handicapped against AMD that the overall even looks close. I mean these games are so far out that even the 970 is faster than FuryX.

Again, look at the whole review, and unless it's these three games you are playing, the 980 isn't competitive with Fury. It's the 390X that's it's competition.

 

[Carfax83]

You have to be purposely avoiding the point of my post just so you can go off on your rhetorical rant. I posted benches from the games that made up the overall performance chart. You on the other hand are talking about metrics that don't apply to the review and then trying to interject them. It would be like me complaining that W1zzard should have unlocked the Fury to a full Fury X. Completely irrelevant to the point.

I'm not avoiding anything. I was responding to a direct statement of yours, as well as the overall implication of your post which was that:

1) The GTX 980 is not competitive with the R9 Fury.

2) Were it not for a few pro NVidia games, the delta would be positively humungus.

This flies in the face of the actual review which shows the R9 Fury only gaining a large performance lead over the GTX 980 in about 2 or three titles, and mostly because TPU uses "copious" amounts of MSAA in their reviews which plays to the Fury's bandwidth advantage..

And them not using MFAA certainly doesn't help the GTX 980 either..

 

[Azix]

Everybody has their way of looking at it. For me its if cards are close in performance and price I want that extra bit. Like Fury and overclocked 980. For the peace of mind. For the 390x vs 390 it's harder given the huge gap in price and smaller gains. I would and did get a 290x 8GB (still need to have delivered). That should match a 390x with minor OC for similar or less than a 390. I wanted those extra shaders even if it's not a whole lot.

If you can wait that long then you might as well. Things could change significantly between now and blackfriday. There could be game bundles along with dropped prices and combos.

There is still the nano to be released and its short.

Got to say, because techpowerup insists on using so many gameworks/nvidia games, I find the results problematic. Especially their summaries. Those would typically be useful, but then it's all tainted by the biased results. Individual charts are fine. A lot of other sites avoid those games, but offer no summary.

outdated drivers do not favor AMD if they use outdated AMD drivers. Fewer drivers means more significance to each driver. Whilst nvidia might put out more, there is likely less of a change in each and some might just be sli

 

[vissarix]

EVGA GTX980 for $375 with shipping. Has 1 year warranty, if purchased with Amex, get an extra 1 year warranty.

However, if paying $480+, get the Fury because it takes more or less a max overclocked 980 to match a stock Fury at 1440P on average unless you are playing mostly NV-favoured games like WOW, Project CARS and GameWorks titles. Otherwise, you are going to need to overclock the 980 to match a stock Fury at 1440P:

I dont know why people take you seriously, your clearly biased and if there is a unreliable and biased review on the internet your going to find it at 100% to back up your bs as usual...

on this pic the r9 290 is within 1% performance difference vs the gtx970?! :hmm: and the r9 290x is 7% faster vs the gtx970?!...now this clearly shows something isnt right. :sneaky:

the gtx970 was constantly beating the r9 290x thats why amd lowered the price below the gtx970 msrp, because it was slower (without taking into account other factors)...and they lost huge market share because of this..

Also the gtx980ti only 3%faster vs fury x at 1440p? , those biased reviewers are showing more like 15%+ but hey, they dont know.

try harder next time, you might find a ''very reliable, unbiased'' review 1 out of 100 on which every amd gpu beats the nvidia counterpart by a good margin

 

[tential]

Some good points being brought up about TPU's testing methodology's (with game selection and with drivers/etc.)

Hope the OP finds all of this useful and can pick the right card for him now, I think we've all put more than enough info to decide.

Considering the performance per dollar of the R9 390 over every single card out there it's hard not to recommend it.

I mean look here:
http://www.tweaktown.com/tweakipedia/91/amd-radeon-r9-390x-crossfire-4k/index.html

Crossfire 390x review but lets just pretend we knock off a couple of FPS here and there and it's a 390 crossfire review.

With the ability to add that second card and still come out under the GTX 980 Ti, come DANGEROUSLY close to the Fury's price ($600 CF brand new, but you'll add that second card probably from the used market later so it'll be cheaper and actually MATCH the Fury price), why not take the R9 390 instead?

I'd have already taken the R9 390s if not for the 4K vsr, but for OP at 1440p, I'd pick up one R9 390 as chances are, the second we all see the nodeshrink's GPU performance, I don't think ANY of us will care about a GPU we currently own. HBM + nodeshrink, I don't think many of us will be holding onto 28nm DDR5/HBM gpus in 2016 come the new cards.

Edit: It's a different game suite.
And yes, the GTX 980Ti is not much faster than Fury X as a REFERENCE card. Look at those non reference cards. That's the point of the GTX 980 Ti. That the nonreference card is $30 more, so of COURSE you would spend that $30 for a massive increase in performance.

If you're going to call a review biased/unfair, don't just look at the performance summary. GO look through the review and write a reasoning why it's unfair/biased. People have pointed out flaws in TPU's review setup (I wouldn't say the results are horrible, I'd say that you should take the flaws into consideration). If you have issues with their testing setup, go ahead and LIST THEM OUT.

I'll do some of the work for you, or you could have just read RS's post, WoW/Project Cars are included in TPU's benches. I think WoW is a perfectly acceptable MASSIVE game people can bench. Project Cars.... considering how many people have purchased it, I don't get why everyone insists on benching it. But the enthusiast community does like these types of games. So it really just depends on the games you play, I think TPU is great for most gamers at the high end who probably also buy a lot more games and buy those games early too.

What that graph shows to me is how great of a decision it is to always WAIT for non reference cards.

 

[Carfax83]

Got to say, because techpowerup insists on using so many gameworks/nvidia games, I find the results problematic. Especially their summaries. Those would typically be useful, but then it's all tainted by the biased results. Individual charts are fine. A lot of other sites avoid those games, but offer no summary.

At least get your facts straight before you accuse TPU of bias... :thumbsdown:

In the TPU review, there are eight AMD Gaming Evolved titles:

1) Alien Isolation

2) Tomb Raider

3) Crysis 3

4) Bioshock Infinite

5) Civilization Beyond Earth

6) Dragon Age Inquisition

7) GTA V (Co-sponsored with NVidia)

8) Battlefield 4

For NVidia's Gameworks, there are eight as well:

1) AC Unity

2) Batman Arkham Origins

3) Far Cry 4

4) Witcher 3

5) Metro Last Light

6) Project Cars

7) Watch Dogs

8) GTA V (Co-sponsored with AMD)

9) CoD Advanced Warfare (Checked and this isn't a GW title. CoD Ghosts is, but not Advanced Warfare.

Yeah, that's so many Anyway, whether a game has Gameworks or GamingEvolved does not determine final performance or give the sponsor an advantage or their competitor a disadvantage. There are gameworks titles where AMD is ahead of NVidia, and GamingEvolved titles where NVidia is leading AMD..

outdated drivers do not favor AMD if they use outdated AMD drivers. Fewer drivers means more significance to each driver. Whilst nvidia might put out more, there is likely less of a change in each and some might just be sli

NVidia are constantly tweaking and optimizing their drivers. A review using a driver that is two three revisions out of date is understandable, but TPU uses drivers that are months old. The driver they used in that review came out in June..

 

[cmdrdredd]

It depends on what games you're looking at specifically. If you want to use any Nvidia specific features in those games then you have no choice.

 

[Azix]

At least get your facts straight before you accuse TPU of bias... :thumbsdown:

In the TPU review, there are eight AMD Gaming Evolved titles:

1) Alien Isolation

2) Tomb Raider

3) Crysis 3

4) Bioshock Infinite

5) Civilization Beyond Earth

6) Dragon Age Inquisition

7) GTA V (Co-sponsored with NVidia)

8) Battlefield 4

For NVidia's Gameworks, there are eight as well:

1) AC Unity

2) Batman Arkham Origins

3) Far Cry 4

4) Witcher 3

5) Metro Last Light

6) Project Cars

7) Watch Dogs

8) GTA V (Co-sponsored with AMD)

9) CoD Advanced Warfare (Checked and this isn't a GW title. CoD Ghosts is, but not Advanced Warfare.

Yeah, that's so many Anyway, whether a game has Gameworks or GamingEvolved does not determine final performance or give the sponsor an advantage or their competitor a disadvantage. There are gameworks titles where AMD is ahead of NVidia, and GamingEvolved titles where NVidia is leading AMD..

I didn't say they were biased did I? I said they use a lot of gameworks games that can skew results. It makes the most useful charts questionable. Individual game charts are still fine.

A game being a GE game means nothing much beyond that AMD helped optimize the game. AMD doesn't do the same things nvidia does. currently their involvement does not seem to result in unfairly skewed framerates.

NVidia are constantly tweaking and optimizing their drivers. A review using a driver that is two three revisions out of date is understandable, but TPU uses drivers that are months old. The driver they used in that review came out in June..

You seem to think nvidia makes significant changes in these drivers (to performance). I do sometimes say nvidia needs driver updates more often than AMD, but really if you look at their release notes there's not usually much about single GPU performance. Not that I have found anyway. They don't seem as open about their driver changes though.

 

[Ma_Deuce]

A game being a GE game means nothing much beyond that AMD helped optimize the game. AMD doesn't do the same things nvidia does. currently their involvement does not seem to result in unfairly skewed framerates.

Exactly. Two completely different programs with different goals. Using the amount of titles from each program isn't an effective way to show that a review site is not biased.