WEBVTT

00:00:15.040 --> 00:00:17.920
yeah there you go

00:00:44.160 --> 00:00:52.559
today we are gonna take this pc

00:00:48.399 --> 00:00:54.559
and overclock overvolt or otherwise

00:00:52.559 --> 00:01:00.320
run beyond the manufacturer's specifications every single part of it

00:00:58.239 --> 00:01:04.799
that we can some of the stuff we're gonna do will seriously improve

00:01:02.000 --> 00:01:10.640
performance and some of it definitely won't big shout out to sonic

00:01:07.360 --> 00:01:13.620
by the way for sponsoring this craziness

00:01:10.640 --> 00:01:16.669
let's get started

00:01:19.759 --> 00:01:27.680
so it shouldn't be news to most of you that the first thing you need when you're overclocking is a solid power

00:01:24.400 --> 00:01:29.759
supply inadequate or unstable power can

00:01:27.680 --> 00:01:34.960
adversely affect stability so we chose the seasonic prime titanium fanless a

00:01:32.720 --> 00:01:38.479
seriously cool unit that is literally silent because it has

00:01:37.040 --> 00:01:41.520
no fan so it's the first thing that we're going

00:01:40.400 --> 00:01:49.200
to tune by plugging it into one of the

00:01:44.040 --> 00:01:51.600
208 volt power outlets in the other unit

00:01:49.200 --> 00:01:58.000
so we're running on 120 volt right now sucking about 270 watts

00:01:54.560 --> 00:02:00.479
let's see what happens here

00:01:58.000 --> 00:02:03.840
all right so after a quick cable swap here for this

00:02:02.000 --> 00:02:08.399
beefy extension cord running out of the other room

00:02:05.040 --> 00:02:10.720
we're running at about 190 volts so

00:02:08.399 --> 00:02:18.280
there was some voltage drop and our same stress test

00:02:14.319 --> 00:02:20.319
is now between 255 and

00:02:18.280 --> 00:02:24.239
260 watts this doesn't make anything faster

00:02:22.480 --> 00:02:28.640
but it does improve your system's efficiency though

00:02:26.720 --> 00:02:32.800
probably not enough that i would if you're in north america like run out and

00:02:30.400 --> 00:02:36.160
get 200 volt power put into your house now before doing anything else you're

00:02:34.640 --> 00:02:41.040
gonna want your motherboard's latest BIOS so you can see ours is so new that

00:02:38.480 --> 00:02:46.720
it has little animations in here you're also going to want the latest os updates

00:02:43.760 --> 00:02:50.239
and a handful of tuning utilities then you want to get some baseline

00:02:48.640 --> 00:02:54.720
readings in your preferred performance tests so you can make sure that any

00:02:52.959 --> 00:02:59.040
efforts you're making any dials that you're turning are actually doing

00:02:56.720 --> 00:03:04.480
something and that what they're doing is positive now we've done more detailed

00:03:02.239 --> 00:03:09.200
CPU overclocking guides in the past so we're going to skip over some details

00:03:06.080 --> 00:03:11.920
here but basically by adjusting our CPU

00:03:09.200 --> 00:03:19.120
multiplier to 51 and disabling svid we were able to get

00:03:14.800 --> 00:03:20.680
our core i5 8600k from 4.3 gigahertz on

00:03:19.120 --> 00:03:28.519
all cores to 5.1 gigahertz with all cores active at

00:03:24.680 --> 00:03:34.400
1.275 volts so that's only a

00:03:28.519 --> 00:03:37.760
0.005 volt increase and it yields a 20

00:03:34.400 --> 00:03:39.920
or higher improvement in our CPU tests

00:03:37.760 --> 00:03:42.799
so we are off to a very good start

00:03:42.000 --> 00:03:48.480
now this is usually where we stop tooling

00:03:45.599 --> 00:03:53.519
around with the processor because CPU cache overclocking

00:03:50.720 --> 00:03:58.400
doesn't really help much outside of very memory intensive applications but not

00:03:56.720 --> 00:04:03.360
today after getting your CPU core dialed in

00:04:00.560 --> 00:04:07.519
you can actually start upping the cache until it's at a similar frequency to the

00:04:06.640 --> 00:04:12.480
core or until it causes uh

00:04:10.159 --> 00:04:19.440
stability problems so ours was able to go from 4 to 4.8 gigahertz with a little

00:04:16.320 --> 00:04:22.479
bit more CPU voltage which actually did

00:04:19.440 --> 00:04:23.919
manage to give us a measurable boost in

00:04:22.479 --> 00:04:28.320
cinebench now when it comes to RAM overclocking we

00:04:26.240 --> 00:04:33.280
generally recommend staying away from xmp or other auto memory overclocking

00:04:30.960 --> 00:04:36.960
since while sometimes it works so it can be worth a shot

00:04:34.720 --> 00:04:41.199
other times it can be questionably stable

00:04:38.320 --> 00:04:46.320
as for our kit well it's rated at 3200 megahertz and our overclocking result

00:04:45.120 --> 00:04:50.880
was 3 200 megahertz even with extra voltage

00:04:49.120 --> 00:04:57.680
so we didn't technically overclock our RAM but we did overclock our system past

00:04:54.479 --> 00:04:58.560
the jdec ddr4 spec so

00:04:57.680 --> 00:05:03.520
yeah we got an extra 27 points in cinebench

00:05:00.960 --> 00:05:07.919
so we're going to call that a success from there

00:05:04.800 --> 00:05:10.479
we did actually try fine-tuning the base

00:05:07.919 --> 00:05:16.800
clock of the system as well but especially given that we already had

00:05:14.720 --> 00:05:21.919
everything else like tuned to the red line it wasn't having

00:05:20.400 --> 00:05:25.520
any of it this whole thing is on the knife's edge of stability

00:05:24.160 --> 00:05:28.560
as it is so then with all the tuning we've done

00:05:27.199 --> 00:05:35.600
so far what do we get for our trouble how has

00:05:31.440 --> 00:05:35.600
this affected our gaming experience

00:05:36.160 --> 00:05:42.400
turns out we've netted a whopping 1.2

00:05:39.520 --> 00:05:46.720
percent increase in 3d mark times by and a whole two FPS

00:05:44.320 --> 00:05:51.280
in far cry 5. so

00:05:48.639 --> 00:05:56.400
it might be time to overclock our graphics my preferred strategy for this

00:05:54.000 --> 00:05:59.840
is MSI afterburner in the foreground so we can make any adjustments that we'd

00:05:58.000 --> 00:06:03.919
like and some kind of stress test running in windowed mode in the

00:06:01.680 --> 00:06:09.120
background so we can monitor stability so you fire up afterburner grab that

00:06:06.000 --> 00:06:10.319
there power knob and rip it all the way

00:06:09.120 --> 00:06:14.639
to the max in our case it's like

00:06:12.759 --> 00:06:19.199
112 some older gpus do let you go higher

00:06:16.720 --> 00:06:25.680
then you incrementally increase the core clock until the system becomes unstable

00:06:22.479 --> 00:06:28.240
so we landed on 150 megahertz for our

00:06:25.680 --> 00:06:32.720
core for our memory clock we ended up with plus

00:06:29.560 --> 00:06:35.280
275 so you do that right after your core

00:06:32.720 --> 00:06:39.120
and this gave us more than a 10 frame per second increase in rise of the

00:06:37.840 --> 00:06:42.479
tomb raider although

00:06:41.039 --> 00:06:48.720
everything else hardly changed for some reason

00:06:45.919 --> 00:06:55.120
that's kind of weird why is that well as it turns out NVIDIA's GPU boost 3

00:06:52.160 --> 00:06:59.759
was already turboing the card up close to 2 gigahertz without our adjustments

00:06:58.000 --> 00:07:04.960
that's much higher than the advertised 1683 that's because the cooling

00:07:02.639 --> 00:07:09.599
capabilities of this card are able to sustain it so that means that without

00:07:07.520 --> 00:07:15.840
overclocking you're actually getting close to the maximum performance you'd

00:07:12.160 --> 00:07:18.880
be able to out of a modern card like a

00:07:15.840 --> 00:07:21.919
which is great for most people but

00:07:18.880 --> 00:07:23.680
overclockers without BIOS modding and

00:07:21.919 --> 00:07:28.400
like physically modifying the card to apply more voltage to the GPU

00:07:26.080 --> 00:07:31.759
don't get a lot of performance gains to brag about

00:07:29.680 --> 00:07:36.400
so up until now everything has been like a nanaimo bar

00:07:35.039 --> 00:07:42.160
just goodness but there has been one downside

00:07:40.880 --> 00:07:46.960
thermals so under full load

00:07:44.800 --> 00:07:52.240
our processor for instance is about 15 degrees celsius hotter

00:07:50.479 --> 00:07:56.800
but what could we overclock

00:07:54.479 --> 00:08:01.919
to make the computer cooler what if we were to give more voltage

00:08:00.080 --> 00:08:06.080
to the fans so stashed in here where a hard drive

00:08:04.560 --> 00:08:12.960
would normally go is something pretty special this is a dc

00:08:09.440 --> 00:08:16.080
to dc converter so all we do is plug

00:08:12.960 --> 00:08:16.960
standard molex 12 volt into this side

00:08:16.080 --> 00:08:22.639
then alex is going to adjust this little potentiometer here

00:08:22.639 --> 00:08:32.240
go for it nice whoa can you hear that

00:08:28.720 --> 00:08:35.200
now we've got 17 volts coming out the

00:08:32.240 --> 00:08:38.560
other side which is where we have our fans attached

00:08:37.279 --> 00:08:42.640
so in ida 64 what kind of an improvement

00:08:41.440 --> 00:08:49.839
will we see bam CPU temps down to 69 degrees a seven

00:08:47.760 --> 00:08:55.519
degree improvement and all we had to give up was sound since this guy went

00:08:53.040 --> 00:08:58.959
from like i don't know 20 decibels or so we're not sure since it was below the

00:08:57.120 --> 00:09:04.000
sound floor of our office thanks in part to the fanless power supply up to 55

00:09:02.160 --> 00:09:08.959
decibels which is around a normal conversation or an air conditioner this

00:09:07.120 --> 00:09:13.920
also by the way won't help the lifetime of your fans our noctuas

00:09:11.680 --> 00:09:18.560
actually might be fine but cheaper ones will likely end up letting out the magic

00:09:16.160 --> 00:09:23.839
blue smoke before long so i think pretty much everything inside

00:09:21.519 --> 00:09:28.240
the case is overclocked now but what about the things outside it

00:09:26.240 --> 00:09:33.200
like your mouse and your keyboard and your monitor can they be overclocked

00:09:32.000 --> 00:09:38.959
yes yes they can okay so for your mouse the

00:09:35.839 --> 00:09:41.360
pulling rate in Windows is 125 hertz

00:09:38.959 --> 00:09:46.720
meaning that there can be up to eight milliseconds of lag between moving your

00:09:44.240 --> 00:09:53.120
mouse and it registering which can lead to noticeable delays in fast-paced games

00:09:50.399 --> 00:09:59.440
in Windows 7 it was possible to overclock the polling rate on your usb

00:09:56.160 --> 00:10:01.920
ports which could be great if your mouse

00:09:59.440 --> 00:10:06.959
sensor could handle it but Windows 8 and up have changed the way that usb drivers

00:10:04.160 --> 00:10:12.160
work and this is no longer possible fortunately a lot of gaming mouse

00:10:09.200 --> 00:10:17.040
manufacturers have found ways to solve this problem for you since most of them

00:10:14.160 --> 00:10:23.360
default to 500 hertz and you can use mouse rate to check yours for yourself

00:10:20.240 --> 00:10:25.440
so click and move your mouse around

00:10:23.360 --> 00:10:28.800
bippity boppity this is going to max out at 500.

00:10:27.600 --> 00:10:33.120
there you have it now with that said

00:10:31.440 --> 00:10:37.920
we can go higher this final mouse right here has a

00:10:35.600 --> 00:10:41.760
firmware tool that you can grab we get one crack at this by the way so

00:10:39.920 --> 00:10:48.160
uh and run

00:10:43.680 --> 00:10:50.640
to flash it to a 1 000 hertz firmware

00:10:48.160 --> 00:10:55.600
and then other mouse manufacturers like Corsair and razer have software tools

00:10:53.440 --> 00:10:58.720
that work similarly let's just let this finish up and then

00:10:57.040 --> 00:11:03.120
we'll see if it worked hey success

00:11:01.120 --> 00:11:06.720
oh i can't even i can't even move it fast enough

00:11:04.399 --> 00:11:11.440
to get all the way to a thousand okay whatever it's definitely over 500.

00:11:09.120 --> 00:11:16.640
as for keyboards this is largely dependent on the manufacturer so this

00:11:13.839 --> 00:11:21.200
Corsair one can actually be adjusted with this switch that's right on the

00:11:19.040 --> 00:11:25.040
back of it bear in mind of course that the reason they make this so easy is

00:11:22.800 --> 00:11:31.600
that it can affect compatibility with your system other keyboards

00:11:27.680 --> 00:11:33.760
may not be possible to overclock at all

00:11:31.600 --> 00:11:39.279
moving on to the monitor overclocking this is actually super easy if you've

00:11:36.640 --> 00:11:44.480
got an NVIDIA card all you need to do is go into change resolution right here

00:11:42.320 --> 00:11:49.279
then click customize tick enable resolutions not exposed by

00:11:46.640 --> 00:11:52.880
the display and create yourself a custom resolution accepting all of the terms

00:11:51.519 --> 00:11:59.200
and conditions and warnings that come along with it from here all you gotta do

00:11:56.000 --> 00:12:02.399
is dial in a new refresh rate then click

00:11:59.200 --> 00:12:06.160
test even a small increase over 60 hertz

00:12:02.399 --> 00:12:08.480
like going to 70 or 75 should make a

00:12:06.160 --> 00:12:12.560
very noticeable difference in games if you've got the GPU horsepower to drive

00:12:10.320 --> 00:12:18.079
it now there ended up being a few things that you can't overclock like your SSD

00:12:15.760 --> 00:12:22.560
or your hard drive like if you've got a 7200 RPM disc

00:12:20.639 --> 00:12:27.920
that's all she's got and the closest thing would be buying a faster drive in

00:12:24.880 --> 00:12:31.360
the first place like a wd velociraptor

00:12:27.920 --> 00:12:33.839
that spins at 10 000 RPM

00:12:31.360 --> 00:12:38.800
but like even that is much slower than an SSD also we tried pretty hard to get

00:12:36.880 --> 00:12:44.959
the webcam overclocked and while it's technically possible to increase the FPS

00:12:42.560 --> 00:12:49.360
while decreasing the resolution our logitech unit just

00:12:47.120 --> 00:12:51.680
wouldn't play ball now

00:12:50.240 --> 00:12:56.079
i was really on the fence about including this but alex insisted that we

00:12:54.000 --> 00:13:01.200
include one last thing that you can overclock and it's probably the slowest

00:12:58.320 --> 00:13:05.200
thing about your computer you

00:13:02.560 --> 00:13:10.320
if you don't touch type learn to touch type also learn all of

00:13:07.920 --> 00:13:13.200
the keyboard shortcuts that you can and if you're doing the same operation often

00:13:12.320 --> 00:13:18.320
enough many modern keyboards allow you to

00:13:15.680 --> 00:13:23.600
program it as a macro key you only have to press one button to execute it and

00:13:20.959 --> 00:13:28.880
there's a joke here it's terrible since although you can't stop ai and machines

00:13:26.079 --> 00:13:32.720
from taking over maybe you can use these tips and tricks to stay useful for a

00:13:31.200 --> 00:13:35.920
little bit longer so big thanks to sea sonic for

00:13:34.240 --> 00:13:39.839
sponsoring this video it's pretty freaking cool that they were able to

00:13:37.279 --> 00:13:43.680
make a 600 watt power supply with literally zero noise

00:13:42.000 --> 00:13:47.760
and that they were cool with us than putting a super loud system around it uh

00:13:46.000 --> 00:13:52.399
thanks for watching you guys if this system this video sucked no the system doesn't

00:13:50.639 --> 00:13:55.760
suck the system is cool if this video sucked you know what to do but if it was

00:13:53.920 --> 00:13:58.880
awesome get subscribed hit that like button or check out the link to where to

00:13:57.279 --> 00:14:02.000
buy the stuff we featured in the video description also down there is our merge

00:14:00.800 --> 00:14:05.839
store which has cool shirts like this one and our community forum which you

00:14:04.240 --> 00:14:09.040
should totally join all we've got to do

00:14:07.440 --> 00:14:13.839
you're in you're in you got it oh you're not in yeah not quite

00:14:11.199 --> 00:14:17.279
dang it alex like close to end okay you gotta you gotta finesse it okay you just

00:14:16.399 --> 00:14:21.519
gotta get the pins kind of in there and then

00:14:19.600 --> 00:14:27.199
you get a you're not in yet you know what you'll know when i'm in

00:14:23.839 --> 00:14:27.199
well yeah