WEBVTT

00:00:05.040 --> 00:00:10.160
oh 6 a.m

00:00:08.080 --> 00:00:12.880
why i just said this oh wait

00:00:11.280 --> 00:00:16.880
the review embargo just lifted on Intel's new CPU i gotta go check out the

00:00:15.360 --> 00:00:23.680
wall of videos from all my favorite channels about the exact same damn thing

00:00:21.039 --> 00:00:27.279
after this message from adrena's fedrina's cold brew coffees are

00:00:25.519 --> 00:00:30.960
available in lots of different flavors and they come in these convenient 15

00:00:28.880 --> 00:00:35.200
fluid ounce cans check them out using offer code Linus at LMG.g slash

00:00:33.200 --> 00:00:46.559
madrina's coffee to get 40 off today we'll have that link below

00:00:46.559 --> 00:00:53.120
i'm gonna have to be honest with you guys this

00:00:49.920 --> 00:00:55.039
is not a very interesting product on its

00:00:53.120 --> 00:00:58.800
own merit and it's a lot more straightforward than the other launches

00:00:56.719 --> 00:01:02.800
that we've had this year but don't stop watching because there is a

00:01:01.520 --> 00:01:08.119
twist so normally we'd talk about what's

00:01:05.360 --> 00:01:14.000
different between the 18 core core i9 9980xc and its brother the 18 core

00:01:11.840 --> 00:01:19.040
7980xe but you might have noticed that i said

00:01:15.840 --> 00:01:21.840
brother not cousin predecessor that's

00:01:19.040 --> 00:01:28.240
because these two are so similar that they're almost twins there's been no new

00:01:25.119 --> 00:01:31.280
manufacturing process no more cores no

00:01:28.240 --> 00:01:33.840
more cash no more pci express lanes no

00:01:31.280 --> 00:01:38.640
memory frequency or capacity improvement nor even a slight refresh of an existing

00:01:37.040 --> 00:01:41.600
chipset so that we can all run out and buy new motherboards

00:01:40.479 --> 00:01:49.040
so then how did they push the clock speeds higher have they raised the TDP so that

00:01:46.079 --> 00:01:53.040
it can just suck back more juice and run laps around the previous extreme edition

00:01:52.000 --> 00:01:59.920
no no they did not so how does it go faster

00:01:57.920 --> 00:02:04.320
well we know from past experience that Intel has been treating the TDP ratings

00:02:02.320 --> 00:02:10.319
of their cpus more like guidelines for cooler design than actual

00:02:07.840 --> 00:02:15.440
power consumption specifications so let's take a look at the one thing that

00:02:12.400 --> 00:02:18.000
they did change just like the 9th gen

00:02:15.440 --> 00:02:22.080
consumer 9900k our

00:02:19.040 --> 00:02:23.599
9th gen hedt though it should be noted

00:02:22.080 --> 00:02:28.560
that architecturally they're more like 8th and 6th generation respectively

00:02:26.480 --> 00:02:33.760
9980xc gets a soldered heat spreader so maybe

00:02:31.120 --> 00:02:40.239
the idea here is that as long as a cooler spec for Intel's TDP rating can

00:02:36.800 --> 00:02:42.400
handle the CPU with nice low temps then

00:02:40.239 --> 00:02:47.599
we're cool and it can it can turbo up more solder should help with that

00:02:45.519 --> 00:02:51.200
let's see by how much now

00:02:48.560 --> 00:02:57.599
given that it comes in at the same price as the outgoing 7980xe

00:02:53.599 --> 00:02:59.920
the 9980xc only really has to perform

00:02:57.599 --> 00:03:05.040
marginally better to come away with a strong recommendation here

00:03:03.440 --> 00:03:10.080
no i'm sorry i'm just kidding so we're going to be comparing it to AMD's

00:03:06.879 --> 00:03:11.360
offerings and Intel's own mainstream

00:03:10.080 --> 00:03:18.000
platform and i think it's going to be a little rough

00:03:14.959 --> 00:03:21.040
yeah it's a little rough even with its

00:03:18.000 --> 00:03:23.519
higher bass and boost clocks gaming

00:03:21.040 --> 00:03:28.400
really isn't this CPU's forte now there's no reason why it can't game

00:03:26.080 --> 00:03:35.440
but it's immediately obvious why it shouldn't Intel's consumer cpus tooled

00:03:31.920 --> 00:03:37.280
the 9980xes predecessors and the only

00:03:35.440 --> 00:03:42.480
changes here are that it picks up a couple points in directx 12 and in the

00:03:39.840 --> 00:03:46.879
very frequency dependent cs go but productivity is the real bread and

00:03:44.319 --> 00:03:51.760
butter of an HDD chip anyway and here we do get some pretty big jumps

00:03:49.120 --> 00:03:56.640
in performance over last gen especially in 7-zip and cinebench both in single

00:03:54.319 --> 00:04:00.560
threaded and multi-threaded modes a trend that continues down the line

00:03:58.480 --> 00:04:04.400
with decent results in adobe premiere and blender where the base clock

00:04:02.159 --> 00:04:08.799
increase becomes more significant the longer the render takes to complete

00:04:07.120 --> 00:04:14.480
with that said red ripper thanks to its lower pricing

00:04:11.680 --> 00:04:19.840
and higher core counts continues to either compete valiantly or make Intel

00:04:17.280 --> 00:04:23.680
look downright stupid across most of the productivity suite

00:04:22.000 --> 00:04:28.560
now some of you may have noticed that the improvement in performance over last

00:04:25.600 --> 00:04:33.199
gen is really inconsistent if this thing is just

00:04:30.479 --> 00:04:37.040
200 megahertz higher across the spec sheet

00:04:34.000 --> 00:04:38.240
then all other things being equal which

00:04:37.040 --> 00:04:43.520
they are we should be able to pretty much guess

00:04:40.639 --> 00:04:48.320
the uplift but we've got tests here that barely improved at all even after

00:04:46.000 --> 00:04:50.880
multiple runs well

00:04:49.120 --> 00:04:55.520
after further digging we think we've reached the bottom of it if we look at

00:04:52.800 --> 00:05:00.080
our thermal results on first glance it looks like we're getting better thermals

00:04:58.080 --> 00:05:03.600
than our 7980xe and with lower power draw

00:05:02.800 --> 00:05:11.919
but by going deeper we think we discovered the reason that Intel has refused to say

00:05:09.360 --> 00:05:16.880
what their all-core boost multipliers are

00:05:12.800 --> 00:05:20.639
so our avx-512 stress test had this CPU

00:05:16.880 --> 00:05:24.560
running at just 2.8 gigahertz across all

00:05:20.639 --> 00:05:28.160
cores that is lower than the base clock

00:05:24.560 --> 00:05:31.759
of the 9980xc and coincidentally the

00:05:28.160 --> 00:05:33.039
same as the base clock on the 7980xc it

00:05:31.759 --> 00:05:38.880
replaces so users of avx 512 workloads

00:05:36.320 --> 00:05:43.199
yeah you're not getting an upgrade today at all

00:05:40.080 --> 00:05:45.120
switching to avx-2 got us to 3.3

00:05:43.199 --> 00:05:51.280
gigahertz that's just a hundred megahertz over the 7980xe with increases

00:05:48.400 --> 00:05:56.639
in both temperatures and power draw then turning avx off all together netted

00:05:54.000 --> 00:06:01.680
us a healthy 3.8 gigahertz all-core boost that's 400 megahertz higher than

00:05:59.360 --> 00:06:07.600
the 7980xes so that's where the performance increase

00:06:04.479 --> 00:06:08.880
in non-avx workloads like gaming is

00:06:07.600 --> 00:06:14.880
coming from but it also resulted in more power draw

00:06:11.360 --> 00:06:17.840
with our 9980xc easily surpassing every

00:06:14.880 --> 00:06:24.400
other CPU we tested in this scenario so the bottom line is this this new sku

00:06:21.120 --> 00:06:26.080
seems to be basically a 7980xe

00:06:24.400 --> 00:06:31.520
with a different voltage curve which allows its better bind die to draw less

00:06:28.639 --> 00:06:36.479
power at 2.8 gigahertz with the side effect being that it requires more juice

00:06:33.840 --> 00:06:40.080
to achieve its higher true all-core boost clock

00:06:37.919 --> 00:06:43.759
as for that soldered heat spreader it doesn't seem to be making much of a

00:06:41.520 --> 00:06:47.680
difference but there are a couple of extra degrees of headroom i guess so if

00:06:45.840 --> 00:06:52.240
you're a real masochist you can slap a fat cooler on this thing and remove the

00:06:49.680 --> 00:06:55.039
turbo limits just make sure please that you have active vrm cooling for your

00:06:53.919 --> 00:06:59.280
motherboard just like for the 7980xc

00:06:58.080 --> 00:07:05.199
bottom line the numbers that you read on the box for the core i9 9980xe should be taken with

00:07:03.520 --> 00:07:11.680
a dose of salt large enough to prune a whale let's start with the 9. this is

00:07:09.039 --> 00:07:16.880
not a 9th gen processor in so much as it doesn't use Intel's 9th generation core

00:07:14.400 --> 00:07:23.039
micro architecture nor is it even the ninth family of hedt processors it is

00:07:20.240 --> 00:07:26.400
6th gen with some extra seasoning namely salt

00:07:24.319 --> 00:07:31.440
as for these clock speeds well they're fine as long as Intel is

00:07:29.599 --> 00:07:37.120
assuming that their customers don't intend to use their 2 000 processor for

00:07:34.880 --> 00:07:42.960
the kinds of professional workloads that might justify a 2 000 processor which is

00:07:40.560 --> 00:07:47.680
obviously stupid and only going to get stupider as use of avx instructions

00:07:45.520 --> 00:07:51.919
becomes more prevalent so in my opinion

00:07:49.440 --> 00:07:57.199
these marketing numbers are misleading at best and making matters worse

00:07:54.720 --> 00:08:01.680
much worse is that Intel is refusing to be transparent about them we asked

00:07:59.440 --> 00:08:05.919
straight up to have a simple table provided to us so that we'd know what to

00:08:04.000 --> 00:08:09.680
expect during our testing and they wouldn't tell us

00:08:07.599 --> 00:08:16.080
what is the purpose of this song and dance we can measure these numbers

00:08:13.360 --> 00:08:20.319
why not just save us the hassle and put it in the reviewer's guide

00:08:17.840 --> 00:08:25.680
i get that giving us these numbers makes the product look worse

00:08:22.080 --> 00:08:28.560
but like the product is what it is we

00:08:25.680 --> 00:08:33.839
were going to figure that out now it's just worse and you've pissed off the

00:08:31.440 --> 00:08:38.640
reviewers who are supposed to be either recommending or slamming your product

00:08:36.640 --> 00:08:43.200
when the embargo lifts and you guys wonder why you have pr

00:08:40.560 --> 00:08:46.320
problems over there anyway

00:08:44.240 --> 00:08:49.279
9980xc then it's pretty cool

00:08:47.680 --> 00:08:53.279
but i can't think of a use case that i would buy one for

00:08:51.120 --> 00:08:57.760
speaking of buying things mass drops nuforce edc in-ear monitors are the

00:08:55.519 --> 00:09:01.600
sponsor for today's video they feature a single six millimeter dynamic driver on

00:08:59.760 --> 00:09:05.040
each side with both channels closely matched for a cleaner more accurate

00:09:03.040 --> 00:09:08.880
sound you can wear them at home in the office at the gym they're in your

00:09:06.880 --> 00:09:11.600
monitors you can wear them anywhere that's sort of the idea they feature a

00:09:10.320 --> 00:09:17.360
lightweight plastic housing which massdrop says is built to stand up to regular use and they come with two

00:09:15.120 --> 00:09:21.519
removable braided cables so pick them up on massdrop for just 59.99 at the link

00:09:19.839 --> 00:09:25.360
in the video description so thanks for watching guys if you disliked this video you can hit that

00:09:24.080 --> 00:09:29.839
button but if you liked it hit like get subscribed or maybe consider checking out where to buy the stuff we featured

00:09:28.160 --> 00:09:32.800
at the link in the video description also down there is our merch store which

00:09:31.200 --> 00:09:37.640
has cool shirts like this one and our community forum which you should totally

00:09:34.640 --> 00:09:37.640
join