guru3dStar Wars: Battlefront Beta VGA graphics performance benchmarks

[crisium]

Something is terribly wrong if your GTX 970 only gets 45FPS at 1080 50%. Isn't that comparable to 1360x768?

 

[Techhog]

Something is wrong I'm playing at 1080p 150% on my r9 290 and I love it! I was going to see if I can go higher but then Ill just buy a second 290.

Edit: I don't even know how you would think that would be normal...

Something is terribly wrong if your GTX 970 only gets 45FPS at 1080 50%. Isn't that comparable to 1360x768?

Sorry. I meant 150%.

 

[AtenRa]

Sorry. I meant 150%.

How is it performing at %100 ???

 

[Techhog]

How is it performing at %100 ???

I've been using 125%. It stays around 60FPS with highs in the 70s (I think the highest I saw was 77) and a minimum of 57.

 

[RussianSensation]

Angry Joe Impressions

Joker - 5 positive things about SW:BF, 10 negative things.

Ars Technica - Star Wars Battlefront beta provides a delightfully unbalanced battle of Hoth

This game's lore/universe would have been perfect for an amazing single-player campaign. We haven't had a solid FPS game with a single player for a while.

 

[AtenRa]

System

Core i7 3770K @ 4.44GHz
2x 4GB DDR-3 1600MHz 9-9-9
HD7950 @ 1GHz

Win 8.1 64bit
Catalyst 15.9.1BETA

Map : Hoth
Graphics : High

 

[happy medium]

A GeForce GTX 780 Ti with its 3 GB of graphics memory normally should run into issues, but that's not the case in Ultra HD. So the game engine adapts with caches, and we think caches a little less to compensate.

 

[Madpacket]

I tried it out on my 2GB 960 today. Game ran pretty smooth at 1080P and looks great on the HTPC TV. Left everything at default settings but I'll do some performance tests later. I don't think it's much more demanding than BF4 which makes sense.

Too bad the game itself is pretty bland. I'll probably pick it up for 10 bucks next year if there's still a community playing it but with how consolized it is I doubt there's going to be one for long.

 

[DeathReborn]

Angry Joe Impressions

Joker - 5 positive things about SW:BF, 10 negative things.

Ars Technica - Star Wars Battlefront beta provides a delightfully unbalanced battle of Hoth

This game's lore/universe would have been perfect for an amazing single-player campaign. We haven't had a solid FPS game with a single player for a while.

Maybe we will see a Force Unleashed 3 sometime, hopefully...

 

[psolord]

Hello. I finally made the game work in my systems.

I had a very weird black screen at start up and it stayed there. The impossible solution was to install origin on top of my current installation and then the game worked fine.

My systems are all upgraded from Win7 to Win10 so maybe something needed a refresh.

Anyhoo, here are my benchmarks with the accompanying videos, in case anyone is interested.

Tatooine single player survival. Not a big fan of MP plus I find SP easier to benchmark, although the game is quite random and dynamic so you cannot do exactly the same things, as I do in my benchmarks.

Still the game seems to be very well balanced witth very little framerate variations.

(60fps videos recorded with an external recorder-spicy wallpapers alert )

Star Wars Battefront 1920x1080 Ultra GTX 970 @1.5Ghz Core i5 2500k @4.8GHz -92fps

Star Wars Battlefront 1920x1080 Ultra 7950 @1.1Ghz CORE i7-860 @4GHz - 62fps

Star Wars Battlefront 1920X1080 High 5850 @950Mhz Q9550 @4GHz ~33fps

Some interesting points

- Although I used the same settings for both the 970 and the 7950, the first showed 2.5GBs vram usage while the second 1.9GBs.

- I read somewhere that 8GBs ram could make the game less fluent. The 7950 system which has 8GBs showed no such behavior even when there was a lot of action going on. MSI Afterburner registered only 4GBs of overall system ram usage anyway. Maybe things will be different in MP, dunno.

- Once more the humble 5850 surprises me. It handled the high preset with its measly 1GB just fine. Actually it showed signs that it could handle Ultra as well but it crashed.

 

[gamervivek]

Interesting that the clockspeed difference might explain the advantage that 970 has.

How about the vram usage of amd and nvidia cards in other games?

 

[Azix]

People are saying image quality is lower on nvidia. Anyone noticed this?

 

[Osjur]

People are saying image quality is lower on nvidia. Anyone noticed this?

Not another IQ-Gate... please

 

[MagickMan]

People are saying image quality is lower on nvidia. Anyone noticed this?

Nope, looks about the same on both my 980Ti rig and Fury rig, w/ the 980Ti (1522 core, 8000 mem) having ~20% higher FPS @1440p Ultra.

 

[gamervivek]

The start here,

http://www.overclock.net/t/1576182/guru3d-starwars-battlefront-beta-benchmark/370#post_24494363

The screenshot like the previous IQ-gate one in BF4

http://cdn.overclock.net/1/14/14ee8628_FXvsTi.png

 

[Silverforce11]

It's probably linked to the same bug that Greg had with his setup, and its likely due to him swapping hardware frequently. Probably a driver setting conflict somewhere that defaults to lower quality filtering.

 

[Erenhardt]

Or quality degradation by shadowplay recording? Were both videos captured using the same software?

https://youtu.be/3MGLQWlXo0Q?t=90

But that memory usage difference... AMD wasn't lying about HBM needing less capacity thanks to higher bandwidth.

 

[Flapdrol1337]

But that memory usage difference... AMD wasn't lying about HBM needing less capacity thanks to higher bandwidth.

Modern games will use what they can. If there's less memory those games will just use less.

4GB is enough, has nothing to do with bandwidth.

The only thruth in it is there's a minimum capacity high speed gddr5 chip, so if you want lots and lots of bandwidth you have to use many chips, so you'd end up with much more.

 

[Erenhardt]

Modern games will use what they can. If there's less memory those games will just use less.

One is using 2 GB, the other 3GB, both have much more than that.

 

[gamervivek]

That reviewer also has a video comparing different nvidia cards where Titan X shows same fire quality as Fury X does in that Fury X vs. 980Ti video.

https://www.youtube.com/watch?v=PYfAcF10QC8&t=2m48s

Not sure what is going on, especially with the rocks in that comparison pic where lightning/shadows seem to be absent on 980Ti.

 

[guskline]

Just ran the single player mode through 5 waves using FRAPS to monitor in the 5960x/GTX980TI SC rig below. All settings at ULTRA and resolution 2560 x1440. Min 60fps; Max 202 and Average 86.9 fps. That's consist with the Guru3d article using the same cpu and clock speed but a reference GTX980TI. Mine is clocked at 1102. I'll run the game on my CF 290s with a 4790k to give some comparison.

In the 4790k/290 CF rig below with all settings at ULTRA and at 2560 x 1440 the Min was 25; Max 202 and Average 69.57 fps. My 4790k is at 4.7Ghz. I'm not sure if the Cross Fire profile is working since I would think the 2 R9 290s would be slightly faster than the single GTX980TI.

 

[ShintaiDK]

But that memory usage difference... AMD wasn't lying about HBM needing less capacity thanks to higher bandwidth.

You mean the 2 engineers sitting dedicated only to handle memory management? Nothing to do with HBM.

http://www.anandtech.com/show/9390/the-amd-radeon-r9-fury-x-review/7

Which is why for Fiji, AMD tells us they have dedicated two engineers to the task of VRAM optimizations. To be clear here, there’s little AMD can to do reduce VRAM consumption, but what they can do is better manage what resources are placed in VRAM and what resources are paged out to system RAM.

 

[psolord]

Interesting that the clockspeed difference might explain the advantage that 970 has.

How about the vram usage of amd and nvidia cards in other games?

Usually my 970@1.5Ghz is faster by about 50-60% than my 7950@1.1Ghz. The clock difference is 36% so the rest must be from architectural improvements of maxwell.

Actually the clock speed itself is an architectural improvement or just an architectural difference, since Nvidia is using quite higher clock speeds for a while now.

Still we are talking about a Tahiti chip. I am sure a Hawai chip at 1.1Ghz would be just as fast as a small Maxwell.

Now regarding the vram, the 970 does have the tendency to use a bit more vram. Always according to MSI Afterburner.

Taking data from my latest benchmark videos, in the Infiltrator demo, the 7950 used 1.5GBs, the 970 1.9GBs.

In mad max the 7950 used 900MBs, the 970 1.25GBs.

In rocket league, 7950 used 500MBs, the 970 730MBs.

In battlefield hardline, 7950 used 1.7GBs, the 970 2GBs.

In ashes of singularity DX11 7950 used 1.25GBs, the 970 2.5GBs (dafuq didn't notice that before)

 

[Tweak155]

Played this 2 days ago on my 970 stock @ 2560x1440... didn't monitor FPS but I didn't notice any slow downs. I was impressed by how the game looked though, I've been out more modern PC games for a while.

 

[Silverforce11]

@psolord

There's really no point comparing a 970 vs a 7950.

The 7950 was a 670 competitor and generally below it back then. The surprising thing from your results is that a 7950 actually gets above 60 fps at 1080p on Ultra settings.