[pcgameshardware.de]Star Wars Battlefront 2 Beta Benchmarks

[kondziowy]

RX480, R9 Nano and Fury X disagree with you, as they gain performance in 1440p.

 

[Muhammed]

RX480, R9 Nano and Fury X disagree with you, as they gain performance in 1440p.

Yeah like literally 2fps, but lose a heck a lot of frame pacing "As per PCGH testing" so no, they effectively gain nothing.

 

[Krteq]

Test of release version

CB.de - Star Wars Battlefront 2: AMD liegt in der Elite Trooper Deluxe Edition vor Nvidia

 

[crisium]

Looking at the computerbase benchmarks, it appears that this game follows TFLOPs pretty linearly - much more than other games. Bandwidth and the number of ROPs don't seem to matter. 570 > 390 by a bit despite half the ROPs, less bandwidth. 1070 Ti barely behind a 1080, and miles ahead of the 1070.

 

[Flapdrol1337]

Yeah, 1070Ti being identical to a 1080 is surprising.

Too bad the game is a pay2win piece of garbage.

 

[Shmee]

Too bad they launched it this way so far. Well, it made me buy the classic swbf2 on steam for $10

 

[Despoiler]

The game runs great on my Fury with everything cranked. DX11. DX12 is a nightmare. Stalls all over the place. The game is mind blowing gorgeous. The textile texture on the uniforms is insane. I've never seen better looking cloth materials. Faces are super detailed. The crafting of the levels and world is very meticulous.

The game itself is good IMO. The pay2win is largely over blown. Most people are so terrible that level 4 cards can't even help them. The campaign is 4 hours long. Some of the worst writing I've ever had to endure. The story was a lot of "well isn't that convenient to move the story along." The gameplay was decent.

 

[PhonakV30]

Test of release version

CB.de - Star Wars Battlefront 2: AMD liegt in der Elite Trooper Deluxe Edition vor Nvidia

WOW , 1070TI is pretty damn close to 1080!

 

[dogen1]

WOW , 1070TI is pretty damn close to 1080!

Yeah, the engine must be super bandwidth efficient.

 

[itsmydamnation]

Yeah, the engine must be super bandwidth efficient.

We are heading down the path larrabee tried to push a decade ago. It was the obvious path but we just didn't have the transistor budget then. Its hard to scale "core" counts and keep linear performance, it makes sense to increase functionality and flexibility of those cores and reduce the fixed pipeline paths that while very fast at what they do create inefficiencies in the pipeline.