WEBVTT

00:00:00.160 --> 00:00:07.160
this video is brought to you by the Intel Core i7 4770k unlocked processor

00:00:05.440 --> 00:00:13.240
if you're a speed freak add an Intel 3500 series SSD to your Haswell system

00:00:10.200 --> 00:00:15.719
for a truly Enthusiast class experience

00:00:13.240 --> 00:00:21.880
welcome to my 1080p performance review of the GeForce GTX 760 from NVIDIA this

00:00:19.000 --> 00:00:28.840
card uses the GK 104 core which is the same as the last generation GTX 670 and

00:00:24.519 --> 00:00:30.519
660 Ti the latter of which it replaces

00:00:28.840 --> 00:00:35.200
however there have definit been some tweaks this card is targeted squarely at

00:00:33.000 --> 00:00:40.200
1080p Gamers so it's in that $250 sweet spot price point and it

00:00:38.120 --> 00:00:45.239
delivers compelling performance in modern games at high details at 1080p

00:00:43.640 --> 00:00:49.399
although you might want to step up to something a little bit juicier like a

00:00:47.320 --> 00:00:53.280
GTX 780 if you want a game at resolutions above that so what tweaks

00:00:51.719 --> 00:00:58.480
have been made so a few things number one is clock speeds have been increased

00:00:55.719 --> 00:01:04.559
number two is that memory bandwidth has been increased we've gone from 192-bit

00:01:00.680 --> 00:01:07.960
bus on the 660 Ti to a 256-bit bus here

00:01:04.559 --> 00:01:11.080
RS have also been increased from 24 to

00:01:07.960 --> 00:01:13.680
32 however there has been a decrease it

00:01:11.080 --> 00:01:18.960
actually has fewer Cuda cores than the GTX 660 Ti that it replaces so it

00:01:16.640 --> 00:01:24.079
remains to be seen how that all translate to real world performance

00:01:22.360 --> 00:01:28.720
speaking of real world performance we test all of our games in as real world

00:01:26.600 --> 00:01:33.119
scenarios as we can meaning timed benchmarks using fraps

00:01:30.560 --> 00:01:37.159
yes we know FCAT is more effective however we are only testing single GPU

00:01:35.159 --> 00:01:40.159
setups here so it makes less of a difference in this case and

00:01:38.520 --> 00:01:44.320
unfortunately we don't have a frame capture setup yet so hopefully that's

00:01:42.520 --> 00:01:51.040
coming at some point for our test bench we've got a P9 x79 Deluxe with 16 gigs

00:01:47.040 --> 00:01:53.640
of Kingston RAM a core i7 3960x at 4 GHz

00:01:51.040 --> 00:01:59.200
an Anta High current Pro 1000 watt that's a platinum power supply 128 gig

00:01:56.520 --> 00:02:05.119
SSD and we are using After Burner Andor precision for our overclocking needs

00:02:01.399 --> 00:02:07.960
with the NVIDIA 32039 driver and AMD

00:02:05.119 --> 00:02:11.319
13.6 beta driver we have switched over to using beta drivers just so we can

00:02:09.440 --> 00:02:15.680
make sure we're getting the most out of all of these cards at the time that

00:02:13.200 --> 00:02:21.440
we're doing these videos just a reminder guys we do run all of our graphics cards

00:02:18.959 --> 00:02:25.360
overclocked at Linus Tech tips because we believe overclocking is relatively

00:02:23.280 --> 00:02:30.640
simple we've done guides on it in the past watch our guides do it because it

00:02:28.400 --> 00:02:34.720
is free performance however that means that sometimes the

00:02:32.560 --> 00:02:37.959
way cards stack up in our performance graphs is going to look a little

00:02:36.000 --> 00:02:40.840
different from how other people's do so if you check out the video description

00:02:39.319 --> 00:02:44.480
we've got a link to the overclocks we're running our cards at most of those

00:02:42.319 --> 00:02:48.120
should be pretty attainable for most of you on your own graphics cards so

00:02:46.519 --> 00:02:52.400
without further Ado let's get into the games starting with matro last Light we

00:02:50.480 --> 00:02:57.400
see a reasonably strong performance for the 760 it comes in a little behind the

00:02:54.800 --> 00:03:03.319
7950 but that stands to reason it is a little bit cheaper and it beats out the

00:02:59.440 --> 00:03:07.400
GTX x660 TI however given that this is a

00:03:03.319 --> 00:03:11.080
generational increase in naming sort of

00:03:07.400 --> 00:03:13.560
because it's a 760 replacing a 660 Ti

00:03:11.080 --> 00:03:18.239
which is typically a slightly uh boosted up skew uh maybe it's not as much as we

00:03:16.480 --> 00:03:21.799
might have expected in terms of performance increase and we're going to

00:03:20.280 --> 00:03:26.720
go through the slides pretty quickly because we're going to see uh some

00:03:23.640 --> 00:03:29.480
pretty uh repeating patterns here so

00:03:26.720 --> 00:03:35.000
Tomb Raider is pretty much the same thing so the 760 is a little bit faster

00:03:32.040 --> 00:03:40.840
than a 660 Ti almost within margin of error for this one next up is BioShock

00:03:37.760 --> 00:03:43.640
Infinite where the 760 is just a smidgen

00:03:40.840 --> 00:03:49.040
faster than a GTX 660ti almost within the margin of error moving right along

00:03:46.000 --> 00:03:51.439
we've got Far Cry 3 where the 760

00:03:49.040 --> 00:03:57.799
performs pretty much the same as the GTX 660 Ti Far Cry 3 blood Dragon shows us

00:03:55.159 --> 00:04:02.360
that the GTX 760 really does perform about the same as the GTX 660 GTI in

00:04:00.319 --> 00:04:08.799
this particular game especially once we overclock both cards and finally Crisis

00:04:05.400 --> 00:04:12.599
3 we see a 10% difference in performance

00:04:08.799 --> 00:04:14.840
between the 760 and the GTX 660 Ti so

00:04:12.599 --> 00:04:18.280
that was a bit of a a patterned sort of performance rundown wasn't it which

00:04:16.560 --> 00:04:24.199
leads us more than anything else to price to Performance analysis you get a

00:04:21.320 --> 00:04:30.080
bit of a savings versus the older 660 Ti you get as good or better performance

00:04:27.120 --> 00:04:35.080
across the board compared to the gtx6 60 TI and you get a very mature card

00:04:32.280 --> 00:04:40.400
running on the GK 104 core as well as a few more options because of that 256bit

00:04:37.400 --> 00:04:43.280
bus some manufacturers May opt to offer

00:04:40.400 --> 00:04:49.120
2 gig and 4 gig configurations whereas with 660ti we were tied to 2 gig and 3

00:04:45.919 --> 00:04:50.680
gig however it would be quite um quite

00:04:49.120 --> 00:04:55.280
debatable whether or not you would benefit from more than 2 gigs with this

00:04:53.440 --> 00:04:59.160
card at this time because it's very optimized for 1080P and not really

00:04:57.479 --> 00:05:03.360
resolutions above that although for those higher res texture Gamers out

00:05:01.000 --> 00:05:10.039
there yeah you might want all the vram that you can possibly get so in short

00:05:06.639 --> 00:05:12.000
the 760 replaces the 660ti at a lower

00:05:10.039 --> 00:05:15.479
price point with more performance does it look quite like the generational

00:05:13.759 --> 00:05:19.240
boost in performance that we were looking for maybe not quite especially

00:05:17.960 --> 00:05:24.880
when you compare it to something like 780 versus 680 but is it a worthy card

00:05:22.720 --> 00:05:28.840
at that price I think the answer is yes thank you for checking out this 1080p

00:05:26.680 --> 00:05:32.000
performance review of the GTX 760 from NVIDIA don't forget to subscribe to

00:05:30.639 --> 00:05:36.520
Line's Tech tips for more unboxings reviews and other computer videos