WEBVTT

00:00:00.280 --> 00:00:08.639
welcome to the Dual gtx980 versus Triple

00:00:05.560 --> 00:00:13.799
gtx970 Showdown both of these Solutions

00:00:08.639 --> 00:00:16.279
cost over ,000 for the video cards alone

00:00:13.799 --> 00:00:21.000
and bring a monstrous amount of gaming horsepower to the table but only one can

00:00:18.760 --> 00:00:25.240
be victorious today stay tuned for the stunning conclusion and also don't miss

00:00:23.199 --> 00:00:30.320
our journey back to the parking lot where it all started for an outdoor

00:00:27.359 --> 00:00:33.439
unboxing of the fantex enu evolve old school LT

00:00:38.000 --> 00:00:44.039
style from December 13th to 20th 2014

00:00:41.399 --> 00:00:48.520
you can save on select Intel CPUs KN and ssds with special holiday rebates from

00:00:46.120 --> 00:00:53.440
select retailers click now to learn more now I often get misquoted as saying

00:00:51.600 --> 00:00:58.559
that you shouldn't buy two graphics cards if there's a more powerful single

00:00:55.640 --> 00:01:02.280
GPU option available but and we've talked a fair bit about SLI scaling

00:01:00.359 --> 00:01:08.439
recently including in this video right here there is one exception to this rule

00:01:05.239 --> 00:01:11.040
that almost always holds true the second

00:01:08.439 --> 00:01:16.439
tier the one just under the flagship graphics card is usually based on the

00:01:13.680 --> 00:01:21.520
same processing core is cut down in a way that doesn't affect performance too

00:01:18.400 --> 00:01:24.799
dramatically it overclocks like mad and

00:01:21.520 --> 00:01:27.640
is much less expensive so if you can

00:01:24.799 --> 00:01:31.799
stretch your budget two of those sometimes doesn't cost much more than

00:01:29.520 --> 00:01:37.880
one Flagship card because they don't carry that best of the best price

00:01:34.479 --> 00:01:40.119
premium but lonus what if one top of the

00:01:37.880 --> 00:01:45.280
line or even two second fiddle graphics cards just isn't enough GPU for me well

00:01:43.159 --> 00:01:50.399
my first response to that would be why the heck not that's an awful lot of

00:01:47.799 --> 00:01:54.360
video card power and my second response would be I'm just kidding I know you can

00:01:52.320 --> 00:01:59.159
never have enough video card power especially if you want a game at 4K or

00:01:56.719 --> 00:02:03.840
at very high frame rates I was just messing with you guys so let's take a

00:02:01.560 --> 00:02:09.080
look at the contenders today NVIDIA's top-of-the-line Uber car dour is the

00:02:06.280 --> 00:02:16.040
gtx980 it features a base clock of 1126 MHz a boost clock of 12116 MHz 4 gigs of

00:02:12.160 --> 00:02:18.120
GPU memory clocked at 7 GHz and 248 Cuda

00:02:16.040 --> 00:02:23.200
course sounds awesome right well it should because it costs over

00:02:20.599 --> 00:02:27.480
$550 each to pick up one of these puppies and just over

00:02:25.360 --> 00:02:35.160
$1,100 for the two of them that you'll need to take advantage of that Rog Swift

00:02:31.160 --> 00:02:37.879
1440p 144hz monitor or that brand new 4k

00:02:35.160 --> 00:02:44.800
60 HZ gaming monitor if you want to play the latest AAA titles but maybe there's

00:02:41.000 --> 00:02:47.319
another way the gtx970 by comparison is

00:02:44.800 --> 00:02:53.239
clocked at 1050 MHz on the core has a boost clock of 1178 MHz also features 4

00:02:50.959 --> 00:02:58.080
gigs of GPU memory clocked at 7 GHz and has

00:02:54.200 --> 00:03:00.519
1664 Cuda cor less than the GTX 980 but

00:02:58.080 --> 00:03:05.319
it's not exactly a slouch either and it weighs in at only

00:03:01.959 --> 00:03:10.599
$340 meaning that three of these would

00:03:05.319 --> 00:03:12.640
actually cost less than two GTX 980s so

00:03:10.599 --> 00:03:17.640
let's start with the raw FPS data here shall we across the board average frame

00:03:15.080 --> 00:03:22.799
rates were really close between our two test configurations on our core i7 5960x

00:03:20.640 --> 00:03:28.360
extreme Edition test platform Tomb Raider continues to be the poster child

00:03:25.519 --> 00:03:36.760
for multi-GPU gaming with about a 20% performance Advantage at a 10% lower

00:03:33.120 --> 00:03:39.319
cost for the three-way setup booya Far

00:03:36.760 --> 00:03:44.200
Cry 3 and Crisis 3 tell a different story however both show solid results

00:03:41.959 --> 00:03:49.799
for the Triple config if we just look at the charts but while sitting in front of

00:03:46.680 --> 00:03:51.480
them at 1440p was fine so you high

00:03:49.799 --> 00:03:56.799
refresh rate Gamers would be happy with this config at 4K the stuttering and

00:03:54.439 --> 00:04:02.280
hitching was unbearable with three graphics cards and dual 980s delivered a

00:03:59.439 --> 00:04:07.159
much smoother gaming experience the rest of our games dirt showdown Middle Earth

00:04:04.519 --> 00:04:10.799
Shadow of Mordor and BioShock Infinite didn't have noticeable hitching or

00:04:08.599 --> 00:04:15.360
stuttering but also didn't demonstrate the same performance Advantage as Tomb

00:04:13.000 --> 00:04:20.000
Raider for our three-way configuration something I would have had to see if I

00:04:17.759 --> 00:04:24.720
was going to recommend three-way SLI for folks who play more than just the games

00:04:22.520 --> 00:04:29.120
that happen to run really really well on it especially when you consider that a

00:04:26.840 --> 00:04:33.320
triple graphics card configuration has other dis advantages as well in-game

00:04:31.639 --> 00:04:37.560
power consumption was actually closer than I expected probably because the

00:04:35.320 --> 00:04:40.960
utilization of the third GPU is never going to be as much as the second which

00:04:39.000 --> 00:04:44.840
is never as much as the first so this was basically a draw but both the

00:04:43.120 --> 00:04:48.919
temperatures at which our cards were running and the experience of sitting

00:04:46.919 --> 00:04:54.360
there next to them were a little bit uncomfortable for me our 970 rig was

00:04:52.000 --> 00:05:00.080
much louder and especially the middle sandwich GPU ran much hotter than either

00:04:57.680 --> 00:05:05.000
of our better space 980s and and that's with a fan blowing directly over top of

00:05:02.440 --> 00:05:09.000
them not to mention that two cards will take up fewer of the available expansion

00:05:07.120 --> 00:05:13.400
slots on your motherboard that you might like to use for PCI Express SSD sound

00:05:11.560 --> 00:05:19.479
cards or other fancy new functionality for your system as it becomes available

00:05:16.479 --> 00:05:21.759
so it's conclusion time for gaming I

00:05:19.479 --> 00:05:28.000
just don't think the added bang for the buck of the three-way config of second

00:05:24.199 --> 00:05:30.160
tier cards compensates whoops for the

00:05:28.000 --> 00:05:34.560
disadvantages associated with this kind of a setup you could possibly make an

00:05:32.720 --> 00:05:38.280
argument for it if you were planning to use your system for things other than

00:05:36.560 --> 00:05:43.280
gaming like 3D rendering or cryptocurrency mining or something like

00:05:40.800 --> 00:05:47.720
that but frankly even then the higher clock speed in additional Cuda cores of

00:05:45.120 --> 00:05:54.039
the 980s means that unless the 970s are scaling nearly perfectly the performance

00:05:51.080 --> 00:05:58.560
is going to be similar anyhow and by the time you're spending $1,000 on gpus

00:05:56.520 --> 00:06:02.280
what's another hundred bucks to get a much more elegant solution solution so

00:06:00.639 --> 00:06:06.280
that's all there is really to it this was a fun experiment but the conclusion

00:06:04.240 --> 00:06:09.880
didn't shake the foundations of the pre-existing knowledge we had on the

00:06:07.880 --> 00:06:16.039
topic it just gave us a clear understanding of what folks who want to

00:06:11.199 --> 00:06:19.199
run at 2560 x440 144 Herz or 4K 60 HZ

00:06:16.039 --> 00:06:21.080
need to power their setups speaking of

00:06:19.199 --> 00:06:24.639
clearer understanding we have a new sponsor online as Tech tips that helps

00:06:22.840 --> 00:06:29.720
you better understand the guts of all your electronic devices I fix it you may

00:06:27.520 --> 00:06:35.479
know I fix it from their tear Downs of a multitude of gadgets like their phones

00:06:32.160 --> 00:06:37.319
like the iMac retina 5K uh their guide

00:06:35.479 --> 00:06:42.000
for which I actually followed to create my upcoming iMac 5K internals upgrade

00:06:39.880 --> 00:06:45.639
video but what you may not know about them is that they also have a full

00:06:43.319 --> 00:06:49.160
lineup of professional Electronics tools the Cornerstone of their tool lineup is

00:06:47.240 --> 00:06:53.120
the prot tech toolkit which includes 70 tools to assist you with any mod or

00:06:51.160 --> 00:06:58.000
malfunction on one of your little electronic gizmos the Protech tool kit

00:06:55.639 --> 00:07:03.039
includes a 54-- bit driver kit which features standard special and security

00:07:00.680 --> 00:07:07.039
bits it's got ESD safe Precision tweezers an anti-static wrist strap or

00:07:05.199 --> 00:07:11.599
you can wear it on your ankle like I do a prying a bunch of prying tools and

00:07:08.879 --> 00:07:14.919
both nylon and metal spudgers I never leave home without my Protech toolkit as

00:07:13.520 --> 00:07:18.479
I can always find a use for it when I'm working with electronics and you can

00:07:16.599 --> 00:07:23.080
purchase the kit over at ifixit.com Linus where it's available for only

00:07:21.120 --> 00:07:27.520
$64.95 which by the way includes a lifetime warranty I actually didn't know

00:07:24.680 --> 00:07:31.120
that but wait if you use our offer code Linus you can actually save 10 bucks on

00:07:29.240 --> 00:07:35.120
the prot toolkit or any other purchase over 50 bucks so if I fix it you can fix

00:07:33.720 --> 00:07:39.879
it yourself and if you guys don't believe this obviously written by Nick

00:07:38.000 --> 00:07:45.479
so like very strong endorsement if I fix it it's actually all true in this case

00:07:42.639 --> 00:07:49.960
um I bought my my Protech toolkit myself I paid my own real money for it over a

00:07:47.400 --> 00:07:53.280
year ago and uh it's was just a fluke that they advertise with us now I

00:07:51.440 --> 00:07:56.759
absolutely love my kit and they're great guys over there so uh I think that's

00:07:55.560 --> 00:08:00.199
pretty much it guys like the video if you liked it dislike it if you disliked it leave a comment if your feelings are

00:07:58.960 --> 00:08:03.479
more complicated than this also check out the link in the video description to support us you can buy a cool t-shirt

00:08:02.400 --> 00:08:06.720
like this one give us a monthly contribution or change your Amazon bookmark to one with our affiliate codes

00:08:05.639 --> 00:08:11.840
so we got a small Kickback whenever you buy new graphics cards or whatever else the case may be thanks again for

00:08:10.000 --> 00:08:24.470
watching and as always don't forget to subscribe