NVIDIA Pascal Thread

[desprado]

Oh boy,

You didn't read those articles eh?

They're talking about the upcoming DX12 patches for Just Cause 3 which will add ROV and CR.

As for stutters, do you have a recent link discussing stutters or just the usual "at launch" stutter issues that affect all Gameworks titles running on AMD Radeon due to there being no AMD driver optimizations for reasons previously mentioned in this thread and others?

Did you read properly that they are using some features in it. I Know you have personal feelings for AMD but fact is a fact.

 

[Mahigan]

Did you read properly that they are using some features in it. I Know you have personal feelings for AMD but fact is a fact.

Please link me to these AMD users having issues with Just Cause 3 so I can compare them to the NVIDIA GTX 900 series users having the same issues here: https://www.reddit.com/r/JustCause/comments/3wy3ux/just_cause_3_massive_frame_drops_and_stuttering/

I was able to play without any problems using an amd radeon r9 290 gpu but since upgrading my pc with an nvdea gtx 980Ti the game is virtually unplayable due to stuttering! I've tried various adjustments without any luck so far. Very frustrating

Must suck to be this guy eh?

 

[desprado]

Please link me to these AMD users having issues with Just Cause 3 so I can compare them to the NVIDIA GTX 900 series users having the same issues here: https://www.reddit.com/r/JustCause/comments/3wy3ux/just_cause_3_massive_frame_drops_and_stuttering/



Must suck to be this guy eh?

http://www.techtimes.com/articles/1...d-graphics-try-this-beta-says-square-enix.htm

http://www.gamespot.com/articles/just-cause-3-pc-graphics-problems-reported/1100-6432700/

https://www.reddit.com/r/Amd/comments/3w86sp/just_cause_3_performance_issues/

https://www.youtube.com/watch?v=QIdtPfn5uv4

That is why this has so many negative review and 90% of them are AMD owners.

 

[Mahigan]

http://www.techtimes.com/articles/1...d-graphics-try-this-beta-says-square-enix.htm

http://www.gamespot.com/articles/just-cause-3-pc-graphics-problems-reported/1100-6432700/

https://www.reddit.com/r/Amd/comments/3w86sp/just_cause_3_performance_issues/

https://www.youtube.com/watch?v=QIdtPfn5uv4

That is why this has so many negative review and 90% of them are AMD owners.

Well, good news is you're taking a situation which affected both NVIDIA and AMD users and trying to paint it as an AMD issue.

The links you provided talk of issues for NVIDIA and AMD users. Oh and this is a DX11 game btw not DX12.

Issues, for AMD users, were fixed with the Crimson 15.11.1 drivers.

The 8GB system memory issues remain for both AMD and NVIDIA users. This game uses upwards of 9GB of RAM. Best to install 16GB minimum.

So you're talking about past issues.

 

[desprado]

Well, good news is you're taking a situation which affected both NVIDIA and AMD users and trying to paint it as an AMD issue.

The links you provided talk of issues for NVIDIA and AMD users. Oh and this is a DX11 game btw not DX12.

Issues, for AMD users, were fixed with the Crimson 15.11.1 drivers.

The 8GB system memory issues remain for both AMD and NVIDIA users. This game uses upwards of 9GB of RAM. Best to install 16GB minimum.

So you're talking about past issues.

No i am just telling you why GTX 980 Ti stock is performing 20% faster then Fury X on 1440p.

 

[tential]

http://www.techtimes.com/articles/1...d-graphics-try-this-beta-says-square-enix.htm

http://www.gamespot.com/articles/just-cause-3-pc-graphics-problems-reported/1100-6432700/

https://www.reddit.com/r/Amd/comments/3w86sp/just_cause_3_performance_issues/

https://www.youtube.com/watch?v=QIdtPfn5uv4

That is why this has so many negative review and 90% of them are AMD owners.

Posts across social media and message boards cite, in particular, a lower-than-desired frame rate. One Reddit user claimed the "game is unplayable" despite apparently booting it on a PC with a high-end processor (Intel i7 5820k) and graphics card (Nvidia GTX 980Ti). "I get 10fps," the Reddit user alleged.

http://www.gamespot.com/articles/just-cause-3-pc-graphics-problems-reported/1100-6432700

So.... you were telling me how great this game was on a 980Ti/Nvidia

Nvidia video cards manifested stuttering and crashing, but in fewer numbers. Nvidia made an attempt to resolve the issues and released a driver for Just Cause 3 and Rainbow Six Siege, but feedback on Steam was inconclusive, as some members of the community still complained that the driver did not fix the problems.

http://www.techtimes.com/articles/1...d-graphics-try-this-beta-says-square-enix.htm
Your own links prove the point that the game was broken on BOTH hardware at launch.... I mean... did you even read the articles you posted?

 

[tential]

No i am just telling you why GTX 980 Ti stock is performing 20% faster then Fury X on 1440p.

When was that ever the discussion?

Edit: In fact, what is going on, I thought this was completely different thread. This isn't even about PAscal....

 

[Mahigan]

No i am just telling you why GTX 980 Ti stock is performing 20% faster then Fury X on 1440p.

Under Just Cause 3 DX11?

A stock GTX 980 Ti performs 5.4% faster than a FuryX. A factory overclocked GTX 980 Ti performs costing over $700 (around $70-100 more than a FuryX) like the Asus Geforce GTX 980 Ti Matrix performs 27% faster.

It's a well known fact that AMD suffer API overhead under DX11. Which is why when you move to 4K, GPU bound scenario, the FuryX pulls ahead of a stock GTX 980 Ti by 4%:

I'm not sure what you're trying to argue here.

 

[Mahigan]

When was that ever the discussion?

Edit: In fact, what is going on, I thought this was completely different thread. This isn't even about PAscal....

I'm equally as confused as you are...

 

[crisium]

I have 16 hours of JC3 and don't have any stutter.

 

[alcoholbob]

Hmm no Titan Pascal until 2017? Would be disappointing if true. I don't know anyone who would upgrade for a 10% perf upgrade for GP104 from a 980 Ti or 295X2, it's basically GTX 780 -> GTX 980 all over again.

Another year of no single card 4K.

 

[thesmokingman]

AdamK47, to further prove your point:

Doesn't look like a regression here, FWIW.

You must be a comedian? A bench does not equal real world game scaling. In the real world getting scaling is like mining for gold. Good scaling is out there but that's like in a handful of games... so it's hardly realistic to expect or even boast about how great one's scaling is.

 

[Arachnotronic]

You must be a comedian? A bench does not equal real world game scaling. In the real world getting scaling is like mining for gold. Good scaling is out there but that's like in a handful of games... so it's hardly realistic to expect or even boast about how great one's scaling is.

Yep, I work part-time as Bozo the Clown at the state fair during the summer. The kids love it when I show benchmarks that some people claim are not representative of real-world multi-GPU scaling -- it really gets the "lulz."

 

[xthetenth]

Dig into AMD's history over the last decade or so. If after doing so you think this is a goodie-two-shoes "ethical" company, then I don't know what to tell you. This company has destroyed billions in shareholder wealth, particularly under Hector Ruiz (though Rory Read did a great job of taking the stock down from ~$8/sh to just $2 and change). This is more harmful than misstating some memory specs and showing off non-working prototypes at a trade show ever could be.

And if you think that such "shareholder wealth" destruction only impacts a few people, think again. AMD stock is/was held by a lot of mutual funds, employees of the company (who were compensated with stock), AMD retirees, etc. It was AMD management's job to try to protect and grow those people's investments and they failed big-time.

They say things like "Fury X is an overclockers dream" when it wasn't, try to dupe people into thinking Zen will have >40% IPC improvement over XV when they conveniently put that claim under the "data center" box in their slide (meaning the comparison is to Piledriver, their last gen server CPU), they say with a straight face to investors that Carrizo is a super great product and show that it's "better" than Broadwell by showing a corner case (HEVC decode), management rewarding themselves with random multi-million dollar bonuses when the stock hits new lows, etc.

This is just the tip of the iceberg, the more you dig, the more dirt you'll find on AMD.

Anyway, this is way OT so I'm going to shut up now, but if you wanna talk to me about "ethical" companies, AMD is the last example you really ought to use.

Oh wow. This is utterly fantastic, but I'm torn. It might be a bit too subtle and prone to poe's law especially since you mix in a legit criticism or two even if they're stretched to the breaking point, but something to make it clear that this is satire might be in order. Maybe saying that it's just like how Standard Oil was the only ethical company in its business during its time, but that might be a bit too on the nose. Considering people continually thought that semi-accurate story about the pascal sample was serious, I'd definitely aim to make it clear that this is in fact satire and that you aren't seriously stating that a company isn't ethical entirely because it has the temerity to not make a profit and totally gloss over any mention of things like anti-competitive practices.

 

[Silverforce11]

Yep, I work part-time as Bozo the Clown at the state fair during the summer. The kids love it when I show benchmarks that some people claim are not representative of real-world multi-GPU scaling -- it really gets the "lulz."

2 GPUs already have plenty of issues with scaling in games.

4 GPUs?

Very niche, and plenty of games do not show any scaling:

https://youtu.be/d8hKhlbrhQ4?t=1m29s

 

[Sweepr]

2 GPUs already have plenty of issues with scaling in games.

4 GPUs?

Very niche, and plenty of games do not show any scaling:

https://youtu.be/d8hKhlbrhQ4?t=1m29s

Well not scaling and negative scaling are different, and so far he failed to provide data to back up this nonsense.

Yep, I work part-time as Bozo the Clown at the state fair during the summer. The kids love it when I show benchmarks that some people claim are not representative of real-world multi-GPU scaling -- it really gets the "lulz."

LOL.

 

[Erenhardt]

Well not scaling and negative scaling are different, and so far he failed to provide data to back up this nonsense.

LOL.

 

[Sweepr]

If that's the best you could come up with I'm sorry. Mostly the same or better performance than 2-way SLI even in this handpicked tests. Don't be jealous.

 

[Erenhardt]

If that's the best you could come up with I'm sorry. Mostly the same or better performance than 2-way SLI even in this handpicked tests. Don't be jealous.

I will spare you all these jagged earthquakelike frametime graphs as it is not even remotely on topic. Cya!

 

[Silverforce11]

If that's the best you could come up with I'm sorry. Mostly the same or better performance than 2-way SLI even in this handpicked tests. Don't be jealous.

Dude, admit when you're wrong, be a man. You asked for examples of negative scaling and @Erenhardt provided it, 4 GPUs scaled backwards. -_-

Some people...

LOL @ 0 fps min ... some major stutter issues there. haha

 

[96Firebird]

I will spare you all these jagged earthquakelike frametime graphs as it is not even remotely on topic. Cya!

Huh, I would expect Titan X charts showing negative scaling, but those are missing... I guess when you can't provide evidence, you run! :thumbsup:

 

[Sweepr]

I will spare you all these jagged earthquakelike frametime graphs as it is not even remotely on topic. Cya!

I don't see any negative scaling here, do you? Sounds like you're making up stuff because you're bothered by the fact that he owns four of the best graphics cards available today, while most people can't afford one.

 

[Sweepr]

If that's not enough, more results: http://uk.hardware.info/reviews/603...-way-sli--4-way-sli-review-insane-performance

Zero negative scaling in any of 8 games tested @ 4K (up to 4-way Titan X SLI). Better than random tests using different VGAs with 4GB VRAM. Cya!

Edit: Minor regression in one game (Total War Attila), probably CPU limited.

 

[airfathaaaaa]

Huh, I would expect Titan X charts showing negative scaling, but those are missing... I guess when you can't provide evidence, you run! :thumbsup:

so when the rest of the cards just doesnt do the job we bring out the gun that isnt even a gaming card and we post results from commercial benches and call it a day?

 

[Jaydip]

If that's not enough, more results: http://uk.hardware.info/reviews/603...-way-sli--4-way-sli-review-insane-performance

Zero negative scaling in any of the 9 games tested (up to 4-way Titan X SLI). Better than random tests using different VGAs with 4GB VRAM. Cya!

To clarify yes using four cards can have negative scaling when your cpu isn't fast enough and we are not using graphically demanding options.While the second option is easy to fix the first not so much, I believe even Intel's fastest isn't fast enough to feed these four beasts.From the conclusion of the review you linked

" Especially for 4-way SLI, we noticed that performance will often times decrease rather than increase."

I believe some site cant recall now tested and concluded that 3 way was prolly the best way to go but that doesn't mean you can't grab one more