Dell G5 5590 (RTX 2060) Gaming Laptop Review

It’s finally here, Dell’s new G5 gaming
laptop. If you saw my CES videos I was particularly interested in checking this one out, as the
previous G5 offered good gaming specs for a fair price, so let’s see how the new 2019
model stacks up and if it’s a laptop you should consider buying. Starting with the specs my unit has an Intel
i7-8750H CPU, Nvidia RTX 2060 graphics, 16GB of memory running in single channel, however
for testing I upgraded to 16gb in dual channel, a 512GB M.2 NVMe SSD, and a 15.6” 60Hz IPS
screen. You can configure it a fair bit when ordering though, for instance to a 144Hz screen,
i5 CPU or lower graphics, you can find updated prices to different configurations linked
in the description. For network connectivity it’s got gigabit
ethernet, 802.11ac WiFi and Bluetooth 5. While I was in the middle of making this review,
Dell updated the G5 with Intel’s new 9th gen CPUs and Nvidia 16 series graphics, so
expect slightly different performance with those, I believe the laptops are otherwise
the same in terms of design. The plastic lid has a smooth matte black finish
with a blue Dell logo in the center, and it’s a similar story inside, although the palm
rest area has a metal finish with blue accenting on the keys and around the touchpad. All edges
and corners were rounded and smooth, and overall the build quality was good. The starting weight of the laptop is listed
at 2.68kg or 5.9 pounds on the Dell website, however this will vary based on hardware customizations.
My unit came in at under 2.8kg and then just under 3.5kg with the 180 watt power brick
and cables for charging included. The dimensions of the laptop are 36.4cm in
width, 27,3cm in depth, and under 2.4cm in height, so on the smaller side for a 15 inch
laptop, but a little thicker. It’s noticeably trimmed down and less chunky looking compared
to Dell’s 2018 model of the G5, and this results in thinner screen bezels. The 15.6” 1080p 60Hz screen has a matte
finish, though no G-Sync available here. I thought it looked alright but it did seem
noticeably dim to me even at maximum brightness. I’ve measured the current colour gamut using
the Spyder 5 Pro, and my results returned 64% of sRGB, 46% of NTSC and 48% of AdobeRGB,
so on the lower side, but I think less of an issue for a gaming laptop. At 100% brightness
in the center I measured just 245 nits with a 1000:1 contrast ratio, so I wasn’t imagining
it when I thought it looked a bit dim out of the box. I’ve taken a long exposure photo in a dark
room as a worst case backlight bleed test, and the results in my unit were quite good,
nothing noticeable during normal use, however this will vary between laptops and panels. Remember that these are the results from the
lower end 60Hz panel, I would expect the 144Hz upgrade option to be better, but I haven’t
tested it so can’t say for sure. There was some screen flex which seems to
be a result of the plastic lid combined with the large hinge in the center of the screen,
giving a bit less stability out towards the corners, however the hinge felt sturdy. There were no issues opening it up with one
finger, with the large battery in my unit up the front and cooling towards the back
the weight felt evenly distributed, though not sure how that will differ if you opt for
the smaller 60 watt hour battery. Despite the thinner bezel Dell have still
included the 720p camera above the screen. The camera isn’t great, about average, and
similar deal for the microphone, although it does a good job at removing its idle fan
noise. The keyboard has blue printed keys, however
unlike the old G5 I had it also has blue backlighting which can be adjusted between two brightness
levels or turned off with the F10 key. You do have the option of paying more to upgrade
to an RGB keyboard though should you want it. Typing was alright, there was a lack of feedback
from pressing the keys down, but otherwise I liked typing with it, here’s how it sounds
to give you an idea of what to expect. The only thing I didn’t like was the small
arrow keys that Dell seem to be known for using, which were annoying to use in some
games. Mine didn’t have a fingerprint reader, however it appears to be an optional upgrade
when ordering through Dell’s website. There was minimal keyboard flex while pushing
down hard, overall it was fairly solid, and about the same in the wrist rest areas, not
much flex. The touchpad has precision drivers and worked
well, it clicks down anywhere while also having separate left and right click areas toward
the bottom, and doesn’t have the strange double press issue I reported in last years
G5. Fingerprints and dust show up easily on the
matte black interior and lid, however as both are smooth surfaces they were easy to wipe
off. They’re even easier to spot up the back, which is either a glossy plastic or
glass. On the left there’s an air exhaust vent
up the back, USB 3.1 Gen 2 Type-C port which is wired for both DisplayPort 1.2 and Thunderbolt
3, although Dell note only the 2060 configuration has Thunderbolt support, followed by a USB
3.1 Gen 1 Type-A port and 3.5mm audio combo jack. On the right there’s a full size SD card
reader, USB 3.1 Type-A port, and air exhaust vent towards the back. On the back there are air exhausts on the
left and right, then for the rest of the I/O from left to right we’ve got the power input,
HDMI 2.0 output, USB 3.1 Type-A port, mini DisplayPort output although that’s only
on the RTX 2060 model and they don’t specify the version, gigabit ethernet port, and I
like the way they’ve put it so you don’t have to lift the laptop up to unplug it, followed
by a noble wedge lock slot. On the black plastic lid there’s the Dell
logo in the center, it has a blue shiny finish which can also look a little green depending
on the angle. Underneath there are only small vents for
air flow directly above the fans, but we’ll see if this is a problem for thermals soon.
It can be removed easily by taking out 9 screws with a phillips head screwdriver, and only
the second screw from the left up the back was a different size. Once inside from left to right we’ve got
the single M.2 slot, two memory slots, WiFi card up the top in the middle, and large 90
watt hour battery down the bottom. You’ve got the option of getting a smaller 60 watt
hour battery instead and that also gets you a 2.5 inch drive bay, and I really like that
Dell are giving you that option here, something I wish many others offered. The speakers are found on the left and right
towards the front, and they sounded alright as far as laptop speakers are concerned with
a little bass present. Here’s what we’re looking at with the volume at maximum while
playing music, though the Latencymon result was one of the worst I’ve seen. Powering the laptop is a 90 Watt hour battery.
I’ve tested it with the screen brightness at 50%, background apps disabled, and all
keyboard lighting off. While just watching YouTube videos it lasted for 5 hours and 32
minutes, one of the best results in this test out of all gaming laptops tested thanks to
that bigger battery, and it was using the Intel integrated graphics thanks to Nvidia
Optimus. While playing the Witcher 3 with medium settings
and Nvidia’s battery boost set to 30 FPS the battery lasted for 2 hours and 16 minutes,
however before you get too excited after an hour and 45 minutes it had 19% remaining and
the frame rate dipped down to 13 FPS and wasn’t really playable anymore. I suspect this would
be less problematic playing less demanding games, but even so this is one of the best
results for this test I’ve had out of all laptops, again thanks to the large battery,
expect lower results with the 60 watt hour variant. I’ll also note that I never saw the battery
drain during any of my testing with the provided 180 watt power brick. Let’s move onto the thermal testing, again
the bottom of the laptop doesn’t appear to have many vents for airflow, and on the
inside we’ve got two fans and heatpipes shared between the processor and graphics,
so a change of temperature in one component will affect the other. The Dell Power Manager software allows you
to change between a few different modes, however I didn’t really find them to do all that
much, as we’ll see soon. I didn’t test with the cool profile at all as I found it
to cap GPU performance to 300MHz. I’ll also note that while you can install the Alienware
control center software, this did not give me any fan control, the best you get is changing
between the power manager profiles. Thermal testing was completed in an ambient
room temperature of 20 degrees Celsius, so expect different results in different environments.
I’ve tested idle down the bottom with the quiet profile, and temperatures seemed about
average while also being quiet, more on fan noise soon though. The rest of the results
are from combined CPU and GPU workloads and are meant to represent worst case scenarios. The gaming results towards the lower half
of the graph were tested by playing Watch Dogs 2, as I find it to use a good combination
of processor and graphics. The stress test results shown on the upper half of the graph
are from running the Aida64 CPU stress test and Heaven benchmark at the same time to fully
load the system. Let’s start with the stress test results,
and I’ll kick things off by noting that there was no difference in temperatures with
either the default optimized mode or ultra mode selected, at least in this particular
workload. When the CPU is at 99 degrees Celsius it’s thermal throttling, and I was only
able to remove this by undervolting the CPU by -0.15v, as shown by UV on the graph. Just
using my Thermaltake Massive 20 cooling pad wasn’t enough to remove the throttling,
however when combining the undervolt with the cooling pad we saw the best results. I saw similar results in the gaming test at
stock, it was running quite hot, however even at this worst case 82 degrees Celsius I was
not seeing thermal throttling on the GPU, that never happened in any of my testing,
it was however power limit throttling as all RTX graphics seem to do. Again there were
nice improvements just from CPU undervolting here, with even further improvements with
the cooling pad. These are the average clock speeds for the
same tests just shown. We can see in the stress test results that the clock speeds are essentially
the same between the ultra and optimized profiles, that’s why I just kept using the ultra profile
for the rest of the testing. We can see the CPU reach the full 3.9GHz all core turbo speed
of the i7-8750H any time we apply the CPU undervolt, which is honestly quite good. Yes,
it is running hot, as we saw in the last graph, however not that many laptops in this test
will reach the full all core turbo boost speed in this combined CPU and GPU workload, and
considering we could lower the temperatures to what I think are more reasonable ranges
once undervolted or with the cooling pad added, at least compared to 99 degrees, I think this
is a good result, or at least better. So to summarise, the CPU thermal throttles
in combined CPU and GPU workloads out of the box, though we can remove thermal throttling
with a CPU undervolt and achieve full speed from the CPU in this worst case scenario,
something not too many machines I’ve tested are capable of. These are the clock speeds I got while just
running CPU only stress tests without any GPU load. At stock the full 3.9GHz all core
turbo speed of the i7 was almost possible, again not that many laptops seem capable of
doing this out of the box with Aida64 from my testing. This performance clearly comes
at the expense of more heat though, as it was averaging 98 degrees Celsius and intermittently
on the edge of thermal throttling, which is why the undervolt boosted the performance,
and as we can see here it’s also giving us a massive temperature improvement. It doesn’t look like the CPU TDP is really
getting pumped up super high either, at least as reported by hardware info. To demonstrate how this translates into performance
I’ve got some Cinebench CPU benchmarks here. No difference to single core speed as it’s
not enough to cause throttling, while the undervolt allows us to improve performance
just a little. Here are the results from the newest Cinebench R20, I’ll move over to
using this version in the future once I’ve got more data from other machines, so for
now you get both. Here are the GPU only clock speeds while under
a graphical only stress test. No real practical changes with a manual overclock applied in
MSI Afterburner due to power limit throttling being reached, and I couldn’t improve this
by undervolting the GPU. This is why there were no differences in temperatures from this
test between stock or overclocked. As for the external temperatures where you’ll
actually be putting your hands, at idle with the quiet profile on it was in the low to
mid 30s in the center towards the back, about average. While gaming the wrist rest area
stays cool, although the middle rises to the low to mid 50s. It was a similar result with
the combined CPU and GPU stress tests running, and once undervolted it lowered by maybe a
couple of degrees. Underneath it was warm while under stress test, however definitely
not hot to the touch at all, no problems in terms of heat with using it on your lap other
than potentially blocking the small air intakes. As for the fan noise produced by the laptop,
I’ll let you have a listen to some of these tests. At idle with the default optimized profile
it was quiet but the fan was still audible. If we manually swap over to the quiet profile
with the Dell power manager the fan stops completely. Otherwise while gaming or with
the stress tests running there was no difference, around average when compared to many other
gaming laptops I’ve tested, maybe a little quieter. As we saw earlier, undervolting the
CPU did remove thermal throttling, however I didn’t find this to change the fan speed. I forgot to record the fan noise difference
from optimized and ultra, but take my word for it they were identical under these workloads
which is why I just stuck to testing with ultra. When it comes down to it, the Dell G5 gaming
laptop is quite hot out of the box when under any sort of sustained multicore CPU load due
to a lack of power limit throttling, whether that be CPU only or combined CPU and GPU workloads
as we’ve seen. A lack of power limit throttling can be seen as both a good or a bad thing,
the clock speeds do seem to be higher than most other laptops I’ve tested, however
this results in more heat. Unfortunately once the CPU gets too hot and starts thermal throttling
the performance in some games starts to become affected, there were many games tested where
I noted other laptops with the same specs performed better. At the same time the G5
also performed better in some games as it was able to reach those higher speeds by default,
so put on the undervolt and you should get the best of both worlds. Undervolting the CPU is the best way to help
address this, this alone allowed us to remove the CPU throttling and allow it to reach the
full all core turbo boost speed even under combined CPU and GPU workloads, an impressive
result despite the heat that comes along with this. Personally I don’t think the temperatures
once undervolted are really too much different from other laptops with the same specs anyway,
and if you’re willing to go with a cooling pad further improvements to temperatures were
seen. Perhaps more air intake vents underneath could have also helped with the high out of
the box thermals, as we’ve only got some small ones just above the fans. I would have
liked to have seen greater fan control options like we were promised at CES, as I found no
noticeable difference between the default optimized and ultra profiles. Finally let’s take a look at some gaming
benchmarks, I’ve tested these games with the these Nvidia drivers and all available
Windows updates to date installed with the ultra performance profile in use. We’ll
start by looking at all setting levels, then compare with some other laptops after. Battlefield 5 was tested in campaign mode
and not in multiplayer mode, as it’s easier to consistently reproduce the test run. The
purple bars show the results with ray tracing disabled, while the green bars show RTX on.
The RTX results weren’t great at ultra and high settings, though it was usable at high,
for a game like this I’d want higher frame rate though, and RTX off at high settings
both looks and runs better than with RTX on at low settings. Apex Legends was tested with either all settings
at maximum, or all settings on the lowest possible values, as it doesn’t have predefined
setting presets. It played alright even with everything maxed out, averaging about 80 FPS,
and averaging about 40% higher with everything on minimum. Far Cry New Dawn was tested with the built
in benchmark. In this test above 60 FPS averages were possible at ultra, and just over 100
achievable at low settings. Fortnite was tested with the replay feature,
and due to a recent game update the replay I’ve used is a fair bit different compared
to my previous tests, so these results can’t be compared with my past results. Even with
max settings over 100 FPS was easily possible in this well optimized game. Overwatch is another well optimized game and
was tested in the practice range, as other players, bots and even different maps in actual
gameplay affect the frame rate and this allows for consistent testing. Even maxed out at
epic settings was giving us around 140 FPS, plenty to take advantage of the 144Hz screen
upgrade. Shadow of the Tomb Raider was tested with
the built in benchmark, the results were alright here, although other 2060 laptops I’ve tested
like the Scar II were ahead at higher settings, but again there will be some comparisons later. CS:GO was tested using the Ulletical FPS benchmark,
and like always high frame rates were coming out of this test. Even with all settings at
maximum well over 100 FPS was possible with much lower at lower settings if you need it. Rainbow Six Siege was tested with the built
in benchmark, and is a game I’ve found to benefit from Nvidia’s new turing architecture.
I’ve only recently started using a 100% render scale here, as the built in presets
use 50% by default, so the results can’t be compared to most of my previous tests.
In any case even at ultra settings over 100 FPS was still possible here. PUBG was tested using the replay feature,
and over 100 FPS was possible even with the settings maxed out at ultra, with closer to
140 possible at very low settings. At the lower settings in particular this game seems
to benefit from better CPU performance, which we do seem to get with this laptop, however
as a result it does run hotter than I’d like. Assassin’s Creed Odyssey was tested with
the built in benchmark, still enough to play the game fine, though the ASUS Scar II with
same specs was 6 FPS higher at ultra high and 14 FPS higher at high settings, a pretty
big difference. The drop here would be the thermal throttling on the CPU, which would
top out at 99 degrees depending on the game. Dota 2 was tested playing in the middle lane
with an average amount of action going on, and it was running very smoothly without any
problems at all. The average FPS is around 20 higher when compared to the Scar II here,
again as this game seems to favour higher CPU clock speeds, which we are getting it
just runs hot to achieve them. Watch Dogs 2 is a resource intensive game,
although I still found it to play perfectly fine even with ultra settings. Give me a solid
30 FPS in this one and I can play it without any problems. The Witcher 3 was also running well with hairworks
disabled, and played fine with ultra settings in my test, although you can get much higher
frame rates if you prefer with lower settings, with almost 150 at low, but personally I’d
just use the higher settings and have it looking better, given I don’t think it benefits
much from huge FPS. I’ve tested 20 games in total on the G5
in the dedicated gaming benchmark video, check the card in the top right corner if you’re
after more results. Let’s also take a look at how this config
of the Dell G5 compares with other laptops to see how it stacks up, use these results
as a rough guide only as they were tested at different times with different drivers. In Battlefield 5 I’ve got the G5 up the
top in red, and in this game I found it was actually a fair bit below the other 2060 laptops
I’ve tested, the Aorus 15 which was about 10 FPS higher, and the Scar II which was about
5 FPS higher. The G5 was only just a little ahead of the throttled m15 in this test. Here are the results from Far Cry 5 with ultra
settings in the built in benchmark, and things have changed a bit here, with the 2060 more
inline compared to the other 2060 laptops, coming in a little ahead of the Aorus 15 and
just behind the Scar II, granted there isn’t as much difference overall here compared to
Battlefield 5. These are the results from Shadow of the Tomb
raider with the built in benchmark at highest settings, and in this one the G5 is back as
being the lowest 2060 result I’ve got, behind both the Scar II and Aorus 15. Overall I thought the gaming performance from
the Dell G5 gaming laptop was alright, however as we’ve just seen there are other RTX 2060
laptops that are performing better. The G5 seems to be behind due to the previously mentioned
thermal throttling at stock. That said there were some games where the G5 comes out noticeably
ahead like Dota 2, so it depends on the title and how it utilizes the CPU. Now for the benchmarking tools, I’ve tested
Heaven, Valley, and Superposition from Unigine, as well as Firestrike, Timespy, Port Royal
and VRMark from 3DMark, just pause the video if you want a detailed look at these results. As we saw earlier we’ve got the option of
overclocking the graphics and undervolting the CPU to increase performance as this removed
the thermal throttling on the CPU, so let’s see how these changes actually help in gaming. Far Cry 5 was tested using the built in benchmark
at 1080p. At ultra settings there was a 4.3% improvement to average FPS with the CPU undervolted
and graphics overclocked, the best case scenario. Thermals could have been improved further
with the cooling pad, but as we saw before the undervolt is enough alone to remove thermal
throttling on the CPU, so it shouldn’t change performance. The 1% low rose a bit further,
5.4%, though there was a much larger 12% boost to 1% low at low settings where we’re more
CPU bound, so we can definitely get some performance improvements with simple tweaks. I’ve used Crystal Disk Mark to test the
storage, and the 512GB M.2 NVMe SSD that was installed was offering excellent read and
write speeds. I’ve also tested the SD slot with a v90 rated card and unfortunately it’s
on the slower side, but still better to have it than not. For updated pricing check the links in the
description, as prices will change over time. At the time of recording in the US the Dell
website has already put up the 9th gen model, so not even sure you can buy this one anymore.
Anyway the base model is $900 USD, while a higher end version with RTX 2060 like I’ve
got here starts at $1450 USD. I’ll also note that I’m glad most of these seem to
be listed with dual channel memory, the 8th gen ones were primarily single channel. Here in Australia the G5 currently starts
at $1900 AUD with i5 CPU and GTX 1050 Ti graphics, while the specs in my review unit are going
for around $3100 AUD. As the 9th gen models have just launched it might be worth keeping
an eye out for sales if you’re still interested in 8th gen. So what did you guys think about the new refreshed
Dell G5 gaming laptop? A year ago pretty much any time I was asked for a $1000 gaming laptop
I’d either be picking between the GTX 1060 Acer Helios 300 or Dell G5 depending on what
the person was after, as it was offering good gaming performance for money. This is why
I was so excited to see all of the improvements the new 2019 model of the G5 was getting at
CES. While they are good improvements, I can’t ignore that for the most part Dell seems to
be playing catch up with the competition that have had these features such as 144Hz screens,
slimmer bodies and even RGB lighting for years now. With these specs with the RTX 2060 it’s
a few hundred dollars more now, granted that also does offer additional performance, closer
to the GTX 1070 based on my own testing. Until I get the new GTX 16 series models to test
with I won’t know for sure how the $1000 config stacks up, so can only speak based
on the 2060 model in this video. Compared to other 2060 or 1070 laptops, the prices
seem pretty fair at first, however the price goes up as you make customizations such as
the RGB keyboard or 144Hz screen. That seems to be the main thing here, it’s
a refresh with a number of nice improvements, but nothing game changing. I really like that
Dell give you the option of removing the 2.5 inch drive bay in exchange for a larger 90
watt hour battery, even if you don’t want that at least it’s a choice for those that
do. I also liked that the I/O places bigger cables, such as power, ethernet and display
to run out the back and are less likely to get in the way. I wasn’t a fan of the dim
low colour gamut 60Hz screen, however when it comes down to it it’s still ok for gaming,
and you do now have the option of getting the better 144Hz option. I’ve already discussed this in depth, but
out of the box under any sort of multicore CPU load it does run hot, however this is
at the tradeoff of also providing better performance than many others, at least until thermal throttling
is reached. With some simple tweaks it was possible to stop the thermal throttling though.
While Dell told us at CES there would be fan control in the new G5, what we got isn’t
what I was expecting, it’s better than what we had in the previous model though it hardly
does the job. Otherwise the only other thing I didn’t like is the small arrow keys, that’s
nothing new, but I wish they’d just move that section down a little and go full sized,
which I think would be appreciated in a gaming laptop. Let me know what you guys thought about the
new Dell G5 gaming laptop with RTX graphics down in the comments. Hopefully I can get
the already updated version with 9th gen CPU soon, so if you’re new here get subscribed
for future tech reviews like that.

Leave a Reply

Your email address will not be published. Required fields are marked *