WEBVTT

00:00:00.199 --> 00:00:06.120
this video is brought to you by the Corsair Vengeance k70 and k95 these

00:00:04.319 --> 00:00:12.679
fully mechanical keyboards are designed for performance gaming visit Corsair.com

00:00:08.320 --> 00:00:16.240
vengance gaming to learn more 770 1440p

00:00:12.679 --> 00:00:16.240
review take two shot

00:00:17.920 --> 00:00:23.599
one now if you're a regular viewer you've probably heard me say before why

00:00:22.080 --> 00:00:28.880
are you buying a high-end graphics card if you're still connecting to your monitor via VGA I would actually go

00:00:27.439 --> 00:00:32.559
further than that now if you're spending a few hundred on a graphics card and

00:00:30.960 --> 00:00:36.800
you're not running at a resolution higher than 1080p you're probably doing

00:00:35.480 --> 00:00:41.120
something wrong and you're probably misallocating your funds in some way

00:00:38.719 --> 00:00:45.079
because monitors are so inexpensive now they've got those cheap ones on eBay you

00:00:43.000 --> 00:00:50.199
can get IPS panels for a few hundred bucks for a 1440p display that you

00:00:48.320 --> 00:00:55.120
should make that part of your upgrade plan in order to have a better more

00:00:51.760 --> 00:00:57.879
immersive gaming experience the GTX 770

00:00:55.120 --> 00:01:03.079
is a direct replacement for the 670 and comes in with a very similar spec to the

00:01:00.239 --> 00:01:09.560
outgoing GTX 680 it's got that same full-fledged GK 104 core it's clocked

00:01:06.720 --> 00:01:15.080
higher than the 680 by default using GPU boost 2 to achieve its its own boosted

00:01:12.680 --> 00:01:18.600
overclocked speeds and to make it operate you know at a fixed temperature

00:01:17.119 --> 00:01:22.960
and all the cool stuff that comes along with GPU boost 2 it has a better cooler

00:01:21.079 --> 00:01:27.479
that looks outstanding in addition to keeping the card cool enough as well as

00:01:25.439 --> 00:01:33.079
an updated fan profile that makes it less U Less uh liable to ramp up and

00:01:31.079 --> 00:01:38.040
ramp down in an annoying fashion so it's got all that going for it but the big

00:01:35.439 --> 00:01:46.399
improvement over GTX 680 for my standpoint is the 7 GHz gddr5 memory on

00:01:42.240 --> 00:01:48.520
it so it's still using a 256bit bus but

00:01:46.399 --> 00:01:53.600
the memory is clocked substantially higher allowing for it to perform better

00:01:51.320 --> 00:01:58.079
particularly at higher resolutions as you are about to see now we're using our

00:01:56.200 --> 00:02:04.399
usual standardized test bench we've got our Nel core i739 60x at 4 GHz 16 gigs

00:02:01.880 --> 00:02:09.479
of Kingston RAM a p9x 79 Deluxe motherboard a 128 gig SSD for our boot

00:02:06.840 --> 00:02:15.160
drive as well as the latest drivers from both the green and the red teams without

00:02:12.720 --> 00:02:20.160
further Ado let's get into the performance

00:02:17.680 --> 00:02:26.440
numbers Metro last Light is an unbelievably demanding game at

00:02:22.480 --> 00:02:28.560
1440p even the GTX Titan and GTX 780 are

00:02:26.440 --> 00:02:33.120
brought to their knees however their version of brought to knees is very

00:02:31.319 --> 00:02:40.480
different from the rest of the cards you can see here that memory bandwidth is

00:02:36.400 --> 00:02:42.879
king So those cards do well the GTX 770

00:02:40.480 --> 00:02:46.519
beats the card that it's replacing as well as the card that it is very similar

00:02:45.000 --> 00:02:51.680
to in terms of spec except for its increased memory bandwidth quite handly

00:02:49.239 --> 00:02:56.239
in this test with both of the cards from the red team also doing phenomenally

00:02:53.879 --> 00:03:00.680
well partly due to the fact that they have great memory bandwidth and 3 gig

00:02:58.800 --> 00:03:05.640
memory buffers as a opposed to the 2 gig memory buffers that we see on the NVIDIA

00:03:03.080 --> 00:03:10.599
cards that we have in this test in Tomb reader we see a hint of the things that

00:03:08.239 --> 00:03:15.959
we saw in our 1080p performance review of this card where it looks like the GPU

00:03:12.920 --> 00:03:18.280
boost on this particular card isn't that

00:03:15.959 --> 00:03:21.959
competitive with the way that our 680 turns up when overclocked remember we

00:03:20.280 --> 00:03:27.840
overclock all of our cards for our performance testing so what that

00:03:24.040 --> 00:03:30.640
revealed was that our GTX 770 is not

00:03:27.840 --> 00:03:33.959
necessarily faster than R 68 once both cards are overclocked to the max I'm

00:03:32.280 --> 00:03:39.760
expecting this to change at some point with some tweaks and I mean if that's

00:03:36.599 --> 00:03:42.239
not the case I'm expecting most 770s to

00:03:39.760 --> 00:03:46.239
boost up higher than our particular one because remember this is a higher TDP

00:03:44.200 --> 00:03:50.239
card which means it can is able to draw more power it has a higher stock clock

00:03:48.280 --> 00:03:55.680
speed and it has that much faster memory to go with it with all of that said

00:03:52.799 --> 00:04:00.239
we've got GTX Titan and gtx780 in completely their own world way at top of

00:03:57.599 --> 00:04:04.040
the graphs then we've got a cluster of cards that are all pretty much within

00:04:01.879 --> 00:04:09.560
margin of error of the ones on either side of them followed up by the GTX 660

00:04:06.959 --> 00:04:15.560
Ti and the gtx670 down at the very bottom in BioShock Infinite we see

00:04:12.159 --> 00:04:17.440
another example of GTX Titan gtx780

00:04:15.560 --> 00:04:22.639
completely running away with it stealing the show and everything else looking

00:04:19.959 --> 00:04:26.680
like a very sort of clustered grouping of cards once again though we see our

00:04:24.560 --> 00:04:31.680
770 in spite of the fact that we did observe some weird ramping of the core

00:04:28.560 --> 00:04:33.759
clock speed ahead of the GTX 680 and it

00:04:31.680 --> 00:04:37.320
looks like remember guys at higher resolutions we're more dependent on

00:04:35.400 --> 00:04:42.840
memory performance it looks like that memory is really helping out the card

00:04:39.479 --> 00:04:44.639
here in Far Cry 3 same pattern those top

00:04:42.840 --> 00:04:49.639
two cards are the top two cards and everything else is below them but we do

00:04:47.240 --> 00:04:55.320
see the 770 pull ahead not only of the 670 in a pretty sort of impressive

00:04:52.919 --> 00:05:01.720
fashion but we see it pull ahead of the GTX 680 as well so we can attribute this

00:04:58.360 --> 00:05:04.320
to the improved memory band with once

00:05:01.720 --> 00:05:08.199
again moving on to Far Cry 3 blood Dragon remember these are based on the

00:05:05.880 --> 00:05:15.199
same engine but have very different demands on your system we see again GTX

00:05:11.759 --> 00:05:17.440
Titan gtx780 and everything else very

00:05:15.199 --> 00:05:23.919
very close after that with the gtx770 being the only GK 104 based card that

00:05:21.199 --> 00:05:28.240
really separated itself in this test we did rerun these numbers and it's pretty

00:05:26.360 --> 00:05:32.039
bizarre and we're not quite sure how we ended up with the same performance

00:05:30.319 --> 00:05:36.880
with all three of these cards especially at such a GPU bound resolution but there

00:05:34.400 --> 00:05:40.000
you have it the 770 was the only one that really separated

00:05:40.720 --> 00:05:47.880
itself Crisis 3 is another game where we see the GTX 770s very incremental

00:05:46.280 --> 00:05:53.960
improvements over the last generation GeForce cards and we see GTX Titan and

00:05:50.720 --> 00:05:55.880
gtx780 just kind of stomping all over

00:05:53.960 --> 00:06:00.560
everything just like they are probably going to continue to do for a while

00:05:58.039 --> 00:06:07.479
until AMD manages to Launch launch some kind of Counterattack that can in some

00:06:02.520 --> 00:06:11.039
way address the GTX 780 and GTX Titan so

00:06:07.479 --> 00:06:13.880
what's the conclusion for this GTX 770

00:06:11.039 --> 00:06:20.240
is very suitable for running at 1440p just like GTX 680 was before it except

00:06:16.960 --> 00:06:21.880
more so it looks better it's quieter not

00:06:20.240 --> 00:06:25.759
only is it quieter but the way the fan ramps up and down is less annoying you

00:06:23.840 --> 00:06:29.240
get GeForce experience which allows you to keep your drivers updated as well as

00:06:27.840 --> 00:06:32.960
gives you recommendations for your game settings which might not be relevant to

00:06:31.120 --> 00:06:36.199
you but might be relevant to someone who you recommend a card to if you know that

00:06:34.720 --> 00:06:43.080
they can't handle handling all that stuff on their own and it comes in at a

00:06:38.720 --> 00:06:45.840
lower price than the GTX 680 that it is

00:06:43.080 --> 00:06:50.880
inheriting most of its technology from but not its looks you get a card that

00:06:48.280 --> 00:06:54.880
looks very premium sounds very premium performs very premium and is suitable

00:06:52.599 --> 00:06:59.319
for gaming at that resolution at a price point that it wasn't available at before

00:06:57.160 --> 00:07:02.599
so it's really hard to go wrong bear in mind guys the numbers you were looking

00:07:00.800 --> 00:07:06.759
at in this test were based on us running at extremely high settings at that

00:07:04.520 --> 00:07:10.759
resolution even with cards like the Titan to get really smooth playable

00:07:09.479 --> 00:07:14.879
frame rates you'd probably end up turning things down from you know ultra

00:07:12.919 --> 00:07:19.720
settings with anti-aliasing all over the place because it is much more demanding

00:07:17.639 --> 00:07:25.080
on your graphics card to run at a higher resolution thank you for checking out my

00:07:21.720 --> 00:07:27.120
1440p performance review of the GTX 770

00:07:25.080 --> 00:07:34.120
from NVIDIA don't forget to subscribe to lus Tech tips for more unboxings reviews

00:07:29.800 --> 00:07:34.120
and other computer videos