maximumpc.com MAR 2020 MAXIMUMPC 73
BEFORE UPGRADING, this machine was fine
for gaming, and could turn its hand to more
serious work at a push. The aging storage
was holding it back, though, and it did feel
sluggish in even day-to-day tasks at times.
After the upgrade (or complete rebuild, if
we’re honest), this machine is an absolute
powerhouse—look at the Cinebench result
below if you need proof (the zero-point
results were taken from the machine before
its upgrade). That storage overhaul means
that it feels smooth and responsive across
the board; no need to go and make coffee
whenever a huge app or game is fired up.
The overall build went well, and there
isn’t much we’d change. Obviously, we’re
luck y hav ing s o much gear to hand—fl ashing
the motherboard BIOS, for instance, just
required slotting in a first-gen Threadripper,
updating the BIOS, then swapping back to
the second-gen chip. Something that took
less than half an hour, as opposed to weeks
getting a replacement motherboard.
There’s still room to add extra fans, and
we may fight the cabling to try to add a little
color, but there’s no need for this rig to be
lit up like the Vegas strip to enjoy its thread-
heavy power. The lighting from the mobo is
subtle, and there’s just enough light from
the RAM and graphics card to see that it’s all
working OK when you hit the power button.
Returning to the performance, there was
one anomaly when testing, and that was in
Total War: Warhammer II—we saw a modest
increase at 4K, as shown, but at lower
resolutions, the frame rate was stuck in the
mid-30s. We knew switching to Game Mode
in Ryzen Master would improve this a little,
but weren’t expecting to hit almost three
times the frame rate at 1080p compared
to Creator Mode—it really does make that
much difference in this one game. At 4K,
the frame rate jumped up to 43fps, which
is a notable improvement. The other games
weren’t affected anywhere near as much,
just a frame or two at 4K. So, we’ll stick withall the cores and threads of Creator Mode
unless we spot sluggish behavior.
The upgrade from a GeForce GTX 1080 to
an RTX 2080 hasn’t blown us away. 3DMark
Fire Strike may be a synthetic test, but it’s
a good indicator of power, and the 7 percent
improvement of the RTX 2080 over the
GTX 1080 speaks volumes. Gaming
performance is improved, especially when
the Threadripper is in Game Mode, but even
here, it isn’t enough for to us recommend
such an upgrade. At least we can experience
real-time ray tracing now, albeit not at 4K.A SERIOUS UPGRADE
(^1)
The tubes on all-in-one CPU
coolers can be unwieldy
beasts, and while we wanted them
to come out of the top of the block
here, we just couldn’t get them to
do so, and so settled with them
coming out of the bottom.
2
We initially intended to place
the radiator in the roof of the
case, but the radiator block and
fans were too l arge to fit above
the motherboard, so we went for
front mounting instead. If needed,
we can always reinstall the front
fans in the top.
(^3)
There’s no obvious place to
mount hard drives in this
case, other than at the front here,
on top of this removable plate.
1
2
3
BENCHMARKS
ZERO-
POINT
Cinebench R15 Multi (Index) 946 4,279 (352%)
CrystalDisk QD32 Sequential
Read (MB/s)^496 3,445 (595%)
CrystalDisk QD32 Sequential
Write (MB/s)^196 2,111 (977%)
Rise of the Tomb Raider (fps) 51 65 (27%)
Total War: Warhammer II (fps) 32 36 (13%)
Tom Clancy’s Ghost Recon
Wildlands (fps)^31 37 (19%)
3DMark: Fire Strike (Index) 17,347 18,571 (7%)
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Our zero-point consists of an Intel Core i7-7700K, 16GB (2x 8GB) of HyperX Savage DDR4 @ 3,000MT/s, an MSI GeForce GTX
1080 Gaming X, and a 250GB Samsung 840 SATA SSD. All tests were performed at 4K at the highest graphical profile.