AlienBabelTech Forums
Nvidia Lagging Behind AMD In Low Overhead APIs? - Printable Version

+- AlienBabelTech Forums (http://alienbabeltech.com/forum)
+-- Forum: Technology (http://alienbabeltech.com/forum/forumdisplay.php?fid=6)
+--- Forum: Video (http://alienbabeltech.com/forum/forumdisplay.php?fid=7)
+--- Thread: Nvidia Lagging Behind AMD In Low Overhead APIs? (/showthread.php?tid=1327)



Nvidia Lagging Behind AMD In Low Overhead APIs? - SteelCrysis - 07-12-2016

http://www.extremetech.com/gaming/231527-new-doom-update-adds-vulkan-support-amd-claims-substantial-performance-boost
Quote:The RX 480 is just one GPU, and we’ve already discussed how different cards can see very different levels of performance improvement depending on the game in question — the R9 Nano picks up 12% additional performance from enabling versus disabling async compute in Ashes of the Singularity, whereas the RX 480 only sees a 3% performance uplift from the same feature. Nvidia has not released a statement indicating whether or not the company expects to see a performance uplift from using Vulkan or which GPUs it expects to see the most benefit but testing by Guru3D suggests that the 1070 is completely flat in Vulkan vs. OpenGL. Guru3D’s testing also shows that the RX 480’s performance improvements are in-line with AMD’s claims.

One of the most significant differences between Nvidia and AMD over the past 12 months has been their approach to next-generation APIs. AMD has prominently talked about DirectX 12 and Vulkan at every opportunity. That’s partly because both APIs are closely related to AMD’s Mantle, which debuted a bit less than three years’ ago, but it’s also because DX12 has often improved AMD’s GPU performance relative to Nvidia. Team Green, in contrast, has had comparatively little to say about the API shift. Pascal’s vastly improved preemption capabilities obviated the performance hit that Maxwell took in some asynchronous compute scenarios, but Nvidia isn’t really putting much of a PR push behind the API shift. Visit the DirectX 12 section on GeForce.com and there’s just one title listed there, and nothing about games like Hitman, Rise of the Tomb Raider, Ashes of the Singularity, or Doom itself.

We’re not implying that Nvidia isn’t committed to DirectX 12, but the company’s public messaging on low-overhead APIs has been anemic these past 11 months. Pascal’s excellent performance has given Nvidia the overall lead at the top of the market, but there are millions of 28nm GPUs from Teams Red and Green still in the field. 28nm GPU performance in DX12 is going to be a topic that people care about for a long time to come, and AMD’s overall messaging on that topic has been stronger than Nvidia’s thus far.



RE: Nvidia Lagging Behind AMD In Low Overhead APIs? - ocre - 07-13-2016

Woah...lame...real lame

The issue has been AMDs all along. They have been using way more beefed up designs to battle nvidia, who has mastered the ability to extract an insane amount of performance from chips that on paper liik much more anemic. Nvidia gets much more real world performance per rated gflop. They were already killing it when it comes to utilization. Nvidia already dealt with and solved overhead, extracting way more performance from weak CPUs. These things were AMD struggles, their issues. AMD invents mantle because they had issues with extracting performance to the full potential. They saw how consoles were doing it and tried to copy and bring this to PC.

But much of what AMD was desperately needing, Nvidia had solved. They already get great utilization, the new APIs aren't much of a help.
It's just how its is. Like Async, amd needs async to achieve more utilization. There isn't a lot of gaps for Nvidia in the pipeline, async does very little when there is little to gain


RE: Nvidia Lagging Behind AMD In Low Overhead APIs? - SteelCrysis - 07-13-2016

More testing: https://www.techpowerup.com/224095/doom-with-vulkan-renderer-significantly-faster-on-amd-gpus
[Image: 45a.jpg]


RE: Nvidia Lagging Behind AMD In Low Overhead APIs? - SickBeast - 07-13-2016

Wow. I wonder if nVidia can solve this with driver optimization. It looks like there is something wrong. Vulkan is having absolultely no effect on nVidia cards whereas it's boosting the AMD cards 30%+.


RE: Nvidia Lagging Behind AMD In Low Overhead APIs? - ocre - 07-14-2016

So, why did they use TSAA? Very strange.

Anyway, I still firmly believe AMD has more to gain in these newer APIs. It's also on the backs of developers..very few will go to great lengths to focus a lot of extra resources for a small portion of the market.

AMD will have to get behind these developers, this is what they will do. They will try to get a few big games, then push a huge campaign like they do with ATOS.

It's just this offs graph is very strange. Why did they use tsaa, I suspect this is an advantage for AMD in vulkan. Its obscured

Any how, this is all being blown up like all the hype AMD pays for. Remember, think back. Battlefield and mantles first showcase.
Nvidia is not gonna hand over the secret sauces to developers or anyone. But, they are masters at what they do. Any project they feel like working on, they get results. They completely stopped mantle with battlefield, nvidia put the effort and for the result. So, it's not some magical thing that nvidia is helplessly dead from. Things will be more the same. The titles AMD invest in will have good results, the ones Nvidia decided to...They will have great results. The rest will fall in the place they have been


RE: Nvidia Lagging Behind AMD In Low Overhead APIs? - SickBeast - 07-14-2016

apoppin found some other benchmarks where the nVidia cards saw similar gains to the AMD cards.  This looks like a driver glitch or some type of testing error.

*edit* Actually no, he didn't! The nVidia cards are showing very small gains with Vulkan. :(


RE: Nvidia Lagging Behind AMD In Low Overhead APIs? - ocre - 07-14-2016

Vulkan is the API closest to mantle. It's so close that it looks so similar you could be fooled by just looking at the code.

Not sure what nvidia would do but we do have examples of mantle in the past, nvidia was able to match and surpass what amd could achieve. So, I dont think they will be helpless.

But still, I am pretty sure that amd has a lot more to gain, it's their hardware that prompted them to come up with mantle. Nvidia will have to let go some of their secrets or go and work closely with the games that they want the best results from.

We are talking about completely different designs, architectures that work completely different, and rhe results of this API which was based on mantle...which, it is based on GCN.

It shouldn't be a surprise that GCN gets a bigger boost, vulkan is very very close to mantle

Nvidia will have to let go of their close secrets to have a low level API take advantage of their architectures the same way. I am not sure they are tasty to do that yet.


RE: Nvidia Lagging Behind AMD In Low Overhead APIs? - SickBeast - 07-18-2016

(07-17-2016, 12:55 AM)24601 Wrote:
(07-14-2016, 06:23 AM)ocre Wrote: Vulkan is the API closest to mantle.  It's so close that it looks so similar you could be fooled by just looking at the code.

Not sure what nvidia would do but we do have examples of mantle in the past, nvidia was able to match and surpass what amd could achieve.  So, I dont think they will be helpless.  

But still, I am pretty sure that amd has a lot more to gain, it's their hardware that prompted them to come up with mantle.  Nvidia will have to let go some of their secrets or go and work closely with the games that they want the best results from.

We are talking about completely different designs, architectures that work completely different, and rhe results of this API which was based on mantle...which, it is based on GCN.

It shouldn't be a surprise that GCN gets a bigger boost, vulkan is very very close to mantle

Nvidia will have to let go of their close secrets to have a low level API take advantage of their architectures the same way.  I am not sure they are tasty to do that yet.

People seem to think AMD have the money to buy developers forever.

They forget that they have just ceded all of the 14/16nm generation to Nvidia.

When Zen fails to live up to it's hype, it's cutting time again.

Also I'm permabanned from ATF again.

The paid shills just can't handle my superiority to them, as always.

BTW, finally deleted Anandtech Reviews from my RSS feed. Holding 1080 and 1070 reviews hostage until AMD releases is pure cancer. And they don't do any decent reviews of anything other than videocards anyways.

Welcome to ABT! It feels good being able to speak freely, doesn't it? Angel