Wednesday, July 4th 2018
AMD Beats NVIDIA's Performance in the Battlefield V Closed Alpha
A report via PCGamesN points to some... interesting performance positioning when it comes to NVIDIA and AMD offerings. Battlefield V is being developed by DICE in collaboration with NVIDIA, but it seems there's some sand in the gears of performance improvements as of now. I say this because according to the report, AMD's RX 580 8 GB graphics card (the only red GPU to be tested) bests NVIDIA's GTX 1060 6GB... by quite a considerable margin at that.
The performance difference across both 1080p and 1440p scenarios (with Ultra settings) ranges in the 30% mark, and as has been usually the case, AMD's offerings are bettering NVIDIA's when a change of render - to DX12 - is made - AMD's cards teeter between consistency or worsening performance under DX 12, but NVIDIA's GTX 1060 consistently delivers worse performance levels. Perhaps we're witnessing some bits of AMD's old collaboration efforts with DICE? Still, It's too early to cry wolf right now - performance will only likely improve between now and the October 19th release date.
Source:
PCGamesN
The performance difference across both 1080p and 1440p scenarios (with Ultra settings) ranges in the 30% mark, and as has been usually the case, AMD's offerings are bettering NVIDIA's when a change of render - to DX12 - is made - AMD's cards teeter between consistency or worsening performance under DX 12, but NVIDIA's GTX 1060 consistently delivers worse performance levels. Perhaps we're witnessing some bits of AMD's old collaboration efforts with DICE? Still, It's too early to cry wolf right now - performance will only likely improve between now and the October 19th release date.
219 Comments on AMD Beats NVIDIA's Performance in the Battlefield V Closed Alpha
I'm still seeing new headlines quoting this article today and morons who don't read into the thread and are just going by the title and op.
www.guru3d.com/articles_pages/geforce_gtx_1050_3gb_review,12.html (BF1@DX12: 580 gets 106FPS vs 1060's 82FPS @1080P)
www.guru3d.com/articles_pages/geforce_gtx_1050_3gb_review,10.html (SF2@DX12: 94 vs 85)
www.guru3d.com/articles_pages/geforce_gtx_1050_3gb_review,11.html (ROTR@DX12: 81 vs 78)
www.guru3d.com/articles_pages/geforce_gtx_1050_3gb_review,13.html (DE MD@DX12: 71 vs 63)
www.guru3d.com/articles_pages/geforce_gtx_1050_3gb_review,14.html (W2C@Vulcan: 86 vs 66)
So, BFV being on the same engine as BF1 SHOULD play much better on 580 vs 1060. If it goes on sale and results show otherwise, it will be a proof that the collaboration with nVidia caused that difference. For anyone willing to debate on this, only arguments and facts on topic please, or else negligence will be my only answer.
www.purepc.pl/karty_graficzne/test_radeon_rx_580_vs_geforce_gtx_1060_9_gbps_msi_gaming_x?page=0,8
www.purepc.pl/karty_graficzne/test_radeon_rx_580_vs_geforce_gtx_1060_9_gbps_msi_gaming_x?page=0,7
but I guess it seems like a huge difference looking through your red glasses.
yes, very genarally speaking gcn is better in vulkan and dx12, but the difference usually comes down to a few percent. 1080 still manages to beat match V64 in BF1. This is stock 1080 FE (10gbps) vs AIB vega. AIB 1080 with 11gbps memory would pull ahead.
www.purepc.pl/karty_graficzne/test_asus_radeon_rx_vega_64_strix_gaming_oc_red_is_bad?page=0,7
www.purepc.pl/karty_graficzne/test_asus_radeon_rx_vega_64_strix_gaming_oc_red_is_bad?page=0,6
btw this thread was never about gcn vs pascal in dx12, why did you make it about that ? Discussing 580 vs 1060 performance is splitting hairs. whatever you can play on 580, you can play at he same level of smoothness and visual quality with 1060, and vice versa. 1060 does it with a smaller chip and less powerful hardware, cause on paper rx580 should be noticeably faster with a bigger die, 14nm process and 8 gigs of 256-bit memory.
There will always be some games where one architecture performs slightly better or worse, sometimes it comes down to small details in the shader design, and it doesn't mean one hardware architecture is better than the other. This is why good reviews rely on a wide selection of representative games, unlike some people who find the edge cases to prove their own agenda.
There is not anything in either Direct3D 12 or Vulkan which inherently benefits GCN, and the games we've seen so far doesn't even use Direct3D 12 natively without and abstraction layer, so judging architectures based on edge cases is just ridiculous.
You want a better comparison? factor in all of the DX12 games, games like Forza 6, Forza Horizon 3, Halo Wars 2, Gears 4, Gears Ultimate, Civ 6 .. etc. GTX 1060 is on top in all of them.
I typically try to stay away from maymays on TPU but I can't contain myself this time.
Worst case scenario for D3D12 is that it uses D3D11 calls so even though it's technically D3D12, it performs like D3D11. A lot early D3D12 games did this--especially games that released as D3D11 and were updated to support D3D12 (The Division comes to mind). Software that's developed for D3D12 from the ground up will see a significant improvement in performance and reduced CPU load across GCN-based cards and Pascal-based (because of async compute) cards. Maxwell and Kepler cards see a minor improvement (mostly because of reduced CPU load).The aggregate is RX 580 is a little bit faster with 2 GiB extra VRAM. Both cards will drive 1920x1080 at 60 fps fine.
www.purepc.pl/karty_graficzne/radeon_rx_480_vs_geforce_gtx_970_test_na_kilku_procesorach?page=0,10
in rotr, even in dx12 mode with 4790k, rx480 cannot performa as well as it should.
www.purepc.pl/karty_graficzne/radeon_rx_480_vs_geforce_gtx_970_test_na_kilku_procesorach?page=0,13
The more I look at the results the more I think there has to be some sort of flaw in design or driver support for gcn cards that dx12 can thoeretically help, but not always does cause it's not able to. They can look great tested on i7's and i9's, but the story about dx12 being able to carry those compute monsters on slow cpus is not true in many cases.
www.sweclockers.com/test/25948-snabbtest-battlefield-v-closed-alpha-sju-grafikkort-i-snoig-batalj
good grief....
The game is a mere 2-3 months from release , built on the same engine as the previous iteration with no obvious big changes. Whoever seriously believes the end product is going to come with major shifts in terms of performance from any of the two vendors is deluding himself.Is it Nvidia mentality to give a damn ?
I don't get you , if you feel like you have provided us with irrefutable proof for whatever the hell is it that you are trying to prove why are you still bothered by that ? You are borderline trolling to say the least.
the game looks amazing,even in alpha. Will probably be my first BF game
It's pointless to simulate the old API through an abstraction layer. Sometimes these abstractions can perform worse than the old API, because the new APIs are designed around a different approach. The point of the new APIs were to leverage lower level control over the hardware, and re-adding wrappers to abstract that away defeats the whole purpose. Utilizing these APIs properly requires entire new engines built from the ground up to leverage this new level of control, and since this is painstakingly hard, we probably wouldn't see any widespread proper adaption of these APIs anytime soon. Most developers will continue to use abstractions to deal with the new APIs, and might even continue to do so. Games are unfortunately mass-produced trash these days, not only in terms of recycling the same concepts, but also in terms of code. They are usually stitched together before the shipping date, and then they move on to the next title.
Give it up man you’re really trolling at this point and it’s sad. Your NVIDIA mentality is rather pathetic at this point.