NVIDIA’s VXAO (voxel based ambient occlusion) To Be Featured In Rise of the Tomb Raider

[spartaman64]

No, I said it did matter. ivan was the one saying results within margin of error (which he incorrectly used) didn't matter.

 

EDIT: That bench is 3 months old and invalid after both the launch of Crimson and more than 6 major Nvidia driver releases. I'll see if I can find something from at least mid November.

ok fair enough but you also need to take into consideration that the 980 is around 500 dollars and the 390x is around 400 dollar price point at this moment but i suppose that your argument for this is that price doesnt matter even if you pay 100 dollars extra for 1 fps difference but you must see how ridiculous that argument is. you call other people fanboys but look back at your previous posts you attack my character instead of my argument at one point and you ignore evidence and use only evidence that is convenient for you are you sure that you are not the one acting like a fanboy. just consider the possibility

[LokiFire]

I prefer venison. Here in New Jersey there's an ample supply.

 

WWII France vs. Hitler. Italy was taken over by a dictator. That's a bit of a different scenario.

You avoided my question, killing a dead rabbit is not possible, it is already dead.

 

To me i just think you like the drama.

[patrickjp93]

ok fair enough but you also need to take into consideration that the 980 is around 500 dollars and the 390x is around 400 dollar price point at this moment but i suppose that your argument for this is that price doesnt matter even if you pay 100 dollars extra for 1 fps difference but you must see how ridiculous that argument is. you call other people fanboys but look back at your previous posts you attack my character instead of my argument at one point and you ignore evidence and use only evidence that is convenient for you are you sure that you are not the one acting like a fanboy. just consider the possibility

 

I didn't attack your character. I attacked the decision making process as terribly shallow and said people who use it alone frankly aren't worthy of reproducing in my mind. 

 

I didn't ignore evidence. I'm the one who provided it ffs. I ignore your incorrect usage of that evidence or even attack and dismantle it. I'm an academic. Evidence is the basis of my graduate career up to this point. Watch where you sling mud.

 

You avoided my question, killing a dead rabbit is not possible, it is already dead.

 

To me i just think you like the drama.

If you subscribe to the bible, a second death is possible.

[spartaman64]

I didn't attack your character. I attacked the decision making process as terribly shallow and said people who use it alone frankly aren't worthy of reproducing in my mind. 

 

I didn't ignore evidence. I'm the one who provided it ffs. I ignore your incorrect usage of that evidence or even attack and dismantle it. I'm an academic. Evidence is the basis of my graduate career up to this point. Watch where you sling mud.

 

If you subscribe to the bible, a second death is possible.

really none of that statement about people who think like that is "shallow" and should be "removed from the gene pool by force" is directed towards me and you just threw that statement in on a whim ok you're the only one who knows the true intention of that statement so ill take your word for it but you understand how that can be interpreted that way in the context. also you used the fable benchmark as evidence for your point but when i showed how it is evidence for mine as well you were quick to dismiss the evidence as invalid which is a bit odd because as you said you provided it. the problem i think is that your argument is too broad and the other person arguing against you has the same problem. of course there is no way that nvidia is better in all cases across the board just like there is no way that amd is better in all cases so you need to focus on something more tangible like if you said the 980 ti is better than the fury x i would agree with you and if a friend asked me for a recommendation for a 600-650 dollar gpu i would recommend the 980 ti but for a 400-500 dollar gpu i would recommend the 390x over the 980

[Prysin]

http://www.techspot.com/review/1081-dx11-vs-dx12-ashes/page3.html

 

Nope, all tiers until 960 vs. 370, unless you can find one benchathon more recent.

those results are seriously fishy. like massively so.

my 295x2 scored higher with a single 290X active.... mind you, i tested AOTS at 3440x1440.... if i get around the same FPS with a single 290X at that resolution as those benches do at 1080p...

I gonna warn you Patrick, be careful with techspot, their benchmark has as of late been a bit off. by notable margins. We saw this with TW3 benches when comparing against Digital Foundry and some other benches, techspot scores were much different with equal settings. I know Techspot has a reputation of being solid, but for the past 8-9 months, they have been a bit weird in their results.

Just a heads up.

[patrickjp93]

really none of that statement about people who think like that is "shallow" and should be "removed from the gene pool by force" is directed towards me and you just threw that statement in on a whim ok you're the only one who knows the true intention of that statement so ill take your word for it but you understand how that can be interpreted that way in the context. also you used the fable benchmark as evidence for your point but when i showed how it is evidence for mine as well you were quick to dismiss the evidence as invalid which is a bit odd because as you said you provided it. the problem i think is that your argument is too broad and the other person arguing against you has the same problem. of course there is no way that nvidia is better in all cases across the board just like there is no way that amd is better in all cases so you need to focus on something more tangible like if you said the 980 ti is better than the fury x i would agree with you and if a friend asked me for a recommendation for a 600-650 dollar gpu i would recommend the 980 ti but for a 400-500 dollar gpu i would recommend the 390x over the 980

 

Am I allowed to assume people who post on LTT are a bit more sophisticated than that? In basic artificial intelligence, you already learn to use multiple evaluation functions in a deterministic system. Neural Networks are a more sophisticated approach, but the principles remain the same. Humans are built to make complex logical decisions, not take the easy way out. It's a sign of evolutionary weakness to do so.

 

No, because the benches a week later throw everything out of whack, but both are 3 months old, so let's exclude them for the sake of sanity. I don't know why no one's revisited them since.

[patrickjp93]

those results are seriously fishy. like massively so.

my 295x2 scored higher with a single 290X active.... mind you, i tested AOTS at 3440x1440.... if i get around the same FPS with a single 290X at that resolution as those benches do at 1080p...

I gonna warn you Patrick, be careful with techspot, their benchmark has as of late been a bit off. by notable margins. We saw this with TW3 benches when comparing against Digital Foundry and some other benches, techspot scores were much different with equal settings. I know Techspot has a reputation of being solid, but for the past 8-9 months, they have been a bit weird in their results.

Just a heads up.

The 295x2 cores are clocked higher than reference and even most aftermarket 290Xs. It's not fishy at all having 1 core active.

 

Thanks for the heads-up all the same.

[Prysin]

The 295x2 cores are clocked higher than reference and even most aftermarket 290Xs. It's not fishy at all having 1 core active.

 

Thanks for the heads-up all the same.

ill downclock my 295x2 to stock 290X, then ill run it with single core active.

i own AOTS, so ill do teh DX12 test, as those in particular is a bit off.

sadly, i have zero Nvidia cards to compare with, as i own zero Nvidia cards. I would have to buy one (which i might do as my mITX rigs HD 7950 died due to careless handling on my part)...

Sadly, there is no Nvidia card in the mid/low range worth buying, as over here, we got R9 380s selling for as low as 950s... i mean... 950 vs 380... cmon man, its not even a fucking competition.

[patrickjp93]

ill downclock my 295x2 to stock 290X, then ill run it with single core active.

i own AOTS, so ill do teh DX12 test, as those in particular is a bit off.

sadly, i have zero Nvidia cards to compare with, as i own zero Nvidia cards. I would have to buy one (which i might do as my mITX rigs HD 7950 died due to careless handling on my part)...

Sadly, there is no Nvidia card in the mid/low range worth buying, as over here, we got R9 380s selling for as low as 950s... i mean... 950 vs 380... cmon man, its not even a fucking competition.

 

I know. The 950 is just so much better.

 

Double check the memory clock too.

[Prysin]

I know. The 950 is just so much better.

nice try....

what is better is the 960 Asus blower.

not the 960, but the blower itself. Damn thats sexy.

[spartaman64]

Am I allowed to assume people who post on LTT are a bit more sophisticated than that? In basic artificial intelligence, you already learn to use multiple evaluation functions in a deterministic system. Neural Networks are a more sophisticated approach, but the principles remain the same. Humans are built to make complex logical decisions, not take the easy way out. It's a sign of evolutionary weakness to do so.

 

No, because the benches a week later throw everything out of whack, but both are 3 months old, so let's exclude them for the sake of sanity. I don't know why no one's revisited them since.

ahh i see you are saying there are other factors but lets say that the cards are identical with the only variable being price which one would you choose? this was only to show that price to performance isnt completely useless as you say it is of course there are other factors but you shouldnt completely exclude this factor. also you are rationalizing somehow i doubt your statement about taking me out of the gene pool by force is genuine concern to better the genetic situation of the human race but i digress

 

also what i noticed about your arguments with me is that you like absolutes which is something i understand as i once thought that way also. fedora is better than ubuntu, nvidia is better than amd but think about it are there really no cases where ubuntu is more suitable than fedora? and are there no instances in which amd is better than nvidia? the world isnt black and white like that. 

[pbenologa]

Stop feeding that kid. Seriously. Just ignore it. It will go away.

[Zah]

sadly, i have zero Nvidia cards to compare with, as i own zero Nvidia cards. I would have to buy one (which i might do as my mITX rigs HD 7950 died due to careless handling on my part)...

Sadly, there is no Nvidia card in the mid/low range worth buying, as over here, we got R9 380s selling for as low as 950s... i mean... 950 vs 380... cmon man, its not even a fucking competition.

 

I would test it with my 980ti, but sadly I do not own the game at this time.

[Prysin]

I would test it with my 980ti, but sadly I do not own the game at this time.

its just a few bucks on steam...

[Zah]

its just a few bucks on steam...

50 USD is not a few bucks.

[Notional]

No revision and no backtracking. Nvidia wins, hands down.

 

Async compute can be done under both. Do your research. It's EXPOSED (not required) by both DX 12 and 11.3.

 

11.3 is basically DX12, but without the need for a large knowledge of graphics architecture. However, we know for a fact that Maxwell is useless at async compute. We even have official statements from the developers of AOTS, that they had to replace async compute threads, with proprietary NVidia shaders to get proper performance on NVidia cards. It's simply incorrect. Maxwell is even more useless at concurrent Async compute, which is where we are headed, and where the performance benefit lies.

 

You do know the margin of error shrinks with a larger sample size, right? That's how statistics works and why political polls only need a couple thousand to have a margin of error less than 2%. The error rate of the benchmark can be mitigated by having multiple runs and taking the average. When the margins of error for each card no longer overlap, there's a clear winner. It's Nvidia.

 

You also know you cannot compare one benchmark with another. Weighted averages are incredibly imprecise, why no self respecting scholar would use it for anything conclusive. I don't doubt the specific proprietary shaders implemented on AOTS for NVidia is doing the trick. What I'm saying, is that it requires a lot of extra work for everyone, and proves that Maxwell is useless at compute.

 

No, they support it 100% the way I said they did. Just because you don't like the fact Nvidia can win in DX 11 vs. AMD's DX 12 implementation at this current moment in time doesn't change the results. The results are Nvidia wins, by however small a margin.

 

I've got no doubt that the 980ti can beat the Fury X by a tiny tiny bit, but that is hardly impressive compared to the price differences. If games starts to use concurrent async compute instead of shaders, we will see AMD cards having the upper hand. I'm sure NVidia will replace the code with shaders or bully the devs into doing it for them, but like We've seen with Kepler, as soon as focus is being removed from the drivers, the performance tanks on older NVidia cards. They simply get shittier with time, due to planned obsolescence; something you've commended NVidia for doing.

 

Considering how gimped Maxwell is on the hardware side, and how dependant it is on driver optimizations per game, I fear that that card will crash and burn in performance after Pascal is released. Maybe an extra year of good support, but even the Fury X should surpass the 980ti when Pascal is here.

 

WWII France vs. Hitler. Italy was taken over by a dictator. That's a bit of a different scenario.

 

Actually France was the only country that tried to stop the German military build up. The English however had no interest in stopping the Germans, because they believed the German army would put France in check, so they weren't a threat to the English. GOOD JOB ENGLAND!

France was also the mainland country that resisted the most, and made dday possible. So it's hardly fair.

[Solkia]

I'm getting kind of tired of graphical gimmicks being used to market PC ports.

 

I actually kind of miss when the only thing used to advertise PC ports was TF2 hats no one used.

[_Grid21]

NVIDIA’s VXAO To Be Featured In Rise of the Tomb Raider

 

 

 

Great. More proprietary tech that is going to kill fps in games probably. GG

 

Source:http://wccftech.com/nvidias-vxao-to-be-featured-in-rise-of-the-tomb-raider-heres-what-we-know/

http://news.xbox.com/2016/01/05/rise-of-the-tomb-raider-coming-windows-store-january-28/

 

Well Just Cause 3 already has this feature built in so I don't know their trying to keep it a secret. 

[suicidalfranco]

Italians, actually.  My people are industrious in making use of bread, tomatoes, and dairy products, but we definitely flip-flop when it comes to saving our arses.

technically, we won every war we took part of

[Mahigan]

Patrick...two words... AMD Crimson.

The results you showed were with the Catalyst 15.7 drivers. 15.7 drivers were before the performance improvements of the 15.10 drivers. Which means AMD was beating nvidia after the 15.10 drivers. Crimson brought a 10-15% performance boost, on top of the 15.10 drivers, under Ashes of the Singularity. Look up the Crimson reviews for more info.

AMD GPUs now lead the pack under Ashes of the Singularity.

Its been that way since November 30.

Anyway, none of this will matter soon enough. We'll have NVIDIA Pascal release followed by AMD Greenland. As things stand right now, Crimson has leveled the playing field even in several DX 11 titles.

Have a look yourself:

http://www.techpowerup.com/reviews/Gigabyte/GTX_980_Ti_Waterforce/23.html

What we have is this..

R9 290=GTX 970, R9 390, R9 290x, R9 390x=GTX 980, Fury, FuryX=GTX 980 Ti, Factory Overclocked GTX 980 Ti.

AMD have made enormous strides. The company that was, in the beginning of 2015, is not the company we're seeing now.

2016 is shaping up to be a very good year for RTG.

That about sums it up right now.