AMD?

[Dackzy]

As it stands right now they are not lying 

http://www.ashesofthesingularity.com/metaverse#/personas/b0db0294-8cab-4399-8815-f956a670b68f/match-details/6b9e8c43-df70-4cec-8ddb-b209c0316fa8

http://www.ashesofthesingularity.com/metaverse#/personas/b0db0294-8cab-4399-8815-f956a670b68f/match-details/ac88258f-4541-408e-8234-f9e96febe303

We can't say anything about other games and AMD has not said anything about other games.

[Dzzope]

Co-operation in some areas is not a problem... Look at to use your example Apple and Samsun.. Apple use Samsung screens.

 

Co-operation can be good to set standards and make certain functions or ways of doing things the same so that everyone can use them.. Eg DX makes a standard.. They could come up with a "standard" for the free/gsync techs.. But I don't think that wil happen unless Nvidia decides it has enough control without supplying the hardware.

[Dackzy]

10 minutes ago, LiquidPanic said:

I think you people are all missing something big here. 2 RX 480s in CrossFire were matching performance with a GTX 1080, while only at 50% utilization. If you translate that to not be confusing babble, they were literally saying that the RX 480 is just as fast as the GTX 1080, which is not going to happen under any scenario. Also, if you watch the demo closely, you can clearly see that the RX 480s in CF were running the game at lower settings than the 1080.

The 1080 could not proces the snow, because bad support for dx12 and async compute, so it was the 1080 that was running it on easy mode, because the snow are way harder to proces.

[Mr. Tsao]

You have to admit that it's kinda fishy that AMD pushed out benchmarks when they knew that the Nvidia card was not functioning properly. Before you reply and blast my ass, keep in mind that I'm not a fanboy of either company

[Prysin]

13 minutes ago, LiquidPanic said:

I think you people are all missing something big here. 2 RX 480s in CrossFire were matching performance with a GTX 1080, while only at 50% utilization. If you translate that to not be confusing babble, they were literally saying that the RX 480 is just as fast as the GTX 1080, which is not going to happen under any scenario. Also, if you watch the demo closely, you can clearly see that the RX 480s in CF were running the game at lower settings than the 1080.

AMD had the correct settings, Nvidia didnt have proper shaders according to OXIDE games. The studio who MADE the game. GPU utilization was 51% under normal batches because AMDs drivers sucks and the game is insanely CPU intensive.

 

i know this, cuz i use Ashes to benchmark my own R9 295x2. WHich is 2x 290X which will in theory be ~ 2x RX 480....

 

[Dackzy]

1 minute ago, Mr. Tsao said:

You have to admit that it's kinda fishy that AMD pushed out benchmarks when they knew that the Nvidia card was not functioning properly. Before you reply and blast my ass, keep in mind that I'm not a fanboy of either company

So does Nvidia, with their hairworks thing. Nvidia will not get good support for DX12 and Async compute this gen, so AMD is just using their flaws against them, like Nvidia does against them.

[LiquidPanic]

11 minutes ago, HKZeroFive said:

Game was never running at lower settings buddy. Starting to sound like JayzTwoCents:

 

https://www.reddit.com/r/Amd/comments/4m692q/concerning_the_aots_image_quality_controversy/

So you're actually going to buy AMD's marketing and really believe that the RX 480, a $200 GPU, is going to match performance with the GTX 1080? Because that's literally exactly what they're saying.

[Mr. Tsao]

3 minutes ago, Dackzy said:

So does Nvidia, with their hairworks thing. Nvidia will not get good support for DX12 and Async compute this gen, so AMD is just using their flaws against them, like Nvidia does against them.

I agree that nvidia also takes advantage of these things, but you also have to account for the fact that third-party reviews have already come out for the 1080, and based off of those, a 480 (390) crossfire shouldn't beat a 1080 in games that don't have limitations

[Dackzy]

1 minute ago, LiquidPanic said:

So you're actually going to buy AMD's marketing and really believe that the RX 480, a $200 GPU, is going to match performance with the GTX 1080? Because that's literally exactly what they're saying.

They are saying that two 200$ or 229$ GPUs will performe the same as one 1080 in DX12 games.

[LiquidPanic]

Just now, Dackzy said:

They are saying that two 200$ or 229$ GPUs will performe the same as one 1080 in DX12 games.

Not at all, because during their demo they showed 2 RX 480s, in crossfire, at only 50% utilization (as in essentially only a single GPU) matching the GTX 1080, which was apparently at ~98% utilization.

[Mr. Tsao]

2 minutes ago, Dackzy said:

They are saying that two 200$ or 229$ GPUs will performe the same as one 1080 in DX12 games.

But they also said that the cf setup was only operating at 50% usage. So theoretically, according to their statements, one card at 100% would operate the same

[Dackzy]

1 minute ago, Mr. Tsao said:

I agree that nvidia also takes advantage of these things, but you also have to account for the fact that third-party reviews have already come out for the 1080, and based off of those, a 480 (390) crossfire shouldn't beat a 1080 in games that don't have limitations

Well a lot of the reviews of the 1080 have been done by people who are kinda Nvidia fanboys, but hides it really well and the 1080 is not that powerful card a OC 980ti is about 8-10% slower than the 1080 and the 480 is more a 980/fury in performance according to the leaks we have seen. The 1080 is also a bad OC card you can get that 2000MHz speed, but you don't get that many extra frames by doing it.

[HKZeroFive]

4 minutes ago, LiquidPanic said:

So you're actually going to buy AMD's marketing and really believe that the RX 480, a $200 GPU, is going to match performance with the GTX 1080? Because that's literally exactly what they're saying.

That's not what I'm saying...

 

If you have two RX 480s in Crossfire, in a game that favours AMD, it's not implausible that it would match GTX 1080 performance.

[Dackzy]

3 minutes ago, LiquidPanic said:

Not at all, because during their demo they showed 2 RX 480s, in crossfire, at only 50% utilization (as in essentially only a single GPU) matching the GTX 1080, which was apparently at ~98% utilization.

 

1 minute ago, Mr. Tsao said:

But they also said that the cf setup was only operating at 50% usage. So theoretically, according to their statements, one card at 100% would operate the same

Look at what Prysin said.

[Simon771]

I don't care what anyone says, just watch this video here and start at 8min 22sec.

 

The right one is GTX 1080, and the left one is 2x RX 480 in crossfire.

Does it seems like it can compare to it? I'm sorry but the right one looks much more detailed, smoother, and in general just better.

[Prysin]

8 minutes ago, LiquidPanic said:

So you're actually going to buy AMD's marketing and really believe that the RX 480, a $200 GPU, is going to match performance with the GTX 1080? Because that's literally exactly what they're saying.

they are claiming TWO. ONE RX 480 + ONE RX 480 = 1080

 

[Mr. Tsao]

1 minute ago, Dackzy said:

Well a lot of the reviews of the 1080 have been done by people who are kinda Nvidia fanboys, but hides it really well and the 1080 is not that powerful card a OC 980ti is about 8-10% slower than the 1080 and the 480 is more a 980/fury in performance according to the leaks we have seen. The 1080 is also a bad OC card you can get that 2000MHz speed, but you don't get that many extra frames by doing it.

First of all, I don't think you can consider any tech reviewer a Nvidia fanboy just because they use their card. They generally just go with the solution what works the best (mainly because nvidia gimps the crap out of AMD cards). But second of all, about the 1089 being a bad overclocked, that's only with the reference model because the power draw is limited by the single 8 pin connector

[DXMember]

27 minutes ago, LukaP said:

Thats like saying it would be nice if Samsung and Apple cooperated. Not going to happen. Not in the companies interest. And even if it were, there are laws against monopolies that cooperation would break...

fun fact... Samsung makes Apples phones and tables and what not

[Prysin]

2 minutes ago, Simon771 said:

I don't care what anyone says, just watch this video here and start at 8min 22sec.

 

The right one is GTX 1080, and the left one is 2x RX 480 in crossfire.

Does it seems like it can compare to it? I'm sorry but the right one looks much more detailed, smoother, and in general just better.

again, Nvidia is using the wrong shaders according to OXIDE Games. The studio who MADE Ashes of the Singularity. There is no better source then the creator of a game. They have all the info and ALL the telemetry.

[Mr. Tsao]

2 minutes ago, Prysin said:

they are claiming TWO. ONE RX 480 + ONE RX 480 = 1080

 

They're saying two cards running at half power=one 1080, essentially saying that that setup running at full power is twice the performance of a 1080 for half the price. That doesn't seem right

[Simon771]

Oh and another thing from that event video

 

 

So if GPU utilisation is at 51% when at crossfire ... why didn't they just use single RX 480 instead and it would have same performance.

Or on another note ... why only 51% utilization? Why it's not using 100% of the GPU and delivering more FPS?

Are they dumb and using Vsync maybe?

However I look at that video, everything seems just wird.

[LiquidPanic]

3 minutes ago, Prysin said:

they are claiming TWO. ONE RX 480 + ONE RX 480 = 1080

 

Actually read my reasoning before you say anything, you clearly don't understand what I was talking about.

[Prysin]

2 minutes ago, DXMember said:

fun fact... Samsung makes Apples phones and tables and what not

not quite...

 

Samsung makes the Memory, CPU/SOC and Screen. Wifi is often Qualcomm or Mediatek i think. The various sensors are apple patents.

[Kinda Bottlenecked]

15 minutes ago, Mr. Tsao said:

You have to admit that it's kinda fishy that AMD pushed out benchmarks when they knew that the Nvidia card was not functioning properly. Before you reply and blast my ass, keep in mind that I'm not a fanboy of either company

 

And you immediately label LukaP as a fanboy for not agreeing with you? 

 

9 minutes ago, LiquidPanic said:

Not at all, because during their demo they showed 2 RX 480s, in crossfire, at only 50% utilization (as in essentially only a single GPU) matching the GTX 1080, which was apparently at ~98% utilization.

 

What's your original account? 

[Dackzy]

1 minute ago, Mr. Tsao said:

First of all, I don't think you can consider any tech reviewer a Nvidia fanboy just because they use their card. They generally just go with the solution what works the best (mainly because nvidia gimps the crap out of AMD cards). But second of all, about the 1089 being a bad overclocked, that's only with the reference model because the power draw is limited by the single 8 pin connector

Nope that OC thing is with all of the models, it is hard locked to 1.25v and you get little to no improment per MHz OC, the 980ti has higher FPS/MHz OC than the 1080. Oh I can call them fanboys, have you not seen the rage jayztwocents made about the 480 and 1080 and a lot of other youtubers has basically done the same.