The Best Video Cards of 2017: All Price-Points Compared
The Best Video Cards of 2017: All Price-Points Compared
2017-11-26
doing a GPU roundup is difficult this
year as chaos and pricing has thrown off
easy price comparisons it also becomes
difficult because many of the direct
matchups are closed when based on MSRP
but gets thrown off when looking at
available prices there's the additional
challenge - of specific games performing
better on one brand than another which
is pretty standard across all years so
you'll ultimately have to exercise your
own thoughts and look at the current
market after the video goes live even to
determine how the video cards stack up
versus when they did when we posted the
video as always links will be below -
the reviews for each product and for the
individual products if you would like to
buy them before we get into that this
content is brought to you by the
Thermaltake flow RGB closed-loop liquid
cooler which is a 360 millimetre
radiator plus 3 120 fans that are RGB
illuminated the if-then we'll take it
ring fans at that this is a 4.5 done a
stack pump which is one of the faster
pumps you can learn more at the link in
the description below so as with the
previous round ups we're looking at only
the products we reviewed and fortunately
we've basically reviewed every video
card on the market this year there are a
lot of them though so we're not going to
talk about every single device we'll go
over the most important ones for our
gaming market audience and then of
course again some of these conclusions
will vary based on the price of the card
so we'll try and make it so that you can
take the the conclusions and then factor
it towards however the price is modified
when you end up going to buy the thin
and finally we're going to be limiting
the amount of data here because it's a
roundup so if you want all of the charts
thermals power noise all the games all
that stuff check the individual reviews
they'll be linked below and keep in mind
that the most recent review will have
the most recent data because as drivers
update as new games come out obviously
the numbers change so the most recent
review to be the 1070 TI which is the
one we'd point you toward for the most
recent data set that said we're going to
be looking at mostly percentages today
rather than flat-out FPS numbers because
we're doing matchups in each price
category so it'll be a straight
percentage performance with a 100%
baseline so the top-performing card will
be 100% in the
chart and then the next card its
competitor will be whatever percentage
it is behind the 100% mark for what is
effectively full speed so let's get
started with Vega 56 on a 1070 work our
way down and then we'll come back to the
high end in our Vega 56 review the TLDR
version was that the card if it could be
found at at MSRP of $400 or reasonably
close to 1070 prices made a whole lot of
sense we later revisited Vega 56 and our
hybrid mod where we applied power play
tables 400 watts worth of power a 360
millimeter radiator and pushed it to the
absolute limit the result was obviously
higher performance as expected but also
completely blowing out the power
efficiency that wasn't the point though
point was that you can push Vega 56
pretty far and basically out reformer
equate Vega 64 in most scenarios the
model was impractical to some extent but
again there's a lot of room to play with
Vega with these six and that proved that
in terms of gaming performance the gtx
1070 and vega 56 cards are reasonably
close to their matchups this is a chart
of relative average FPS performance and
shows that the gtx 1070 SC leads
primarily in ghost recon and for honor
while Vega 56 leads in the rest so that
would be DirectX 12 and Vulcan titles
mostly these numbers are what
precipitated the launch of the gtx 1070
TI nagan 56 has a wider spread of games
here where it winds and has a price
which is theoretically lower and here's
a chart of power performance the gtx
1070 is much more efficient but not in a
way that makes Vega 56 significantly
worse for room ambience or for the power
bill in most places if these numbers do
concern you either because your
electricity costs more or you don't have
AC or something like that well your
choice has been made or you need to look
a bit further into the impact of the
difference if not we're still supporting
Vega 56 as a buy but only if it's under
GTX 970 TI prices we should also note
that the reference blower performs
poorly like always even with Nvidia
cards despite having an excellently
built PCB and BRM for Vega 56 in fact
it's the same as being a frontier
Edition which shipped at about $1000
you'll want to either mod the card which
adds cost or wait for partner models the
reference model is
loud as you can see in our rpm to noise
response chart here and pushes high
temperatures across the entire board
Vega is sensitive to those high
temperatures just like Pascal and so
you'll have direct clock benefit from a
better cooling solution and is the
single biggest point against making 56
right now until partner models come out
learn more about Vega under bolting
overclocking and more like our reviews
linked in the description below you can
also find a purchase link to Vega 56 if
you like along with a link to an air
cooler mod that pretty much fixes the
problem so that had vega 56 generally
preferable over the gtx 1070 here's a
look at how the 1070 TI changed the game
we're starting with a non overclocked
numbers but we'll get to overclocking
because it does matter without an
overclock running the reference clocks
because that's what nvidia and forests
on manufacturers we're already running
the gtx 1070 TI at anywhere from 89 to
90 8.4 percent of the GTX 980
performance rx Vega at 56 is listed
strictly for perspective operating at
74% to 88% of the GTX 1080 remember this
is not percent performance of the 1070
TI forbacon 56 but of the 1080 we're
scaling verses a baseline of 100% for
1080 again to be very clear the gt-r x10
70 Ti is a 1080 with one less SM Jai was
like with 1063 gigabyte is 1066 gigabyte
with one less sm the difference amounts
to one of clocks more than shaders as
we'll show in this overclocking chart
and the GDR 5x of the 1080 isn't really
missed on the 1070 TI a lot of folks
jumped on the bandwagon that the 1070 TI
is a pointless card but ever since
launch we've mostly disagreed with some
of that rhetoric the gtx 1070 TI even
before an overclock largely invalidates
the more expensive gtx 10 80s not the
other way around
at $50 cheaper on average the 1070 TI
provides 90 to 98 percent of the
performance when stock and mostly
achieves parity when both are
overclocked as you can find a GTX 1080
on sale for under 500 or unless the 1070
TI's end up around the $480 plus range
exclusively we think about the 1070 TI
makes more sense presently than the more
expensive 10 80s you can save a few
bucks and as we showed in our recent man
vs. mission
overclocking you can get back some
performance with 15 minutes of overclock
and work power consumption isn't even
that different as you'd expect and
that's because it's the same silicon
so to recap everything depends on price
at price parity or favor toward being a
56 we're recommending Vega 56 over the
10 79 TI at a price advantage of 40 to
50 dollars over the GTX 1080 we're
recommending a 1070 TI and maybe a short
overclock under $500 we're recommending
the 10 80s and this is of course
assuming that the individual models are
actually good and not just some more
rebranded flour garbage cards before
getting into all of the mid-range
hardware that includes AMD we're first
addressing the differences between the
1063 gigabyte and 16 byte cards because
this came out a while ago as a recap we
concluded our GTX 1063 gigabyte card
review by declaring that it should have
been named the 1050 Ti this is I think
before the tentative DTI existed the
card doesn't just have the vram it
removes one SM or simultaneous multi
processor from the GPU reducing pore
count by 10% base clock is the same
between them but the core account and
memory capacity are both lower as for
what this means for fps it really
depends on what game you're playing
here's the relative performance for
average FPS across several games the GTX
1066 kilobyte card is represented at
100% and runs at the same clock as the
three gigabyte card for the most part
the GTX 1063 gigabyte card always
maintains at least 88% of the
performance of the six with its average
performance closer to 92 to 93 percent
of the total 1066 gigabyte performance
scalability gets a bit rough when
blasting vram but we're more likely to
become bound by Rob's or cores and
clocks before we go in bound by VRAM
there are a few games where frame time
consistency dips at higher texture
qualities represented by our 1% and 0.1%
low converted FPS values the 1% lows are
mostly consistent with the average as
shown in this chart and we tend to sit
between 88% and 95% of total performance
potential the 0.1 resent low as shown in
this chart are also largely consistent
accept more vram abuse of games like
Mordor dragged us down to 70 percent of
total performance potential
fortunately although occasionally
noticeable this isn't generally
experienced ruining
and also generally speaking if the two
cards cost the same this six gigabyte
version is going to be a better buy they
generally don't cost the same though
have the price gap in years forty to
fifty dollars it may be better to buy a
three gigabyte version unless you're
running specific applications that are
vram intensive for anyone attempting to
make the argument that higher capacity
will last longer for games in the long
run keep in mind that again we're more
likely to become bound by Rob's or Core
count or frequency prior to becoming
memory bound now we've gotten that all
the way we can turn toward the RX 580 on
GTX 10 to 60 comparisons as a refresher
the RX 580 uses the same polaris GPU as
the RX 480 and it was functionally a
refresh the card significantly increased
power consumption over the rx 480 and
the GTX 960 with improvement only to
idle power consumption versus the 480
this was a launch where AMD and its
board partners opted to blast the
polaris clock effectively pre
overclocking the cards and increasing
the voltage going to the core which gave
that extra clock Headroom the result was
a highly competitive product with the
gtx 960 when looking at gaming
performance a higher power draw as a
result whether that's relevant will
leave up to you as for the average gain
in performance the gtx 960 and our RX
580 trade back and forth the award for
quote best gets passed between them
largely depending on game with nvidia
generally pulling ahead in DirectX 11
titles and Andy generally pulling ahead
in Vulcan or DirectX 12 these results
are somewhat expected based on what we
know of the architectures the RX 580 has
room to play with overclocking and under
bolting making it a good option for
people who don't mind being under the
hood so to speak the trouble is
availability in price as ever if you can
find these cards at price parity we
recommend picking based on a per game
performance look figure out which has
the performance most representative of
the titles you intend to play and then
consider power consumption for the
Nvidia argument or consider free sync
for the AMD argument and pick primarily
based on price and availability
afterward this discussion does overlook
the rx 570 though we found the RS 5 so
that would need to be one of the most
promising mantras of the 500 series the
570 cards have been difficult to get
near amaz RP since launch and that
problem was exacerbated with the mining
trend if you can find one below rx 580
prices ideally around 1
eighty dollars then we still like the RX
570 much like Vega 56 and mega64 this
difference is one of shaders and most
games favor the clock difference on and
the card over the shader difference you
can easily compensate in most titles by
overclocking in rx 570 and it's much
more worthwhile as a purchase than the
GTX 950 TI which costs one hundred fifty
two hundred and seventy dollars that
said it has to be available at 180 once
the arts 570 goes above that the
argument gets a bit weak for its price
to performance the gtx 750ti is however
one of the few cards that for a few
months this year was left untouched by
miners making it an odd unexpected
anchor for cheap gaming pcs we never
fully recommended the 1050 Ti the rx 570
was much better at the high end than
before that the 470 and at 1050 or rx
560 fulfilled the low apt it was not
until the GPU shortage that the 10th
SETI developed more value and that's
mostly solved by now moving on to
low-end GPUs we now look at the arch 560
and gtx 950 cards at present the RX 560
ranges from 110 to $140 to save everyone
some time the $140 units are Ripa don't
bother
the only challenger to this level of
worthlessness is the GTX 1050
at 150 dollars also a ripoff bTW I by
the way is not too distant from that so
we're not talking about the 1050 Ti at
150 we're talking about the 1050 and the
RX 560 at 140 to 150 ignoring the
overpriced units be more reasonably
priced at GTX 950 s cost between 112 and
120 dollars with the more recently
priced RX 5 60s around 110 dollars after
rebate if you count those toward price
or 124 flat these cards should both be
close to $100 ideally but the GBO market
behavior has also affected the low-end
unfortunately we'll make do with what we
have though looking at the average
performance charge the gtx 1050
maintains a lead in the same games as
the previous pairings Ghost Recon and
for honor representing our dx11 tests
with firestrike representing one half of
these synthetics the lead tends to be in
the range of eight percentage points the
RX 560 meanwhile maintains a lead in
times by a significant lead in doom and
another significant lead
sniper with small leads in ashes when
the RS 460 launched it was a bad
purchase compared to the 1050 flat out
the RX 560 changed that and showed
performance advantages in direct x12 and
vulcan titles and some gains in DirectX
11 thanks to the extra clocks the 560
doesn't have huge deficit versus the
1050 and DirectX 11 anymore the deficits
are workable and even recoverable with
an overclock but it maintains a large
leads in DirectX 12 and Vulkan API is
that we've tested not shown in these
charts but shown in our review charts
the arts 560 also maintains more
consistent frame time performance in
some games partly thanks to its 4
gigabyte frame buffer as with other
cards it comes down to price at price
parity our pick is the rx 560 but we'd
like to see the 560 closer to $110 and
we'd like to see the 1050 closer to $100
but there's not much we can do about
that and at the high end it's the gtx
1080i
there's no head-to-head matchup here
because there's no head-to-head
competition we have plenty of videos on
1080 T is in fact we have a roundup of
1080 T eyes which we did separately
strictly because there's no competition
so you're just picking between them so
if you want to learn more about that
card check our 1080i round up as for
other devices so things to know we have
a strong stance of don't buy for two
cards on the market right now one of
them is Vega 64 we think it's a bad
purchase if you can spend a few minutes
doing just even an HBM overclock on vega
56 you don't even have to flash the bios
just a basic overclock I'm 56
largely invalid age 64 if you're
comfortable with flashing the BIOS you
can gain back even more of that
performance Delta the only thing you're
going to be missing is the shaders but
you can get back the clocks they
overclock about the same technically
with Vega 56 taking a slight lead
because the extra heat generated by the
extra shaders on Vega 64 so don't buy 64
the price is not good
56 is the same card just slightly fewer
shaders and the shader difference
doesn't really matter and it's a much
better deal and if you overclock it you
basically just got yourself a 64 for
much cheaper the other card we have a
strong don't buy stance on is
titin XP this one's a bit weird for
gaming the Titan XP is invalidated by
the 1080i the difference between them
there are a couple differences but the
main ones one gigabyte of video memory
and that difference is primarily
beneficial from what we understand for
users of deep learning setups machine
learning things like that we actually
worked with one of the users who
purchased a Titan XP for the purpose of
machine learning noting that the extra
gigabyte mattered a lot for his
particular data set size so for most of
our audience being gamers there's no
real point to buy get a 1080 Ti if you
want something that high-end you don't
even have to overclock it some of the IV
partner 1080 eyes will outperform a
tight and I speed out of the box so
those are the two we'd say don't buy
Vega FPE also I guess it's been on sale
for $700 lately that doesn't mean you
should buy it because it's still Vega 64
and 56 just with more vram unless you
know you're gonna use that and you know
who you are if you will use it then
there's not much value in it and you now
have to deal with the difference in the
drivers being the pro drivers and then
and duel drivers and things like that
the rest though we've got a lineup here
that we just went through for all the
cards and what to buy at each category
so hopefully that helps you as always we
have a lot more content coming subscribe
for all of that you can help us out
directly on patreon.com slash gamers
nexus or you can pick up a shirt like
this one or one of our stickers on store
that gamers nexus net where a lot of
them are presently on sale for the
holidays thank you for watching I'll see
you all 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.