My Failed EK Supremacy EVO Threadripper Edition Review
My Failed EK Supremacy EVO Threadripper Edition Review
2017-08-31
cooller masters new master set ms 120
sets you up with a gaming keyboard and
mouse for just $89.99 the keyboard
features many mechanical switches rubber
dome hybrids with a clicky mechanical
feel perky RGB backlighting and 9 preset
LED modes the mouse has durable Omron
switches a 3500 DPI picsArt optical
sensor and matching RGB lighting click
the link in the description for more
information what's up guys how's it
going welcome back to Pauls hardware I
am moving on from my Vega testing
because I think I want to give that a
little bit time for maybe some drivers
to mature and also to figure out how the
pricing is going with that so for
today's video I'm switching back to
thread Ripper mode because I've just had
this guy arrived in the mail this is the
ek supremacy Evo thread Ripper Edition
and if you guys want to check out a cool
video done by gamers Nexus actually
before or third or even launched they
sort of took a look at where the dies
are placed on thread Ripper and what
kind of coverage you currently get with
the existing ace tech blocks that you
can use with the adapter that comes in
the package and my question with Red
River today is since the dies are spread
out and there's four of them since the
heat spreader itself is so big and since
the ACE tech blocks aren't completely
covering the heat spreader is there
going to be a marked improvement in
thermals when we switch from using those
ace tech liquid coolers to something
specifically designed for thread river
like this block from ek for now I'm
gonna get started actually I need to do
some cleanup first - and I'm going to
get my thread Ripper test bench set up
and for my test bench I'm actually going
to be testing out a new X 399
motherboard for today this is the
gigabytes X 399 or is gaming 7
the block itself I have positive things
to say for at least at least from the
get-go I haven't tested it right now but
taking it out of the box pretty simple
it's just an EVO block I did take it
apart to take a closer look at the block
itself and the channels that are under
there you will notice that the actual
contact plate is is huge it's
substantially larger than one of these
standard ace tech copper blocks that
makes contact with the CPU heat spreader
in side-by-side you can see it will
probably make a lot more contact here
and now that should aid and heat
dissipation theoretically I also notice
that the micro fin channels inside the
block which is where the most efficient
heat transfer takes place is a bit
larger than it has been on previous
blocks as well so again theoretically if
that is directly above or mostly above
where the actual dyes are on the thread
ripper then it should theoretically give
a little bit more heat dissipation and
might lead to better temperatures
potentially even better overclocking
mounting the unit itself is also not too
challenging at all thankfully thread
Ripper tr4 socket just like intel's
high-end desktop socket has a backplate
that's pre installed so it's just a
matter of tightening down the four thumb
screws of course after you've applied
thermal paste i will say that the
thermal paste provided is nice it's
actually called thermal grizzly but i
used the entire tube on this one
application and i'll see when i take it
off if it actually got enough coverage
beyond that the water cooling loop
itself i was setting up using the
existing fractal Kelvin cooler that I
had already installed I totally forgot
that the Praxis wet bench is made for
water cooling and that I could have just
removed this bracket to remove this
whatever that will help for putting it
back on it's been a few months at least
now it's been it's been several months
since I actually built this thing but
thankfully this is a copper radiator and
it does have G 1/4 fittings so I was
able to remove the fractal pump / block
and that was a little bit of a challenge
but I did get it off and then my plan
was to use this unit over here which is
from the ek predator 280 which I've had
in storage it's a big alternate 80
millimeter radiator and
I'm here on this end it has a pump combo
that's part of the radiator so I would
have been able to just connect that all
up and I would have my pump reservoir
essentially I would have had to top it
off of course but when I took the box
out of storage it's just been in the
garage for the past few months since I
reviewed this and this this fitting just
broke just snapped so water was leaking
out of the box when I took it out of the
out of the garage and I was like oh
that's not good and eventually
determined that's just just broke I'm
not sure how it happened like I said
it's just been in storage but can't
really use this because of that I might
be able to disassemble this thing and
make use of some of the parts that are
in there but we'll figure that out at a
later date so the solution is to set up
a more standard type of loop with a
reservoir and a pump and then a radiator
and then of course the block over here
so I'm going to reuse the pump reservoir
from arctic panthers since it's not
doing anything right now
that should get the job done and then I
could reassemble all this things and
then I can actually start testing so
let's let's get back to work Shelly
so this is one of those projects that I
was expecting to get done and maybe a
day or a day and a half and it's now
like over a week later but that's okay
there's various reasons for that but
let's focus on the project at hand which
is getting this water cooling loop set
up and how I have accomplished it
block of course is installed right there
as you probably saw earlier and then I
had to accommodate the rest of the loop
which includes the reservoir and the
pump right there as well as the radiator
which is installed back here so I'm
still using the radiator from the
fractal Kelvin which is a three hundred
and sixty millimeter copper radiator I'm
using the tubing and the quick-release
connections from that ek kit that I
showed you which had broken at least the
reservoir and and pump part of that but
salvaged the tubing since I actually
don't have very much flexible to me I
have a lots of I have lots of a PE TG
hardline tubing but not a lot of
flexible stuff so this is nice because
it allowed me to use the the
quick-release is still and that means I
could pop this off and pop on a
different block if I swapped out the
motherboard over 2 X 299 or something
like that so that's convenient and then
for the actual pump and reservoir down
here I just used the same existing pump
and res that I have salvaged from arctic
panther that was still in fine shape so
after getting everything installed and
getting the fittings in place and
whatnot I had to cycle through and do a
bunch of cleaning because some of this
equipment has been used before so I had
to stock up on my distilled water and
then I just basically refilled that
reservoir multiple times over and over
again cycled it through emptied it out
and so on and so forth just to make sure
that everything in there is clean and
tidy so now I can finally finally move
on with my testing so I'm gonna try to
keep things pretty simple since I'm just
testing the 1920 X in there and since I
haven't tested that before I need to get
some baseline numbers first of all and
then I will do a burnin test and see how
hot things get and then of course I will
be switching over to the H 100 i from
corsair to give us a little bit of
comparison so let's begin our testing
the EK water block is currently
installed and this is the idle test so
for idle it's just been sitting here at
idle I was going for about 15 minutes
we're actually closer to 25 minutes now
to let everything settle
down protip I've discovered something
about the rise in master software which
is that if you turn on the act of
monitoring it actually takes up about
five they're like four or five percent a
CPU load and the CPU temperatures go way
up so here's my idle temperature which
is right around it's been about bouncing
between 33 and 35 for the most part
although if you do turn on monitoring
suddenly it becomes like 40 to 45 so
that's why I'm doing things as I am
right now
over here I'm using Hardware monitor to
also monitor temperatures and here you
can see both the offset temperature at
the top as well as the actual
temperature right here there is a 27
degree offset so the top value will
always be 27 degrees warmer than the
current value and here we can again see
we're idling around 33 to 35 degrees
Celsius I also wanted to point out when
it comes to the actual processor since
this is the 1920 X and we have 12 cores
you can see how the course listed right
here and at stock it's actually going to
run at about 3.7 gigahertz when it's
running on most of the course and if I
drill cores though when it's using say 2
or 4 of them can actually boost up to
well beyond 3.7 gigahertz so we're
actually seeing about 4.1 if you look
over here at the maximum values about
4.1 4.15 is what the cores are actually
hitting individually but of course when
we're actually running a load test it's
going to be running that across all the
course so I'll double check that again
when we get back to that and now the
burden test is going well it has been
going now for about 25 minutes
I was actually kena intending to stop it
at 20 minutes and I lost track of time
but that's ok so our maximum temperature
was recorded over here by hardware info
64 and that's actually a little bit
higher than it then it actually was
spiked to seventy six point three but
actually it was running at about 68
degrees Celsius
pretty stable and you can you can tell
that a little bit by the histogram here
the CPU is the blue line going across
the top here which is very stable for
the most part running at about 60
degrees and I don't think it's gonna
change too much more beyond that so I'm
gonna call that my value for the burn
test and just in case you're wondering
systems
ability test is running with CPU FPU and
cash but not system memory which does
tend to lead to slightly higher temps
for the burn test at least from what
I've been told and then also just to
scroll up here and take a look at our
core clocks we're actually running at
about 3.5 gigahertz on most of the cores
although it does fluctuate a bit going
from between about 3.3 to about 3.6
overall and since I reset this right
after the test starter we can see that
since it's running on all cores it never
got over 3.7 gigahertz while it was
under load so next up I'm gonna swap in
the H 100 IV 2 and I will be making use
of my quick disconnects right here to
pop off the third Ripper block and then
we'll get that test up and running
so I now have the the trusty h100 IV to
installed so let's go ahead and fire up
the test bed here and I've applied
plenty of thermal paste again just like
I did with the the ek block in case
you're wondering I was using the ek
thermal paste the same stuff that they
provide it's basically the same stuff
they provide with their GPU blocks and
that kind of thing when you buy them uh
and we're just gonna load into Windows
here see you see how things are looking
those fins appear to be hitting it
pretty hard already
they're running pretty loud there's
hardware info 64 and let's take a look
at our initial CPU temperatures looks
like we're at a nice chilly ninety three
point three degrees Celsius appears to
be climbing fairly quickly ninety four
ninety four point eight ninety five huh
yeah of course they reach one hundred I
it doesn't seem to be doing a very good
job keeping things chilly here and to
running loud you know what happens when
temperatures get really hot and they
keep climbing and climbing and climbing
theoretically oh there goes that is
called thermal junction max TJ Max and
the system just shut itself off to
prevents the 1920 X from being damaged I
know what's going on my cursor each one
hundred Ivy to pump has died
I've couldn't tell my sarcasm as I was
starting off this little bit this is one
of my main issues one of the reasons
that I have stopped using iia OS in any
of my production machines one of the
reasons why I sort of have gotten less
and less inclined to recommending a iOS
and it's just it's just one of those
points of failures that can be very very
challenging to deal with but yeah pump
dead so guys I'm gonna be as candid as I
can right now this has been the video
from hell everything has gone wrong
starting off like over a week ago was
setting up the water cooling loop just
just I
there was so much frustration okay I
can't even talk about a little
individual things that went wrong over
the course of time ending with now my
pumps dead I'm H 100 IV - so the AIO
cooler that I've been telling you guys
I'm going to compare against the entire
time is it's not dead now of course
there's other things I could do I could
pull another AO out I have other options
but at this point I am so sick and tired
of this actual video that I am just
going to end it so I'm sorry for those
of you guys who are looking for a more
direct comparison of the threader for
waterblock to something else maybe I'll
come back to that in the future we'll
see how that goes but for now that's all
for this video guys I hope you have
enjoyed it such as it is and I'm gonna
try to come back soon because I had a
bad week last week - I didn't get many
videos posted
I just need again I kind of get back on
track and and sort of reassess my
situation and hopefully put myself in a
better mood moving forward anyway guys
thank you so much for watching this
video and if you did enjoy it hit the
thumbs up button we'll see you guys the
next one
We are a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for us to earn fees by linking to Amazon.com and affiliated sites.