Asus ROG Phone 5 review

Snapdragon 888 synthetic benchmarks

Asus has pulled-off some pretty impressive binning and overclocking stunts for ROG Phone models in the past. It has always been the company's primary goal to get the best possible chips and even individual units for its gaming phones. The modern mobile silicon scene is quite a bit different, though. Qualcomm has been doing its own in-house binning and optimizing for some time now, mostly eliminating the capabilities or viability for an enthusiastic third-party like Asus to do its own cherry-picking.

As a result, the ROG Phone 5 essentially has the exact same Snapdragon 888 chipset that you can find in competing flagships. This time around, there is no mention of CPU or GPU overclocking and small performance bumps on a hardware level, which used to be the same with some of the older ROG Phone models. Don't get us wrong, though, the Snapdragon 888 chip you are getting is still the cream of the crop in terms of performance currently out there. Plus, there are still plenty of software-level tweaks and optimizations that can be done to squeeze more performance out of it. That hasn't changed and is, in fact, more in-depth than ever.

Asus has implemented the Snapdragon 888 in a pretty unique "unlocked" manner, quite unlike anything else available from competitors. So, on the one hand, the engineering team has expanded huge efforts towards internal and thermal design for the ROG Phone 5 in every way imaginable and on top of that, you are permitted to tweak the behavior of the internals on a low level for yourself even further through advanced tweaks in Armoury Crate.

Just to be perfectly clear, you can't really "overclock" anything in the traditional way. The CPU you are getting is still capped at 2.84 GHz on its primary Kryo 680 core, 2.42 GHz on its three other "big" Kryo 680 ones and 1.8 GHz on the four "small" Kryo 680 ones. To the best of our knowledge, the Adreno 660 GPU is also operating at its regular stock clock. What you can definitely tweak, however, is the thermal behavior of various components of the system. In other words, how far you would like to push your maximum and sustained internal temps. Which, depending on how well you do it, can ultimately get you better peak and/or sustained performance, since thermal constraints are the main limiting factor within a closed passive, or semi-passive with the AeroActive Cooler 5 system, such as the ROG Phone 5.

You can really dig deep and drown yourself in multipliers, limiters curves, governors and all of that geeky goodness that is going on under the hood of the ROG Phone 5 and is generally tucked away in every other consumer smartphone. And tweaking can be done both ways as well. If you know what you are doing, you can tactically limit your system performance in a controlled manner, even on a per-app basis, to squeeze the most battery life out of the ROG Phone 5. And Asus has you equally well-covered even if you don't want to get your hands dirty and mess with any of that by offering pre-made performance modes and even a default Dynamic mode, which will automatically kick in and do the basic tuning on its own. It's all about options, and nobody does it quite like Asus ROG.

Transitioning into some actual synthetic benchmarks, we made sure to run all of our tests at the default Dynamic mode, X Mode, as well as X Mode+, which only gets unlocked and substitutes the regular X Mode when you attach the AeroActive Cooler 5 to the ROG Phone 5. We left all of the other in-depth tuning sliders on default, cooled the phone between runs and had the AeroActive Cooler 5 manually set to max for those tests. With the methodology out of the way, we can kick things off with GeekBench and a pure CPU test.

GeekBench 5 (single-core)

Higher is better

• Asus ROG Phone 5(X Mode+FAN)
  1127
• Asus ROG Phone 5(X Mode)
  1121
• Asus ROG Phone 5
  1110
• Samsung Galaxy S21 Ultra (SD)
  1109
• Xiaomi Mi 11
  1085
• Huawei Mate 40 Pro (perf. mode)
  1020
• Samsung Galaxy Note20 Ultra (SD)
  988
• Asus ROG Phone 3 (144Hz)
  975
• Huawei Mate 40 Pro
  920
• Xiaomi Mi 10 Ultra
  901
• Sony Xperia 1 II
  897
• OnePlus 8T
  893

GeekBench 5 (multi-core)

Higher is better

• Asus ROG Phone 5(X Mode+FAN)
  3745
• Asus ROG Phone 5
  3710
• Asus ROG Phone 5(X Mode)
  3709
• Huawei Mate 40 Pro (perf. mode)
  3704
• Xiaomi Mi 11
  3489
• Asus ROG Phone 3 (144Hz)
  3357
• Sony Xperia 1 II
  3318
• Samsung Galaxy Note20 Ultra (SD)
  3294
• Huawei Mate 40 Pro
  3275
• Xiaomi Mi 10 Ultra
  3248
• Samsung Galaxy S21 Ultra (SD)
  3244
• OnePlus 8T
  3126

The ROG Phone 5 is making the best out of its Snapdragon 888 and as such, is scoring as good or slightly better than other flagships with the same Qualcomm flagship chip.

Since most benchmarks are basically measuring peak bursty performance, there isn't enough time for thermal constraints and throttling to properly kick in, running the tests at X Mode and X Mote+ with the AeroActive Cooler 5 doesn't really make that much of a difference. It's a perfectly expected outcome.

Proving our point about the importance of thermal-throttling and endurance over time, we have AnTuTu, which is not only a more-compound benchmark, but also a significantly lengthier one. Hence, we can see the effects of either allowing more lapse thermal constraints on the hardware in X Mode and then also adding an active cooler on top of that. The differences are small, but over a longer period, like a lengthy gaming session, they do compound and can start to make a tangible difference.

AnTuTu 8

Higher is better

• Asus ROG Phone 5(X Mode+FAN)
  718864
• Asus ROG Phone 5(X Mode)
  708531
• Asus ROG Phone 5
  708216
• Huawei Mate 40 Pro (perf. mode)
  686835
• Xiaomi Mi 11
  668722
• Samsung Galaxy S21 Ultra (SD)
  657273
• Xiaomi Mi 10 Ultra
  638497
• Asus ROG Phone 3 (144Hz)
  601858
• OnePlus 8T
  576625
• Samsung Galaxy Note20 Ultra (SD)
  571312
• Sony Xperia 1 II
  534701
• Huawei Mate 40 Pro
  531270
• Asus ROG Phone II (120Hz)
  483239

For the sake of thoroughness, we will mention once again that this review and all of the benchmarks in it were done with a regular ROG Phone 5 unit (not Pro or Ultimate), with 16GB of LPDDR5 RAM and 256GB of UFS 3.1 storage. It is safe to assume that AnTuTu, in particular, will award extra points to the ROG Phone 5 Ultimate edition with its ludicrous 18GB of RAM.

Moving on to graphics tests, we can see the Adreno 660 provides a nice generational boost up in scores, especially if you compare it to the vanilla Snapdragon 865 (up to 35% faster) instead of the boosted Snapdragon 865+, as found in the ROG Phone 3. Qualcomm's Snapdragon Elite Gaming suite of technologies now also includes Variable Rate Shading (VRS) technology, which could provide an additional up to 30% theoretical performance improvements in certain scenarios.

GFX Manhattan ES 3.1 (offscreen 1080p)

Higher is better

• Asus ROG Phone 5(X Mode)
  123
• Asus ROG Phone 5(X Mode+FAN)
  123
• Asus ROG Phone 5
  122
• Huawei Mate 40 Pro (perf. mode)
  117
• Xiaomi Mi 11
  111
• Samsung Galaxy S21 Ultra (SD)
  109
• Huawei Mate 40 Pro
  97
• Samsung Galaxy Note20 Ultra (SD)
  94
• Asus ROG Phone 3 (144Hz)
  92
• OnePlus 8T
  88
• Xiaomi Mi 10 Ultra
  86
• Sony Xperia 1 II
  84
• Asus ROG Phone II (120Hz)
  78

GFX Manhattan ES 3.1 (onscreen)

Higher is better

• Asus ROG Phone 5(X Mode+FAN)
  104
• Asus ROG Phone 5
  103
• Asus ROG Phone 5(X Mode)
  103
• Samsung Galaxy Note20 Ultra (SD)
  86
• Asus ROG Phone 3 (144Hz)
  82
• Xiaomi Mi 10 Ultra
  78
• Huawei Mate 40 Pro (perf. mode)
  78
• Asus ROG Phone II (120Hz)
  68
• Huawei Mate 40 Pro
  64
• OnePlus 8T
  60
• Sony Xperia 1 II
  59
• Samsung Galaxy S21 Ultra (SD)
  58
• Xiaomi Mi 11
  57

GFX Car Chase ES 3.1 (offscreen 1080p)

Higher is better

• Asus ROG Phone 5(X Mode)
  72
• Asus ROG Phone 5(X Mode+FAN)
  72
• Asus ROG Phone 5
  71
• Xiaomi Mi 11
  67
• Samsung Galaxy S21 Ultra (SD)
  66
• Huawei Mate 40 Pro (perf. mode)
  64
• Samsung Galaxy Note20 Ultra (SD)
  57
• Asus ROG Phone 3 (144Hz)
  56
• Huawei Mate 40 Pro
  56
• OnePlus 8T
  53
• Sony Xperia 1 II
  51
• Xiaomi Mi 10 Ultra
  51
• Asus ROG Phone II (120Hz)
  47

GFX Car Chase ES 3.1 (onscreen)

Higher is better

• Asus ROG Phone 5
  59
• Asus ROG Phone 5(X Mode)
  59
• Asus ROG Phone 5(X Mode+FAN)
  58
• Samsung Galaxy Note20 Ultra (SD)
  51
• Asus ROG Phone 3 (144Hz)
  48
• Xiaomi Mi 10 Ultra
  46
• OnePlus 8T
  46
• Huawei Mate 40 Pro (perf. mode)
  43
• Asus ROG Phone II (120Hz)
  40
• Sony Xperia 1 II
  39
• Samsung Galaxy S21 Ultra (SD)
  33
• Xiaomi Mi 11
  33
• Huawei Mate 40 Pro
  25

Overclocked or not, we can clearly see that the ROG Phone 5 is squeezing a few frames more, on average, out of the Adreno 660, compared to competitors like the Xiaomi Mi 11 and Samsung Galaxy S21 Ultra. The lead here is definitely more pronounced than in the CPU section. Again, this boost is likely due to a combination of factors, like excellent thermal design, a higher thermal ceiling and a slew of other performance tuning and tweaks.

Aztek Vulkan High (onscreen)

Higher is better

• Asus ROG Phone 5
  43
• Asus ROG Phone 5(X Mode)
  43
• Asus ROG Phone 5(X Mode+FAN)
  43
• Samsung Galaxy Note20 Ultra (SD)
  34
• OnePlus 8T
  31
• Xiaomi Mi 10 Ultra
  30
• Asus ROG Phone 3 (144Hz)
  27
• Asus ROG Phone II (120Hz)
  27
• Samsung Galaxy S21 Ultra (SD)
  25
• Xiaomi Mi 11
  24
• Huawei Mate 40 Pro (perf. mode)
  21
• Sony Xperia 1 II
  20
• Huawei Mate 40 Pro
  17

GFX Aztek ES 3.1 High (onscreen)

Higher is better

• Asus ROG Phone 5
  40
• Asus ROG Phone 5(X Mode)
  40
• Asus ROG Phone 5(X Mode+FAN)
  40
• Samsung Galaxy Note20 Ultra (SD)
  34
• Asus ROG Phone 3 (144Hz)
  32
• Huawei Mate 40 Pro (perf. mode)
  32
• Xiaomi Mi 10 Ultra
  30
• OnePlus 8T
  29
• Sony Xperia 1 II
  27
• Asus ROG Phone II (120Hz)
  26
• Huawei Mate 40 Pro
  26
• Samsung Galaxy S21 Ultra (SD)
  23
• Xiaomi Mi 11
  22

A few things worth noting here - these numbers only represent peak performance in short loads. The margins will inevitably shrink with longer workloads, like real-life gaming. Also, comparisons on a device-to-device basis are only fair for off-screen tests since the on-screen rendering is dependent on the panel resolution. In the case of the ROG Phone 5, 1080p+ provides a clear benefit to on-screen fps, compared to the QHD+ displays on the Xiaomi Mi 11 and Samsung Galaxy S21 Ultra.

GFX Aztek Vulkan High (offscreen 1440p)

Higher is better

• Asus ROG Phone 5
  32
• Asus ROG Phone 5(X Mode)
  32
• Asus ROG Phone 5(X Mode+FAN)
  32
• Xiaomi Mi 11
  28
• Samsung Galaxy S21 Ultra (SD)
  25
• Samsung Galaxy Note20 Ultra (SD)
  23

GFX Aztek ES 3.1 High (offscreen 1440p)

Higher is better

• Asus ROG Phone 5
  28
• Asus ROG Phone 5(X Mode)
  28
• Asus ROG Phone 5(X Mode+FAN)
  28
• Samsung Galaxy S21 Ultra (SD)
  25
• Xiaomi Mi 11
  25
• Samsung Galaxy Note20 Ultra (SD)
  22

All of these tests were also done at a forced 144Hz refresh rate on the ROG Phone 5, so as to allow the on-screen portions to push and score higher. In fact, the ROG Phone 5 almost managed to reach its 144fps cap while running the older and no longer particularly useful Manhattan 3.0 run in GFXBench. We are dealing with some truly powerful hardware here.

3DMark Wild Life Vulkan 1.1 (offscreen 1440p)

Higher is better

• Huawei Mate 40 Pro (perf. mode)
  6679
• Huawei Mate 40 Pro
  6170
• Asus ROG Phone 5(X Mode+FAN)
  5763
• Asus ROG Phone 5
  5744
• Asus ROG Phone 5(X Mode)
  5740
• Xiaomi Mi 11
  5673
• Samsung Galaxy S21 Ultra (SD)
  5547
• Samsung Galaxy Note20 Ultra (SD)
  4194

As yet another example, we did put the ROG Phone 5 through the usual SlingShot runs in 3DMark, but you might notice that they are not listed here. That's simply because we literally got the maximum possible score and the 3DMark refused to give us a number.

3DMark max score

Before we move on to thermal-throttling behavior, it is important to note that the ROG Phone 5 gets noticeably hot while doing intensive work, like gaming and the heat compounds over time. It never actually gets uncomfortable to hold unless you run deliberate stress tests, but you should still be aware that this is intentional behavior. Asus has one primary goal - keep the internals as cool as possible, so they can perform better and has deliberately designed the entire chassis, including the thick metal frame, to act as a heatsink.

If you are not a particular fan of that and would rather prioritize hand comfort, though, you can definitely do that by either limiting performance on the phone in one of the many available ways or alternatively by getting the AeroActive 5 cooler to have the, in a sense - "have your cake and eat it too". It provides a nice and cooling stream of air to your palms, perfect for mitigating sweat.

AeroActive 5 cooler

The AeroActive line of cooler attachments has been one of the staples of the ROG Phone family since the very start. Kind of a spotlight feature, in fact, since it essentially kickstarted the admittedly still niche but growing endeavor of trying to introduce active cooling to mobile phones. While v1, v2 and v3 of the accessories certainly had their share of differences and innovations from one to the other, the AeroActive Cooler 5 represents the biggest change to the accessory in more ways than on. Unfortunately, not all positive.

Let's start with something positive - the AeroActive Cooler 5 now has two physical AirTrigger buttons of its own. This is a new development that adds yet another layer of controls that can be mapped using the industry-leading Asus solution. The buttons themselves are just hairpin triggers, with little travel, despite their appearance mimicking an analog appearance. They have a decent click to them and are shaped in a way that is convenient to hit with a ring finger. So, overall, perfectly decent, if not mind-blowing in terms of quality, feel and tactile feedback.

Another positive we can instantly point to with the new AeroActive Cooler 5 design is its streamlined profile and relative lightness compared to most previous iterations. All the while, both the convenient kickstand and the "full-featured" dual-LED ROG RGB logo are present. Perhaps the lack of any moving attachment mechanism helps in regards to keeping the construction light.

However, with no moving parts, you are forced to kind of "snap" it into place on the phone with a fair bit of force. This is one of our first gripes with the design. And things go from bad to worse quickly when you consider the new pogo pin design of the Asus Side Connector. Where the ROG Phone 3 had a Type-C port, flanked by a slightly narrower custom port, with essentially the same protruding plug design, the ROG Phone 5 simply has a set of contact plates flanking its Type-C port. It definitely looks slicker, but it makes the spring-loaded contact pins on the AeroActive Cooler hard to align and very vulnerable to damage.

The previous AeroActive design had a bit plug enter the Side Connector of the ROG Phone 3 in a reassuring manner, with a passthrough Type-C plug then included on the accessory.

Now, the AeroActive Cooler 5 expects you to line-up the pogo-pins without much of a guide to go by and then apply quite a bit of force to the other end of the accessory to clip it on. A simple cutout allows you to still access the now recessed Type-C port. A seemingly more-elegant, slicker and definitely cheaper solution, but a far less convenient one. In fact, our AeroActive Cooler 5 managed to get badly damaged during this review (thankfully after the performance testing), snapping off one of the pins. We are counting this particular design as a clear downgrade from previous iterations.

We also have mixed feelings about the actual fan inside the AeroActive Cooler 5. It works just fine but is definitely noisier than the one on the AeroActive Cooler for the ROG Phone 3. We measured it at all of the available speed settings and compared that against the official data, provided by Asus.

Our conclusion - we most-likely got unlucky with an early unit since the extra noise is almost certainly due to some form of scraping. Perhaps a blade is getting stuck on something. Knowing Asus' high standards of manufacturing quality, we are fairly confident that this is a one-off thing and won't be holding it against the ROG Phone 5.

In more practical terms, once we were past the struggle of actually getting the AeroActive Cooler 5 on the phone, we found using it quite pleasant. It doesn't add too much extra weight nor does it get in the way. And while we can't exactly validate the official claims of 15-degree surface and 10-degree CPU temperature drop, there was definitely a tangible difference to both in-hand feel, as well as thermal behavior of the internals with the AeroActive Cooler blasting air onto the back panel.

Thermal-throttling test

Now that we introduced the AeroActive 5 cooler, as sort of our "secret weapon" in battling thermal-throttling, it's time to see how the ROG Phone 5 handles prolonged loads and heat generation. Arguably a much more valuable and relevant metric when it comes to those real-world gaming sessions. For that, we turned to CPU Throttling Test - an excellent app that we've used in the past to see how ramping-down performance is handled. This is actually a process manufacturers have some degree of control over through hardware and software solutions.

ROG Phone 5 thermal throttling: Dynamic Mode

While losing some performance over time is inevitable, the way that happens is important. What you ideally want to see is a gradual, controlled ramp-down, with the worst-case scenario being jarring, sudden dips, indicating that the phone pushed too hard, reached a hard limit and was forced to rapidly dial-back, potentially causing a stutter in-game.

In its default Dynamic mode, the ROG Phone 5 is very well-behaved, gradually dropping its performance over time with no jarring stutters. Exactly the desired behavior. It managed to retain 89% of its original performance after a full hour of torture, which is a great achievement. Mind you, this is an entirely synthetic scenario that keeps the CPU pinned at 100% all of the time and is absolutely not representative of real-world conditions. Purely done as an example of the absolute-worst potential situation to put the ROG Phone 5 in. A fact further backed up by the scorching temperature of its frame mid-testing.

ROG Phone 5 thermal throttling: X Mode

Running the exact same stress test in X Mode, without the AeroActive Cooler 5 on leads to an interesting but expected behavior. Initially, the phone tries to push itself harder, as signified by the higher maximum performance number attained. It even manages to score a higher average number in X Mode. However, this also means that even with the phone's best efforts to stick to a gradual decline in performance, eventually, a couple of dips appear. Mind you, these were not severe enough to cause stutters, as evidenced by the yellow color on the graph. So, bottom line, X Mode will give you more performance, but over time may lead to less stable performance. There's just no avoiding heat build-up in a closed system.

ROG Phone 5 thermal throttling: X Mode+ with AeroActive Cooler 5

Unless you introduce the AeroActive Cooler 5 to the mix, that is. The results are stunning. We don't believe we've ever seen no CPU throttling occur after a full hour of testing on a phone before. Granted, Asus seems to have tuned the performance curve of X Mode+ a bit differently than the regular X Mode, with a tamer initial push. It clearly works out great, however, and achieves truly smooth sustained performance.

This is no small achievement and one that the ROG Phone 5 should rightfully wear as a batch of honor, proving its status as the ultimate gaming top-dog.