Battlefield V: GPU Benchmark, DX12 a Stuttery Mess [Beta]
Battlefield V: GPU Benchmark, DX12 a Stuttery Mess [Beta]
2018-09-05
welcome back to harbor unboxed and
firstly yes my voice is a bit off today
my kids have been sharing at the Germans
with me again I've been sick for the
past a week I haven't had a voice pretty
much for the last 3 days
today it has improved a lot so I can
actually talk now so I'm wasting what
little voice I have on this video but
it's well worth it
because yesterday I got my first look at
battlefield 5 and I spent all night and
what I had of yesterday benchmarking the
game so that I could show you guys the
performance so I've spent a lot more
time benchmarking that I have actually
enjoying the gameplay but that probably
won't surprise to many of you anyway
this one it is a better version of the
game so I suppose we should just call
this more of a preview of the
battlefield 5 performance the game will
be coming out in two months time I don't
expect any major changes to performance
probably if anything we'll just see some
driver optimizations from AMD and NVIDIA
the purpose of the open beta is to test
technical aspects like service stability
latency levels and then of course the
matchmaking to see how well that works
and then things like gameplay elements
you know such as weapon balancing and so
on those of you that pre-ordered the
game and I know we highly recommend that
you never pre-order things at least
until checking out reviews but the point
of us pre-ordering it is so that we can
tell you how well the game does or
doesn't play on your hardware anyway
those of you that did pre-order the game
and now have early access to the beta
while those that haven't will be able to
get access in two days time though by
the time you watch this video it's
probably more like one day I'm not gonna
waste time explaining what battlefield 5
is I'll provide any kind of introduction
I'm just going to assume if you're
watching this video you know what it's
all about and I'm also gonna try and
save my voice as much as I can I will
just say that like 2016 battlefield 1
and 2013 battlefield 4 yeah the name is
kind of confusing on that one
battlefield 5 uses the same 3rd
iteration of the frostbite engine
Kalas other quality games have also made
use of the game engine over the
battlefield 5 will be hitting multiple
platforms including PlayStation Xbox and
Windows the PC version of beta allows
gamers to adjust the quality levels of
things such as textures a filtering
lighting effects post-processing mesh
terrain undergrowth anti-aliasing and
ambient occlusion the beta also includes
options for resolution scaling UI
scaling field of view sliders motion
blur lens distortion vignettes and more
as you might expect there are limited
maps and game modes available in the
open beta but doesn't really matter for
testing as I'm only using a single map
anyway getting into the game was quite
difficult as all the servers were
constantly full but I did manage after a
fair bit of messing around to get in to
where I needed to be to run my tests I
chose the Grand Operations mode on the
Narvik map and I had to keep loading
into the game until I ended up on the
right team in the right section of the
match so that I could run my pass it was
a massive amount of messing about but I
managed to test almost a dozen of the
latest generation GPUs at 3 resolutions
I also have some quality preset scaling
data to go over as well towards the end
of the video as usual I'll be using my
GPU test rig which is built inside the
corsair crystal 570 X and comprises of a
core i7 87 okay clocked at 5 gigahertz
with 16 gigabytes of ddr4 3400 memory
for the GeForce GPUs the three-nine 9.0
7wh ql driver was used and frame d the
Radeon adrenaline 18.82 driver was used
okay so I think that's about everything
let's get into the results before
getting too deep into the testing I
decided to check out how the DirectX 11
and DirectX 12 api's compare in this new
title after all if AMD worked best with
DX 12 and video with DX 11 that's how I
would have tested the GPUs from each
company going forward however I quickly
discovered that DirectX 11 was the only
way to go for both teams are using
DirectX 12 the game suffered a major
stuttering issues with both the AMD and
NVIDIA GPUs and this is something we saw
in battlefield 1 upon release so I guess
it's not entirely surprising
developer a a dice clearly out
prioritizing the low-level API so for
the remainder of this video we will be
testing exclusively with DirectX 11 ok
so first up we have the 1080p results
using the Ultra
deep reset and here the rx 580 was up to
14% fast in the GTX 1066 year light the
Radeon graphics card also maintained a
frame rate above 60 FPS on our tests and
we see that Vega 56 was also 8% fast in
the GTX 1074 the average frame rate but
did suffer a larger disparity between
the average and 1% low result making it
5 percent slower for frame time
performance this weaker than expected
frame time performance was also seen for
both the Vegas 64 models and both were
down on the gtx 1070 TI despite beating
the 1080 for the average frame rate so
some driver works still need to come
from AMD but no surprises there as we
are only testing the beta version of the
game and of course it is an Nvidia
sponsored title the gtx 980ti also sees
a rather large difference between the
frame time and average frame rate result
that said we're likely running into some
kind of system bottleneck here is the 1%
low figure still broke 100 FPS moving to
1440p and we see that the gap between
the 1% lower average frame rate for the
1080i has shrunk and this is likely due
to the fact that we're more GPU bound
here are taking the pressure off
whatever the performance of limiting
component was previously perhaps the CPU
or memory for example again we see that
the Vegas series is a little lacking
when it comes to frame time performance
which is a shame as we're seeing
competitive average framerate results
hopefully AMD can address this before
the game's official release at 1440p the
RX 580 and GTX 1060 are able to deliver
play performance though the RX 580 it
was noticeably better still if you're
seeking around 60 FPS for the 1% low
result than the gtx 1070 will have you
covered and for smooth gameplay at this
resolution that's really all you will
require then at the extreme 4k
resolution you'll either have to dial
down the quality settings or bring at
least a GTX 10 a DTI but even then the
performance wasn't really amazing its
playable just not amazing here we can
have a quick look at how the presets
compare using four of the previously
tested GPUs please note I have had to
manually configure the preset each time
as for some reason the developer has
decided to embed the vsync option within
the presets and for some of them it's
forced on very strange that but
hopefully they'll address that one
before release anyway the results are a
little unexpected going from ultra Ohio
only boost the performance of the rx 580
in GTX 1060 by about eight to nine
percent and we saw a slightly larger 11
percent bump for the GTX 1080 however a
much larger 23% bump was seen for Vegas
64 though the frame time performance was
still slightly slower than that of the
GTX 1080
using the same quality settings then
from high to medium we saw around a 40
percent boost for all GPUs tested and
now Vegas 64 in the GT X 1080 are able
to maintain over 100 FPS at all times
during our test battlefield 1 was one of
the first games to really kill quad-core
CPUs the 64 player multiplayer that is
as expected battlefield 5 is no easier
on quad cores I regularly saw my core i7
87 arcade test system which features all
cores clocked at 5 gigahertz exceed 70
percent utilization and at times peaked
at over 80 percent so good luck with
just 4 threads VRAM usage wasn't
actually that extreme surprisingly the
GTX 280 I saw just 3 point 4 gigabytes
allocated at 1080p and then about three
point eight gigabytes at 1440p and this
explains why the three gigabyte 10
succeeded so well at 1080p however you
will want to ensure that your gaming rig
has at least 16 gigabytes of RAM as we
saw memory usage constantly exceed 8
gigabytes and again we saw the same
situation with battlefield 1 8 gigabytes
of memory would cause stuttering when
playing the multiplayer portion of the
game well the good news is it looks like
the game plays very well on mid-range
Hardware even a three gigabyte GTX 1060
offered a solid gaming experience at
1080p I haven't had much time to test
Laura and hardware yet like the GTX 1052
irx 570 for example I'll save the big 50
something GPU comparison for when the
game is officially released for now
though it's nice to know that those
rocking a GTX 1060 or rx 580 have a lot
to look forward to
even at 1440p the game can be enjoyed in
all of its glory with Vega 56 or the gtx
1070 and i expect things to only improve
from here unless of course you're using
ray tracing as we saw you can of course
reduce the quality settings to increase
frame rates though going from ultra high
didn't really
for big gains for the most part for that
you will need to drop down to the medium
quality preset it's hard to say at this
point if any unexpected performance is
down in the games beta status display
drivers or if it's really just how the
game behaves or I suppose another
variable being that the multiplayer is
less consistent and single-player
despite the fact that we are reporting
on a three run average the stuttery
directx12 performance wasn't that
surprising this exact issue was seen
when testing battlefield 1 for the first
time in 2016 I guess the only surprise
being at the developer wasn't more on
top of it this time around but I'm sure
we'll see that addressed before too long
also for those of you wondering I am yet
to test out any multi GPU technology
such as SLI for example but I will have
a bash at 4k with a pair of GTX 1080i
cards later today and maybe I'll relax
and play a few rounds
I'm also keen to try a few different
CPUs as well such as the core i5 8600 k
and even the 8400 and then of course the
full range of Verizon CPUs and depending
on demand I may make better CPU
comparison about for the really in-depth
comparison where I'll test a Holley per
CPU I will wait for the official release
and that is going to do it for this one
thanks for sticking in there with me and
my crook avoid fillet it will be goodbye
tomorrow if not the next day Tim will
probably have a video tomorrow actually
I might hmm
go easy on my voice I think anyway if
you did enjoy the video feel free to the
like button for us subscribe for more
content and if you're preciate that
we're doing our box and consider
supporting us on patreon thank you for
watching I'm your host Steve and I'll
see you again next time
We are a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for us to earn fees by linking to Amazon.com and affiliated sites.