ASUS ROG Ally Deep-Dive Review: Thermals, Gaming, Power, SD Card, & More vs. Steam Deck

[Music] this review of the Asus Rog Ally is going to be dense we'll be going through gaming performance battery life noise frequency spectrum and the GN hemianchoic chamber thermals discussion of the SD card situation and how actually great it is to take the Ally apart we're very excited about doing this review and it's been fun to work on because we got to work with a bunch of different handheld devices but there's also a lot of complexity to them so there's a mixture of things that we really liked like the assembly and things that we didn't like like the micro SD card reader breaking we compared the Ally against the steam deck the Nintendo switch and some older Ionia devices we're also working on some competing handhelds next and all that means this review will be long we're coming in late the ReUse cycle here because we waited to buy this unit we were on a waitlist for it pre-ordered it and we got it in and we decided to just throw everything in here so we expect that many of you will jump around the review and that's totally fine we've time stamped things below so you can do that they're mostly containerized sections some kind of rely on others but for the most part they're Standalone and we're mixing things like objective analysis of the FPS the battery life things like that and some subjective discussion of the hardware as in physical feel of the device and software alongside some shared thoughts from both Patrick and I on the teardown process so we've got a bit of everything in here let's get started before that this video is brought to you by us and our brand new 15-year anniversary edition large anti-static mod mat for PC Building the gn15 large mod mat features four grounding snap points one at Each corner to make it easy to move around your work area for PC builds plus it has a ton of new diagrams and useful pin out information the mat is designed to not only protect your table and your parts from dins and scratches during the build but also provide useful information on common front panel connector wiring standard fan orientation pin outs and it offers functionality with screw tracking grids and sorters so that you can more easily reassemble components this highly durable surface is also great for working with heat guns like for Tube Bending and for water cooling projects the mat is already available and it's shipping now on store.gamersaccess.net and it's the best way to help us fund this type of ultra Deep dive review content so that we can produce more of it thanks for your support and grab a mod mat today on store.gamersnexus.net the Asus Rog Ally had a weird start to life because it began as an April Fool's joke that wasn't funny enough to be fake and actually it's two handhelds there's two allies one has a ryzen Z1 extreme processor basically an R7 7840u at 700 and the other allies a cut down ryzen Z1 which will be 600 both of these are called the Rog Ally rc71l and both come with Zen 4 plus rdna3 apus but one has eight cores and 16 threads with 12 cus while the other has six cores and 12 threads with four cu's the 700 Z1 Extreme has a chance at being the handheld price to Performance leader between competitors like the top of the line gpd14 and the io Neo Quinn on One end and and the relatively inexpensive steam deck on the other however there have been some widely reported problems with the Allies since its launch first with performance degradation from a forced bios update and then a flood of broken microSD cards we'll talk about all that as well but let's start with some Hardware Impressions starting with what's included the Allies packaging is pretty simple especially for an Rog product the Ally comes with the power adapter a cheap small cardboard stand and that's it based on our experience with other handheld PCS a docking station with USB Ethernet display out and a carrying case are necessary accessories the Allies 1080p 120 hertz freesync screen is a major selling point on paper but we have mixed practical feelings about it on the positive side 1080p is still the most common gaming resolution whereas the steamedex 1280x800 screen which is actually rotated requires letterboxing or stretching in some games the Allies screen displays brighter colors out of the box than the 512 gigabyte steam deck meaning the steam deck can appear washed out in comparison on the other hand it can be difficult to take advantage of the Allies High refresh rate since a lot of AAA games will struggle to run on this device above 60 FPS at 1080p but the extra pixels versus the deck do help with other smaller details like text legibility an extreme mobile unit with a laptop 3080 not a real 3080 cost fifteen hundred dollars from Best Buy that's insane while Asus is selling a laptop 4090 which is definitely not a real 49d for two thousand dollars we have nothing against egpus but the proprietary standard is awful then the price isn't great either rog's flow connector is annoyingly close to the non-proprietary oculink currently used by gpd's G1 egpu and there's not a great reason to not use a standard the Ally is slightly smaller than a steam deck on paper it's 28 centimeters for the Ally versus 29.8 for the deck and the Ally has Slimmer grips in practice the Ally takes up basically the same amount of space in a bag or in your lap I'm more significant factor though is the Allies weight it's 608 grams to the Dex 669 grams plus or minus some error anyway that way maybe a bad long-term choice for a device that's meant to be touched constantly but it looks great out of the box especially with the RGB lighting around both the joysticks the Dex controls are pretty different the Ally on the other hand sticks to a 100 Xbox series X layout including the joystick positions And the round d-pad the Ally is more familiar while the Dex twin touch pads give it more versatility actual feel is highly subjective it's hard for us to comment on that meaningfully in a way that helps you Patrick thinks that the valve's oversized grips make the deck more comfortable but he thinks the Allies rear buttons are easier to use I personally find that the Allies rear buttons are more difficult to use and while the decks are easier for me I don't think they're good enough to be treated as a standard so in other words two people have different opinions on a subjective matter go figure we really wish the Ally included more than one built-in USB port which would give it a big boost of over the deck we're tired of single port devices here there's space for more we're also disappointed that Asus limited SSD compatibility is at 2230 drives this was a huge opportunity to beat the deck 2230 drives are more expensive per gigabyte than 2280 drives and their smaller physical size limits their capacity in most cases starting with the first thing we noticed Asus has gotten far too comfortable forcing automatic bios updates updates aren't inherently bad and they often add security but forcing them without a warning is bad and so is shoving out a bunch of half-baked hot fixes in Rapid succession which they did as soon as we connected our Ally to the internet Windows update immediately downloaded and flashed a bios update two days later the my Asus app forced another bios update in the background without any input from us in fact it didn't even tell us that it was a bios update it just told us to save our files and buckle up for some reason my Asus is separate from Armory crate another software Hub responsible for yet another set of soft for firmware updates and the one that's used as the default front end for the Ally the second bios update that my Asus forest was version 319 a version which has known performance issues and which Asus had specifically advised against installing nearly a week prior to forcing Adonis Asus said quote some users are reporting a lowered performance on the latest bios 319 if you want to hold off on that bios update you can thanks for the clear language apparently we can't Asus we opened my Asus to disable automatic updates but we found that it's impossible to controller interact with the app at all without creating an account and logging in even as it actively installs updates in contrast to my Asus Armory crate almost looks good it's still disgusting that Asus and gigabyte and maybe MSI now injects bloatware into Windows via motherboard firmware but at least it has a purpose on the Ally the Armory crate overlay Works similarly to the steam Dax but with customizable shortcuts for things like power plans FPS limiters and Radeon super resolution the biggest strike against it is that almost all of its functions other than LED control are just shortcuts for controls that already Exist Elsewhere also having not one but two physical buttons on the Ally bound to both open an Armory crate gives us a bad feeling like the OEM specific buttons on Old keyboards the windows 11 touch keyboard makes a massive difference for Windows on touch screen devices and this is actually one of the stronger points of having windows on a device like this we've come a long way since the Aya Neo founder shipped with a physical button to bring up the on-screen keyboard the windows 11 touch keyboard different from the windows on-screen keyboard behaves exactly like a touchscreen phone it appears whenever a text field is selected we also can't overstate how much easier this makes it to control windows on a handheld it's completely possible to use the Ally without peripherals maybe more so than with desktop mode on Steam OS Asus has also predefined some macros for things like taking screenshots and opening task manager that's on top of the most obvious Windows Advantage which is that the apps these handhelds are meant to run are often native to Windows not Linux of course windows can be installed on a steam deck and steamos can unofficially be installed on an Rog Ally but the out of the box experiences were beginner friendly on the Ally Windows defaults to running at 1920x1080 with 150 scaling on the Ally which is cramped on the screen this size you can change it though 720p and 1080p are the only display mode supported without enabling VSR or connecting an external display so these are the only full screen resolutions that games offer on the Ally we're now moving to our hemianchoic chamber and projecting in here is always very strange because of how quiet it is but we have a separate video talking about how we built it we have a news video talking about why we built it at a basic level and then our fractal Terra ITX review was the first product that we actually used the chamber on and that video contains a lot of background on what we do with it and what the numbers mean so we won't be going into as much of that background information in this video but we'll link all those Below in case you really care about why and how it works so although we have data for each of these devices under idle conditions low load conditions and full load conditions we're only going to focus on presenting the full load conditions with different Power States so that'd be like turbo performance plugged in not plugged in things like that and the reason we're not bothering with idle States for the most part anyway we have some basic numbers in there but is because the devices are already so quiet that it kind of becomes irrelevant but it's the load where the fans have a lot of chance to differentiate themselves in frequency we might have a higher pitched approach versus a lower pitched approach between devices for the numbers we're going to present frequency spectrum and then we're also presenting the DBA spls just the noise level so to speak and additionally we have some noise samples for you so we'll start with those so before we go to the objective plots here's a couple noise samples to get you accustomed to the sound we're boosting these by the same amount in each file and post so you can hear them more clearly but will have unboosted numbers for the analysis here's a sample of the Ally and its loudest scenario that's plugged in and with turbo mode in this clip you can hear a generally higher tone from the fan here's another sample of the Ally in a more common unplugged scenario with performance mode it's much quieter overall and the frequency is also lower unlike the Ally the steam deck caps its frame rate and disallows tearing at a system level by default here's a sample of the steam deck in its loudest condition uncapped with terrain aloud so full frame rate and unplugged that's at a typical position in the SteamVac fan ramp cycle but it has a couple of those so this is the same task but with the RPM and a higher part of its fan rap cycle In that clip you can hear that it ramps purges the Heat and then slows back down and here's one of the steam Deck with a 60 FPS cap and no tearing and finally here's a sample of the Nintendo switch you won't hear much what we can hear is a little bit of a higher frequency but the total noise is very low it's close to the noise floor that we had for testing okay now we have a foundation for the frequency analysis this chart is going to get kind of crazy first some updates we've truncated the frequency start to 20 Hertz this time we've also changed the major grid line step in or the white lines to increment in steps of 10 DB this is because a 10 DB jump is widely regarded as effectively doubling and perceived volume to the human ear finally we reduced the vertical scale from -75 last time to minus 40. these improvements to the Chart legibility came from viewers of our tarot review if you have further suggestions on this and you have some background with this kind of data we'd love to hear from you as we're making improvements to the presentation of each test final background info here this chart is a snapshot of a set of data across a moment in time with the frequency on the bottom and the SPL dbrms on the left axis the data is representative of a common steady State fan behavior for each device we do have some extra notes on the steam deck though you heard it was more variable than the others okay let's get past the blank chart the Ally starts us off it's plugged in and set the turbo so it's the loudest possible condition we noticed the device was overall subjectively loud particularly at the more annoying higher frequencies when compared to the other devices we can see this manifest particularly in the 2000 to 5000 Hertz range with a steep fall off as we approach 6000 Hertz speaking with RS from cybernetics and Hardware Busters one of our peer reviewers for acoustic data he believes the following of frequencies for humans quote two to five kilohertz is the most important frequency zone for us humans our ears are most sensitive there this doesn't mean that lower bands can't be annoying also depending on age frequencies above 15 to 16 kilohertz don't play a role especially if you're over 30. this quote from RS should help give us some perspective note that noise is of course highly subjective so although his beliefs regarding the most sensitive range are commonly agreed with including by us each person may have some differences our other peer reviewer for this content was Mike Chen founder of Silent PC review and an acoustic consultant today he had this contact X to ad for our viewers the three to five kilohertz region is where tonal sounds bother us the most a primary reason often cited is that this is the range where baby crying Falls but we generally hear the entire middle range say 120 hertz to five kilohertz very well and can be sensitive to any tonalities in that region think of the frequency range of a standard 88 key piano the lowest note fundamental is 27.5 Hertz the highest is 4186 Hertz middle C or C5 is 261.63 Imagine middle C held down continuously on an organ even at low volume it wouldn't take long for you to get bothered by it and that's just 261.63 Hertz and again that quote is from Mike Chen who contributed significantly to the fundamentals of Silent PC component testing and has been helping peer review all of our acoustic content so far but the Ally isn't normally plugged in if it's handheld so let's go to unplugged and performance mode this is the more common use case in this mode the Ally drops significantly an overall volume which we'll get to it also has a softer falloff and frequency towards the upper end we no longer have the elevated power and the three thousand to five thousand Hertz range and the tone more gradually softens above that adding silent mode we instantly see that it definitely does what its name says by reducing the volume the entire spectrum is shortened following the reduction in total DBA but particularly we see a reduction in the 2500 to 3 800 Hertz range adding the steam Deck with vsync on we see a noise level that most closely follows the silent mode Ally except for a deviation in the 1400 to 2000 Hertz range keep in mind that the steam DAC has two fan suppliers but we plotted their differences previously ours here is the Hawaiian fan disabling vsync allows the steam deck to get much hotter in this workload from uncapped frame rate drastically increasing the noise levels overall but especially in the 1200 to 2000 Hertz range and then again in the 3500 to 4000 Hertz range the point it falls is similar to the turn turbo and plugged an ally disabling vsync also causes a lot more fan ramp and dramp which is generally more observable from the change but not reflected here since this is only a moment in time we noticed it was typically two to four DBA change between fan speeds during the course of gaming this segment was one of the louder ones finally completing our chaotic chart is the switch similar to the silent Ally and v-synced steam Deck with the noteworthy difference that it's also quieter at the lower end of the frequency spectrum and off of that those sections will get shorter as we get more experience presenting them but for now we're still learning how to present that data let's move to a simpler chart noise levels this shows the total volume so to speak and DBA and under certain conditions all tests are at 20 inches the noise floor was about 13.6 DBA across all tests our chamber can be as low as 11 DBA or as high as 14 but it was 13.6 here the switch was quiet enough to be difficult to measure it was effectively at the noise floor from a front-on perspective of the user or the microphone the steam Deck with vsync on in its lower part of the ramp cycle was around 15.4 DBA and the next quietest device next is the Ally in silent mode at 16.4 DBA so about the same volume as the prior steam deck entry but slightly noticeably louder than the switch the higher end of the fan ramp cycle in the steam deck brought us to 19.7 with vsync on a noticeable increase in volume valve could tune its fan curve for a less observable change here vsync off was loud and has three entries we'll highlight 24.1 DBA at the low end of the ramp 28.2 typically so that's what you'd normally hear and 32.1 DBA during a bursted High wrap cycle which is loud the Ally maxed out at 26.8 DBA which is similar to the deck during its typical cycle with vsync off now we'll talk about the disassembly process we didn't take this apart until after we completely finalized testing in all charts including stuff after this I first took it apart to wire up some thermocouples and then later Patrick took it apart to investigate the the SD card reader integrated circuit basically to try and look for signs of failure so let's go to Patrick first for the thoughts of just the assembly itself Steve has torn down the Ally a little bit and I've torn it down a little bit further I was hunting for the SD card controller but I think both of us have been pretty impressed by how easy it was to take the Ally apart and how maintainable it is in general so I'm just going to walk through starting from the outside there's only six screws holding in the back shell here and then once you have taken the screws out it just pops off the steam deck shell is very difficult to pry apart the Ally not so much the thumb sticks and the battery I think are the most likely candidates for replacement and those can be taken out without removing other pieces heatsink can be left on the motherboard you know you can take out the motherboard as a full unit with the heatsink and fans on it I've taken the heatsink off here but these fans should be pretty easily replaceable as well you know you can just undo these cables and then they're just taped on here and screwed in I think overall this is a pretty modular unit the back shell and the internal parts all of the ones that I removed there's only three different types of screws among them one of them is specifically for the speakers there are no tamper screws on the outside of the unit and there are no hidden screws yes there are anti-tamper stickers but they're not going to stop you from opening up the unit and they're not going to stop you from I think replacing the SSD or some of the other obvious candidates for cleaning and maintenance like the thumb sticks because this is a mobile all-in-one unit um you know in terms of upgrading really you're just going to swap out the SSD and there's not a whole lot else to do here but overall I'd say this is a pretty good effort by Asus um I had some issues with the software but on the hardware side uh very very maintainable back to me for benchmarking methodology and setup so we're covering a ton of this review these tests are just a representative sample we're comparing the Rog Ally again the top of the line 700 model against the 512 gigabyte steam deck as well as two older ion Neo devices that's the next and the founder and we're adding some more devices to this in the near future when benchmarking individual components for especially frame rate we download the latest available drivers and for the Ally we did one different thing which is we replicated a typical user experience that means applying updates through Windows update my Asus and Armory crate rather than manually hunting down the drivers we then paused the updates for the duration of testing this is the experience that Asus largely steers its users towards because they want people to treat it like a handheld console rather than a gaming PC so we followed that direction as end users and we did it that way for other controls freesync was disabled on the Ally and unless otherwise noted vsync was disabled and tearing and allowed on the steam deck we use the term vsync out of convenience but vsync isn't technically accurate on the steam deck the steam decks panel is actually portrait mode and so the screen refreshes move side to side with vertical tearing we also experienced some bugs around tearing and frame rate caps on the deck which we discussed with valve and we came away with some interesting information firstly even if you want viasync which is valuable mostly for battery life preservation purposes of not tearing there's no real reason to enable it in game on the deck according to valve here quote by default gamescope always synchronizes output frames to the physical display unless that allowed tearing option is checked it enables immediate flips in the kernel driver we also had issues with the OS level 60 hertz cap being applied on boot without being shown in the UI but we were eventually able to gather a full set of uncapped tearing enabled deck results and if you want to learn more about all of this actually game scope and many other elements of the Steam OS are open source and fully available on GitHub first some important background about how TDP numbers are often not representative of real power consumption the Ally has three operating modes they are windows power plans they're silent performance and turbo it defaults to performance on battery and turbo unplugged in or a mode can be selected manually through Armory crate Asus describes silence as 10 watts performance as 15 watts and turbo as 25 watts on battery and 30 Watts on wall power as usual though TDP numbers should be ignored in favor of observable measurements TDP is often not fully representative of real power consumption but it depends on how the company calculates TDP some are closer than others TDP also changes between CPUs so the older AMD formula doesn't necessarily apply to all of their CPUs under load a plugged in Ally with a full battery 30 watt TDP can easily draw the full 65 Watts that the AC adapter is capable of delivering while the steam deck a 15 watt TDP pulls 30 Watts under the same conditions without confirmation of actual power draw it is inappropriate to assume Asus is 15 watts is the same as valve's 15 watts so 15 watts does not necessarily equal 15 watts in terms of how unlocked or enabled the device is it could be different now that's stated to really fully understand these you'd have to tap into the battery and measure the actual power exchange there but realistically an end user will just set one of these modes or the other so our solution to this is to test in all relevant modes and then also test battery consumption on all relevant modes and give you both charts separately cyberpunk 2077 starts us out we're using the low preset running at 1280 by 720 without upscaling and using these settings on all devices for like for like test when plugged into the wall and with turbo active the Rog Ally is the clear front runner in performance averaging 47 ahead of the deck at 63 FPS however its lows are not scaling well with that average we'll look at frame times for that momentarily in handheld mode with the default performance profile the Allies averages slightly behind the decks at 40 FPS and the older IO Neo devices lag behind at 31 FPS for the next and 23 for the founder we confirmed through testing that the Ally is the only one of these devices that significantly Alters performance when plugged in so all of our non-ally results were logged on battery power handheld performance is what really matters though and starting out at parity with the steam deck is bad news for the more expensive Ally what's even worse for the Ally is the frame time consistency holding a 9 FPS 0.1 percent low when on battery as compared to the Dax 30fps lows has the deck clearly advantaged let's look at a frame time plot here it is as a reminder these represent the frame to frame interval in milliseconds or the required time for the system to prepare render and flip the frame to the screen lower is better but more consistent is best the Ally on battery starts with a wide 10 to 20 millisecond frame to frame range but it starts running into big problem as the test advances we encounter a series of excursions from about 22 to 30 milliseconds all the way up to 100 milliseconds that's terrible where it's the technical word for it we're looking for an 8 to 12 millisecond range but the Ally has a range or a Max minus Min upwards of 90 milliseconds the steam DAC by comparison is much smoother in frame delivery the beginning of the test is the smoothest once again and then deviates more as the test progresses however the deck stays within our expected range for the entire test this is a significant gameplay experience advantage over the Ally and it shows that average once again doesn't tell the whole story Dying Light 2 is next with Dying Light 2 manually configured to the low preset and we manually switch to dx12 with asynchronous compute enabled and we controlled the settings because the low preset has changed its settings with updates so we manually set those the Ally is plugged in performance was 47 FPS average 32 percent ahead of its 36 FPS average on battery power trimming the Z1 on extremes power allowance hurts relative performance in some games more than it does in others because of that we also ran a turbo pass with the Ally actively charging to see why they're charging the battery would eat into the apu's power but the performance difference was minor some of the swings and when the power allowance benefits more versus less have to do with which component specifically is consuming that TDP allowance so in other words how much of a CPU or GPU component binds there might be the steam deck averaged 29 FPS in this test behind the battery powered results for the Ally and that's unlike the tide cyberpunk result one thing that is consistent though is the inconsistency of frame times the Ally once again suffers as compared to the deck here's a frame time plot for Dying Light 2. once again the Ally struggles hard its first major set of spikes go to 170 milliseconds so 17 percent of a second you're sitting there staring at the same image the next frame is back down towards 16 milliseconds then it's back up towards 70 milliseconds this presents as a stuttery set of images we then encounter similar spikes around frame 250 450 and 780. the Ally is inconsistent in its frame Time Performance this is traditionally a combination of drivers and power delivery and the power gating behavior on the Ally may be causing some of the spikiness the deck meanwhile is a much flatter line there's a single large Spike to about 125 milliseconds for the deck which is visible in gameplay but the rest of the line remains remarkably and impressively flat Resident Evil 4 is next we did most of our testing at low settings since that's what these handhelds are generally capable of at 720p for Resident Evil 4 though we chose the prioritize Graphics preset again without any upscaling this is another instance where we also double check the settings manually and here the battery-powered Ally tied the steam deck at 26 FPS and 27 FPS respectively the plugged in Ally managed 36 FPS or 34 ahead of the steam deck the lows are much better in this particular game our estimation again is that the frame time consistency behavior is probably a mix of game characteristics in terms of the low load allocation and of the power gating whatever it may be the end result is observable and the Ally is tripping over itself in some games but not others the good news is that this doesn't appear to be indicative of performance in all scenarios the Ally did better overall here although tying a cheaper device still isn't a winning position we'd have to look at the quality of life features discussed elsewhere in the review to find the value the frame times for Resident Evil 4 are much better the Ally is still more variable frame to frame than the deck but we're not running into the massive stuttery spikes Red Dead Redemption 2 at 720p is our final game test using the Vulcan API and a favorite performance preset which again we then fix those settings to ensure they don't vary as usual the Allies wall power turbo mode was significantly ahead of the handheld performance mode 55 percent in this instance the Allies performance mode is nearly tied with the deck and frame times are actually okay on all devices here sometimes the Ally on battery power is better than the steam deck but usually it's tied it's clear that the Allies maxed out turbo mode puts it far ahead of the steam deck and performance but it's also not practical to run turbo mode without plugging the Ally in which defeats its purpose on that subject let's take a look at the battery life on all devices speakers were muted screen brightness was set to 50 and screen dimming was disabled sleep was also disabled unless otherwise noted note also that screen brightness at 50 is not a perfect control and we want you to be aware of that to completely control for this we need to take Lux measurements and then match the locks we're looking into this Improvement as a possibility for our testing in the future device performance settings were left at out of the box defaults unless otherwise noted we used our special patent-pending battery life spring tension tester something we developed in the lab to run these tests with constant user input fortunately this device is extremely hard for anyone to copy so we're okay with showing it battery life is measured from the beginning of the load to the point where the device automatically shuts down or hibernates we opted to mostly test at the native resolution of the individual devices we're not comparing frame rates here so it doesn't matter we're seeing how long the batteries last under normal use conditions we'll start with the simplest battery life test to Loop in one of our own 4K 60 videos at full screen and the native desktop version of VLC for each of our devices note that full screen in this instance is 1280x720 letterbox for the non-ally devices and 1920x1080 for the Ally the Ally outlasted the deck here by 67 percent at 216 minutes versus 129. the DAC audibly ramped fans and struggled for the duration of the test our Aya Neo next with 5800u and integrated video graphics actually outlasted both at 246 minutes and both the founder and the next are equipped with 47 watt hour batteries to the deck and allies 40 watt hours moving to games we first ran the test with f122 at medium presets with vsync disabled and variable rate shading disabled this is with the frame rate uncapped and we'll look at the capped results next f-122 easily exceeds 60fps with these settings so this represents a best case scenario for taking advantage of the Allies 120hz display the power plan was clearly the limiting factor on the Ally using the default all performance operating mode of a 720p and 1080p the Ally lasted 100 minutes while manually bumping the power plan up to Turbo at 1080p dropped battery life to 63 minutes remember turbo on battery power has lower power limits than turbine wall power 100 minutes in performance mode isn't as bad as it sounds the steam deck for example lasted almost exactly as long with vsync disabled and that lines up with our previous dmc5 testing from the decks launch let's take a moment to consider what those results mean using out of the box default settings the Ally and the steam deck are effectively tied in battery life for this test and they're usually but not always tied in terms of FPS under the same conditions and ignoring frame times this is an underwhelming result for the Ally but it does have more performance Headroom next we ran dead cells this is a 2d game with pixel art so we ran it at Native screen resolution with vsync enabled to represent a common handheld gaming scenario one which we already tested for the steamvax launch on the Ally vsync means 120 hertz unless settings are tweaked which is is likely why the Ally only lasted 167 minutes to the deck's 365 less than half the total play time even with a 60 hertz cap applied though the Ally only lasted 207 minutes if we need to cap the Ally to 60 hertz in order to reach half the deck's battery life it defeats a lot of the purpose of having that 120 hertz panel other than other differences like resolution time to talk charge time the Ally was the fastest of the handhelds to recharge from zero percent with the system turned off and charging for a full cycle Peak power draw was 53 Watts which dropped sharply after 30 minutes to about 30 Watts then gradually fell until the charge cycle completed at about 98 minutes in comparison the steam deck took 168 minutes to charge despite having the same nominal battery capacity there are two points we'd like to make here first plug in a higher wattage adapter doesn't always make these devices just charge faster with the systems off and doing nothing except charging none of them hit the max capacity of their Chargers or 45 Watts for the deck and 65 for the others the Allies adapter has a little little Headroom for charging and gaming at the same time a steam deck can run games at full performance while charging whereas gaming without charging already maxes out the Allies charger we didn't see any significant performance degradation from gaming while charging on the Ally but it may extend charging time of course let's look at Thermals and fan Behavior by proxy of looking at The Thermals this is done first with f-122 as a gaming load and then we'll look at blender in this chart you're seeing the torture scenario on the Ally it's plugged in running turbo mode and it's at 100 screen brightness so this is about a worst case scenario for it that hard rap to about 96 degrees Celsius that you're seeing is when it's waiting for the fans to respond Asus has stated it went with a less aggressive fan curve to try and temper the noise something that it talked about publicly closer to its launch and that's evident here we think that they should probably kick the fans into higher speeds sooner in this curve though the high temperature is from a delayed start on those fans ramping to the higher speeds which eventually do bring up the RPM and that brings the temperature to more reasonable levels the occasional spikes are during loading screens where the fan Behavior oscillates here's the chart at steady state the hottest set of tests we have at steady state once we're past the initial fan ramp where the Apu is hitting TJ Maxx would be the torture worth load hitting between 75 and 77 degrees Celsius on the Apu and SOC components and the drive here was in the 60s which is okay this test was with the screen Max at 100 and the charger plugged in so once again a worst case unplugging the charger drops turbo Mode's power consumption so it's no longer a pure like for like but the end result is similar we're at about 73 to 75 degrees Celsius adjusting only the brightness drops that by another few degrees Celsius and remember that changing the modes like performance and turbo or changing whether it's plugged in will possibly affect the fan speed depending on the configuration which affects the temperature too not just the power screen brightness here definitely has an impact on thermals but that makes sense so you can touch the screen and feel the difference it runs pretty hot as most screens do go into performance mode drops a massive 15 degrees or so so from the turbo 50 brightness result at 71.7 to 56.9 degrees Celsius that's a crazy drop and it aligns with the power reduction that performs claims moving to silent mode drops even further around another 10 degrees depending on which metric you look at and we've got Acoustics separately for all these modes as well of course thermally only the brief slams against the 96 degree celsius wall are concerning but those appear to only briefly pop up while waiting for the fans to really kick into gear Asus could adjust the hysteresis here or effectively a fan curve to tackle this and this could be done via bios so they could change it at any time if they wanted to ultimately though the temperature came back down rapidly and it didn't stay elevated in a way that would concern us about the CPU or GPU components and the SD card however is a separate matter we'll get to that later well the CPU torture workload using blender so nearly zero percent GPU and 100 CPU we can even more clearly see the fan Behavior reflected in The Thermals the CVT die reading was 95 degrees Celsius for nearly three and a half minutes ads that's far too slow of a fan ramp here because you're already losing clocks in that time Asus needs to rework the hysteresis we think it's going to get loud anyway but throttling for several minutes before getting loud or if not throttling just running this hot it's really not a good trade-off you can just accept a little more noise they also cool it to a temperature that at least for the Apu is unnecessary anyway Asus could meet in the middle they could improve both the initial load thermals and the steady state noise levels while maintaining an acceptable combination of both it's possible that some other component needs the fan aggression though without probing everything on the board we can't be certain if that's the case in order to test the micro SD they're all problem that's been so widely discussed we first needed to get a Micro SD card detected in the Ally which proved a lot more difficult than we thought it would be we tried a SanDisk Extreme Plus 128 gigabyte a SanDisk Extreme 512 gigabyte a PNY Premiere x256 and a SAM Amazon pro plus 256. we bought a couple different brands all of these are completely functional cards that show up on other devices as well as the Ally itself if connected through a USB card reader we format it as xfat we tried NTFS as recommended by the official Asus guide we tried fat FAT32 and we tried disabling and re-enabling the SD card reader in BIOS and Device Manager we also tried downloading and installing every possible driver or software and firmware update from the Asus website as well as Windows update my Asus and Armory crate we tried booting into the built-in my Asus in winreave diagnostic tool which has tasks for pretty much everything except for the SD card reader and we tried running a system repair so all of that that wasn't to test the problem that was to try and get an SD card working in the device at all micro SD in the top here and uh because we couldn't get one working we think we've come to the conclusion that our device has already been a victim of the micro SD card problem much of the secondary and user reporting has focused on the SD cards themselves but upon reviewing asus's official statement which it posted on Discord the wording was quote under certain thermal stress conditions the SD card reader May malfunction they also said quote if you are currently experiencing issues with your SD card reader well RSD card reader is malfunctioning and it has been since basically we got it we had thermocouples wired to the micro SD card reader internally and to some of the other chips around it but we never were able to make use of any of that because since we can't get a card to reliably detect to generate any kind of load to thermally stress it like transferring files they were just they were useless it was it's effectively the reader itself appears to be more or less dead on our device one thing we're wondering is whether this is an issue with the Genesis logic GL 9755 as the card reader controller in general not just thermals either way we have confirmed that there is a problem we just never got to test what may be causing it Patrick even checked under a microscope for visible damage to the SD card controller the SD card reader and we can't see any now to get any further here we'd have to send it out to an external lab for failure analysis which we might not do we'll see if it's worth it or not but what we can confirm is that yes this is a problem because the the micro SD card reader issue did occur for us Asus has acknowledged it as well so we know it's a real thing we just don't know what triggered it precisely So in theory Asus says some of its updates like the BIOS should prevent the SD card reader from malfunctioning it's very hard to prove a negative because we'd be trying to prove the it no longer fails aspect of it but to do any further testing we would need to buy another Ally install the updates and then use it and at some point draw a line and say okay it's probably fixed so for now what we can say is the issue did happen and Asus has said that it's been working on addressing it via its updates but we basically just have to all look at the community to see if more of these types of failures pop up so wrapping up we're not going to recap every single section here in the conclusion we're going to focus on more broad wide sweeping stuff so if you've jumped here and you're looking for micro SD card or noise or thermal any of that you're gonna have to go back for those specific things now at launch the Z1 extreme was the most powerful mainstream handheld gaming PC and there is always a market for most powerful thing in a category because you're not really fighting on value as much now the market is already shaping just a little over a month or two months later at this point and GPD is shipping wind for 2023s and Aya Neo has announced it's Quinn which both of these devices will compete very directly with the Ally handhelds obviously need a mix of being cool quiet and really a more important way than say a laptop sometimes and they run on tiny batteries so battery life matters in this way being the most powerful thing is maybe not as important as it would be for a desktop GPU because there's a lot more downside sides to it now it's possible to cap the Allies refresh rate and its power draw but at that point it starts to make sense to maybe buy a cheaper device that is built around those characteristics instead and the steam deck is super competitive right now because they just introduced the refurb program where you get them even cheaper if you're okay with something that's got maybe some wear and tear to it at this point the Ally non-xtreme Z1 at 600 without having tested it of course we don't think is worth considering because there's simply too many Alternatives we can look at the on paper specs versus say this one the Ally and the steam dags and we could draw that conclusion without needing it in hand to do so knowing everything we know about this device and it's uh it's competitors so the specs for the non-xtreme are massively decreased and it's beyond what a 14 discount really justifies this Ally is definitely still ahead of the steam deck specs wise but the deck competes based on value not just tours hour the cheapest deck is four hundred dollars it can be upgraded to double the Ally storage capacity for maybe another hundred dollars and seven hundred dollars for an Asus handheld is impressively low actually given asus's General value of its brand but valve's got a secret Advantage which is that it owns the damn store you're buying the games from and it can and probably has sold Hardware at a loss on top of that looking back at some of our gaming charts the Allies performance Advantage becomes much smaller when running on battery power if you want a handheld PC and intend to use it that way not just plug it in the deck is the better price to Performance option but the Ally has more Headroom to scale that performance up the Ally does have one at least clear advantage to us which is the quality of the screen where regardless of what you think of the refresh rate 120 hertz whether it's useful or not the actual image quality that you get is better we think on the Ally than on the deck if that matters to you maybe that swings it the other way but for now that's it for the testing of this device and we hope you enjoyed the review it was a massive amount of work and we really enjoyed working on it so we're looking forward to doing more of these we will get these a little bit shorter as we continue to iterate on the process um but for now this gives you at least the the sort of foundation for what we have coming up the GPD Win 4 is likely to be the next one if you didn't catch on uh but we've got some others we're looking at too thanks for watching please support us if you like this content and or found it useful because we need your help to be able to continue paying for the extremely long production times on stuff like this you can go to store.gamersnexis.net if you want to grab a useful item like one of our soldering or project mats our mod mats our shirts if you want something simpler and our tool kits are there as well subscribe for more thanks for watching we'll see you all next time

Channel: Gamers Nexus

Views: 459,545

Keywords: gamersnexus, gamers nexus, computer hardware, asus rog ally review, asus rog ally benchmarks, asus rog ally microsd card, asus rog ally sd card, asus rog ally vs steam deck, asus rog ally vs aya neo, asus rog ally vs nintendo switch, asus rog ally noise, asus rog ally battery life, asus rog ally thermals

Id: Na1y7DyDe2w

Length: 44min 44sec (2684 seconds)

Published: Sat Sep 09 2023