Lenovo Legion Go Handheld vs. ASUS ROG Ally & Steam Deck | Deep-Dive Review & Benchmarks

[Music] this is a weird one the Lenovo Legion go can do this and it can also do that but it also has some severely broken features so the legion go is a mix of cleverness like we just saw there and also unrefined and it has really strong latency in some of our testing but it has trouble with frame time pacing it also occasionally has obnoxious fan acoustic issues we bought this Legion go for our testing today it was $700 when we bought it although we did buy two for reasons we'll talk about later neither of those really compares to the thousands of dollars we've spent in testing time on it but that's a different story and the handheld PC market is really heating up you can check out our deck OLED or Ally reviews for some of the specifics on those but now we'll focus on something new the Lenovo Legion go before that this video is brought to you by Arctic and the liquid freezer 3 which we've tested using our standard methodology and found to be one of the most noise efficient CPU coolers on our charts the liquid freezer 3 includes ryen offset mounting to improve AMD performance and uses a thicker radiator to further lower the temperatures Arctic has priced these competitively for aios and you can learn more at the link in the description below there's a ton of room for creativity in these devices and they're still figuring out some of the basics so things like the ergonomics of the handheld itself they haven't been perfect Ed yet that means there's a ton of room for advancement with each different device that comes out we'll have a separate tear down for this as well going up pretty soon we couldn't fit it all in just one video so uh keep an eye out for that but today the focus will be on some subjective Impressions Hardware features uh also we talking about software and firmware bugs Acoustics gaming performance latency performance as it relates to gaming uh battery life thermals charge time all that stuff there's a lot to cover we first start with the spec sheet the ghost spec sheet also lists the much less powerful non-extreme Z1 as an option so get ready for some Ally style naming confusion when that releases but for the one we have the ghost memory is 16 GB of LP ddr5x the same capacity as asus's Ally handhelds but clocked higher to 7500 megatrans per second 16 GB is starting to feel cramped for gaming on Windows these days especially when a chunk of that is set aside for vram the battery capacity is also more comparable to the deck olet here's a recap of just a fraction of the handheld PC market right now the Ste deck OLED we just reviewed cost $550 to $650 on the channel if you want to see it the deck LCD is 350 with an emmc SSD to $550 the Asus Rog Ally not to be confused with the Rog Ally is $400 the r Ally meanwhile is $600 lenovo's Legion go meanwhile is $650 to $700 depending on how much storage capacity you buy aan Neo seemingly has an infinitely filling bag of holding overwhelmingly erupting with handhelds because we can't can't keep up with the amount they release they seem to really Excel though with custom colorways and we actually just ordered an Air1 s to look at something a little on the smaller side compared to these the most recent launch is the Quinn which was priced between 1130 bucks to $1,889 they're also soon releasing the flip DS for 700 and up they're a lot more on the market like the 300 and up Ayn Odin 2 with a Qualcomm Android solution also making a name for itself with colorways but maybe that's for another time and then of course there's the nint switch which won't be focusing on much because it's got a completely different gaming and software ecosystem but it is worth considering if you're looking to buy a handle right off the bat though the legion go looks ginormous compared to just about everything else but in actuality it's not that much larger than the deck the difference does matter though it is however a good perspective for the Nintendo switch's sizing and why the power is a little more limited the go is bigger than the deck but only top to bottom and only by 1.4 CM that 1.4 CM matters though especially since the steam deck is already notoriously large but it's not the whole reason that the go looks gigantic that impression is created by a combination of the Go's 8.8 in screen with narrow bezels to the steam deck 7 in and the Go's chunky detachable controller grips Lenovo claims a total weight of 854 G or pointlessly 640 G without the controllers we measured 863 G in our testing against calibrated weights close enough but that does make it the heaviest handheld that we've tested by far the previous record holder was the steam deck LCD which we measured at 676 G the detachable grips are definitely the most unique aspect of this so far though so for the Nintendo switch the use cases of the detachable joycons include being able to hand one to a friend so it's a fully discret usable controller for games that support multiplayer on the same device uh or other devices and additionally you can connect them into an interlocking component to turn it into one semide decent controller beyond that there's the option of docking the switch of course and turning it into functionally a living room console and using them remotely for that one or two player and you can hook them into accessories like the rain fit or the labo the only one of those the go can do out of the box is the first one using it as a standalone display with detached controllers and maybe the peripheral dock if you count FPS mode we'll cover that separately next the two Lenovo grips can't re reasonably be used as two fully discret controllers like they can on the switch we tried a lot and we just couldn't get them to consistently be usable as two inputs in instances where you could get them to work as two inputs which is technically a possibility by switching to dual D input mode and using Bluetooth you'll still only have two separate very unergonomic halves of a controller the games would have to be very simple or likely older to be playable at all with the buttons available which is fine as long as you're aware of that remember though that you can always just bring a separate controller and connect that through native Windows functionality for two complete controls and inputs or you could remap the two halves of the goons for emulator games or simpler titles but don't expect this to work like a switch because it doesn't and it will not only be frustrating but also have limited compatibility and to be clear Lenovo is not marketing on it in that way so that's fine but we want to make sure everybody who's thinking about buying this understands the limitations it'd be easy to get the wrong idea if you don't do a lot of research uh additionally it does not ship with any connecting grip to link them into a single separate controller if you were to want that it' have to be purchased separately there is still value in detachable controllers but we think it's important to objectively identify what that value is before buying into it because you should not think of it as an equivalent to this the FPS mode we mentioned is a really unique and clever implementation it entails flipping a switch on the bottom of the right controller sliding it off of the go popping it into the Teflon backed magnetic ring and pulling the thumb sticks off so that it doesn't stick into your palm despite other limitations it's a clever design and we're relatively fond of how it came together there's an optical sensor on the bottom of the controller turning it into a vertical mouse that looks like a joystick but as Lenovo repeatedly and dramatically warns every single time FPS mode is turned on it is not a joystick we found the control schem of using the right controller as a mouse and left as a joystick equipped controller worked surprisingly well in games like cyberpunk 2077 we were skeptical at first but it worked out in our opinion though the so-called Mouse is unpleasant to use in a non-gaming scenario like just on desktop it would work better if the controller snapped firmly into the base as it is it's free to wobble around the base itself is narrow enough that the whole Mouse can tip back and forth alternatively we've discovered another use case you can leave the controller attached to the go and Slide the entire thing around on your desk as a mouse lovo characterizes the go as a base module and two Legion true strike controllers which could imply some other controller options in the future from a ports perspective this has USB four type-c ports but the most important aspect is the fact that there are two of them the Ally and the deck variants only have one port which means if you want to charge and plug something in at the same time you need a dongle for that and there are limitations with that the Go's display runs at 144 htz but doesn't support free sync or other variable refresh rate methods the image is extremely crisp but we've already noted that even the Allies 1080p screen is at the point of diminishing returns for Fidelity in such a small panel held at arms length one big thing with this is the insane amount of inputs though so in addition to the normal buttons where we have a d-pad we've got two analog sticks there's two kind of option buttons up here you have the face buttons of course you've got two more option esque buttons down here there's also on the back and this is like completely ridiculous uh the fact that as we stated this turns into a mouse so it has an optical sensor it additionally has a scroll wheel so they're trying to accommodate every function on a mouse which adds a lot of buttons you've got the trigger and the bumper up here there's another button in here as a grip button and then uh of course we've got the rest that I just described plus a touch had on the other one oh actually there's there's also two buttons right there that I had not mentioned on the other one though we have even more there's two more grip buttons here there's again the bumper and the trigger so maybe an upside completely inan amount of buttons uh we did have accidental button presses but fortunately they're mapped in a way where it didn't tend to cause any problems time to get into testing it's a really complicated test Matrix this time so the Go's native display resolution is 2560 x600 that complicates things that is four times the pixel count of the steam deck and as a result of that uh because there are only three supported modes out of the box at 1280 by 800 1920 by 1200 and then the 2560 x600 Solutions we have to make some compromises previously we've used 1280 x 720 for our testing as the highest common resolution to ensure a fair and likeforlike comparison between devices uh it's scientific method for these tests we've used 1280 by 800 whenever possible and then in some additional testing we have 2560 x600 exclusively on the go and that is looked at less from a comparative standpoint because these can't do it and more of a can you use it at all or what is the usefulness what can you expect the performance profiles are critical to testing though need some explanation that's because on the go they need some improvement using the basic Lenovo user interface there are three main performance options there's thermal mode or TDP OS power mode and fan the fan curve setting is bound to the TDP setting and to easily change these the legion go bundles the settings into two customizable presets named performance and power saving when we first got our go it was set to power saving but the out of thee box values were different from the ones that we got by hitting restore default in other words our Legion go had outdated presets we dumped an unreasonable amount of time and effort into straightening this out like literally a week of testing and this eventually led to us just buying a second brand new off the shelf Legion go now we've doubled our investment in just the hardware the out of the box preset had a different name for this one because it was a different batch this one was called custom one but the settings were the same as they had been on the original the power saving and the performance profiles were set up correctly but the legion go doesn't auto swap between profiles when a charger is connected like the Ally so the out of the boox setting are very relevant here's a summary of our internal naming you can come back to this time stamp if the charts get confusing 1A is completely unmodified out of the box TDP is balanced for this OS power mode is performance 1B is the upto-date power saving profile TDP is quiet OS mode for power is efficiency for this one 2A isn't important this was the old performance profile on our original unit with settings completely maxed out which leads to actually terrible performance on battery power this won't show up on new units as far as we can tell but it's on the charts if you're curious and as a note the second unit that we bought just this week the last week of February uh it had November updates on it so it's been on the shelf for a little while so it's possibly you get something old enough 2B is the up-to-date performance profile TDP and Os power mode are both performance 2C is the legion go completely maxed out 30 W TDP with a wall charger connected which allows the Apu to operate at full power all other tests were done on battery power unless stated otherwise note that the alli's performance profile is not the same as the Go's performance profile the Ally has silent performance turbo 25 watts on battery and turbo 30 Watts on wall power the go has power saving and performance time to get into testing we'll start with latency this testing evaluates the total end end system latency which is the entire pipeline that we're showing you on the screen now we noticed a lot of people are still mixing up the input latency sometimes called input lag and endtoend system latency on the steam deck and the Ally subreddits in particular to be very clear we testing the totality of the entire pipeline you're looking at from input to Output that means the PC latency and the render queue and pipeline are included some of which is frame time dependent because this is a holistic look at a complete device and not an isolated component this is the best way to evaluate latency it factors in the screen and the input with the rendering pipeline testing is done with a minimum of 100 test passes per instance and if you want to learn more about all of this you can watch our Deep dive interview and discussion with an expert Linked In the description below we did our first set of latency tests using dite 2 as representative of a 3D game that's limited by the Z1 extreme chip the results broadly correlate with average frame rate the goh's power saving profile with restored defaults is at the bottom of the stack at an abysmal 186 milliseconds the original performance profile and restored default performance profile are above it effectively tied with each other at 148 and 152 milliseconds and the restored default performance 2B is the exception as it performed significantly better in terms of average frame rate but not system latency past that the stack returns to frame rate hierarchical ranking the original power saving profile is there and then the OLED then the Z1 extreme the go with its maxed out out of the box performance profile and wall power tops the chart at 106 milliseconds the power saving 1B results is bad enough that you'd notice the latency in the games for our next latency test we used a sample application from Nvidia that just flashes white when a button is pressed the render rate was intentionally capped at 120 Herz and if you take a look at a breakdown of system latency from the prior pipeline image by controlling both the software and the render rate we can control most of the central Chunk from game latency through render latency allowing us to compare peripheral latency and display latency in near isolation although the render rate was capped vsync was not enabled wasn't necessary we used this synthetic test so we could compare running the Go's controllers wired versus Wireless it doesn't represent gaming performance it's attempting to isolate for that wireless variable here's the chart Baseline is 27 milliseconds with wired controllers using the right side controller as our chosen input device which means disconnecting it and running it wirelessly ended up adding 9.3 milliseconds to the average latency that landed it at 36 6.4 milliseconds that's slightly more than the duration of one frame time at 120 FPS which would be 8.3 milliseconds we also ran this against the Rog Ally which is why we limited the render rate to the Allies maximum 120 HZ refresh rate rather than the Go's 144 Herz and again we're controlling the render rate to intentionally make it irrelevant we ran the Go's latency test in the restor default performance mode on battery power so the rough Ally equivalent is Turbo socalled 25 socalled y hot mode on battery power with these settings the Ali added 24.4 milliseconds of latency versus the go with wired controllers given the equal render rates and the same CPU and GPU responsibility for that 24.4 milliseconds can mostly be divided between the quality of latency from the input device and the display in other words the go has a superior latency when isolating against render rate and time to get into the gaming benchmarks for frame rate Resident Evil 4 is up at prioritized performance with FSR 2 quality the result with the ghost default out of the boox 1A power saving profile was 44 FPS average with it just below the deck oled's 46 FPS average and the Z1 xtrem's 47 FPS battery powered average that allows the deol a lead of 3.6% and since we were forced to use a slightly lower resolution on the Ally these three devices can be considered tied for performance Straight Out of the Box on battery power the real default power saving profile labeled 1B here significantly lowered performance bringing its average down to a complete complely unplayable 17 FPS if you manually switch to the power saving profile on a legion go these are the settings you'll get on the other end of the spectrum plugging the go into wall power with the maxed out 30 wat profile labeled 2C led to one of the highest averages on the chart effectively tied with the Alli Z1 extreme on wall power at 61 FPS that's a 38% uplift over the completely stock profile 1a's 44 FPS average using the normal performance profile on battery power led to a stable 58 FP PS average nearly as good as the Mac settings at 2560 x600 the goh's behavior remained the same the default power saving profile had terrible performance as did the bugged out original performance profile at this resolution and with these settings only the performance profiles both restored default and the original with wall power make the game playable but well below the screen's 144 HZ refresh rate we're not going to waste a bunch of time looking at unplayable settings suffice to say youd need to employ heavy FSR to make this work the higher pixel density and the resolution are more useful in non-gaming scenarios maybe like video playback or if you're going to do some kind of lightweight word processing or something next up is f123 a game we can rely on to achieve decent FPS numbers on almost any platform the original power saving mode average on the go was 77 FPS with the deck OLED only 6% ahead at 82 FPS average but with vastly better 1% and 0.1% lows frame pacing is problematic on several of these devices with the mixture of Windows and the am the driver for this game the battery powered Ally Z1 extreme is 9% ahead of the go 1A but taking into account the Go's 11% higher render resolution and therefore total pixel count as rendered they're about the same the deck leads the 1A by 6% ISO resolution in average FPS and More in lows and is objectively superior as a result restoring the go to the default power saving mode settings tanked performance again we would only use that profile with 2D games the legion goes maxed out results with wall power nears the Z1 extremes equivalent wall powerered result at 124 FPS average but even that was accompanied by unpleasant stuttering although to be fair that's true of every non deck result on this chart it's something AMD and Microsoft need to work out separately after this point we whittel down our 2560 x600 testing we selected settings with the goal of getting playable frame rates at 1280 x 800 initially for benchmarking comparisons so running 2560 x600 test at the same setting is generally beyond the goes capabilities the power saving profile with restore default ran so slowly here that we were forced to turn performance mode on during the loading screens now for cyberpunk we've dropped the wall powerered results now to focus exclusively on the battery powered results that really matter in cyberpunk 2077 at 1280 by 800 with the low preset and FSR 2.1 set to Quality the fully stock go with its original power saving profile ran at 36 FPS average ahead of the deck OLED by 10% but again with significantly worse lows the higher average and lower lows is due to frame pacing if they can even it out that average might come down a little bit but the lows will also improve in dlight 2 the established pattern continues both the Ally and the go have unfavorable lows in comparison to the dec OLED while in terms of overall averages the go with its restored performance profile leads the more appropriate 1A profile close to battery parody with the Z1 extreme and later testing shows the Z1 Extreme has a lead of 11% that aligns again with the pixel count difference so normalizing for that they're roughly the same the deck OLED holds a legitimate lead over the go 1A by 6.5% but is experientially Superior in the lows Red Dead Redemption 2 had more stable frame times across the nonv valed devices so that's good to see it's not a universal problem the frame pacing but it pops up in some of these games this one is fine though since we didn't test with wall power the highest average frame rate was on the go with performance profile restored to defaults selecting the boosted power profile clearly makes frame rates higher but we'll have to examine battery life as well to get the full picture of how these devices compare the deck OLED leads the 1A result by 8.7% with the 2B result leading the OLED by 18% finally in balers Gate 3 we continue to see low performance cross allall devices the deed and the 1 a go are about the same the Ally Z1 extreme stands out here as having better 1% 1% lows but we've seen similar behavior in previous tests where the Ally has some good passes and some that are just as bad as the rest this is effectively luck of the draw you can't read too much into that set of lows now we're getting into Acoustics this testing was done in our newly built Hemi anicake chamber which we debuted last year this was a multi-week construction project that cost us a quarter of a million dollars and it was only possible thanks to your support on the GN store visit store. Gamers nexus.net and consider grabbing one of our brand new Ultra comfortable warm tear down style hoodies that we just launched they're in stock and shipping now and use a highquality tri blend of materials water-based inks and a soft interior this video alone cost us thousands of dollars in testing time not even counting the video production time and we need your help to keep making them so head over to the store or throw us a few bucks on patreon.com Gamers Nexus here's the testing we'll start with noise levels the quietest device on this chart Remains the Nintendo switch which operates only slightly above our noise floor and not enough to be detectably different so it's effectively silent the legion Go's first appearances with 1B which is the restor defaults using quiet mode for tdb control it makes sense the Ally in handheld in silent mode is about the same noise level under load as the go objectively here and 2B appears next using performance TDP mode that has the noise levels at 19.5 DBA are about the same as the steam Deck with a 60 FPS cap and it's loud as cycle and Under full load the outof boox power saving configuration is clearly the worst for the go other than the 100% fan speed entry Lenovo needs to work on its outof boox setup the noise was 22 dbaa here is just under the uncapped deck and it shouldn't be set up this way finally the 100% fan cycle is similar to the deck when uncapped and in its high ramp cycle frequency spectrum is up now looking first only at the legion go frequency spectrum the 100% fan speed test is unbearably loud to the extent that it becomes functionally unusable in any public setting unless you're okay with being that guy or without headphones testing was particularly loud in the range of 2200 to 3,000 Hertz which is definitely in the annoying end of the spectrum and the performance mode tests both the restored and the uh not restored were significantly quieter in this frequency range and had their biggest spikes around the 600 to 900 HZ range with another at 5,000 to 6500 Herz the power saving mode tested out of the box was actually louder than one of our performance mode tests and this was repeatable and is just a result of a bad fan configuration by Lenovo and here's the competitive chart the Ally plots first with generally louder aspects of the fan curve in the middle of the chart around 1200 HZ and again around 3,000 HZ the de OLED with uncapped FPS as next and it's significantly louder particularly around the 1500 to 2,000 HZ Mark and then again at 800 to 1200 it also has another Spike around 3, 3500 HZ the capped deck LCD is next with a significantly quieter fan as a result of the FPS cap the curve is also much more evenly distributed finally for the legion go the power 71b profile is a mix of the deck capped and the Ally behavior for the frequency with 100% being clearly louder for everything now for a noise comparison of these this is important test testing is done in our chamber with a high Precision microphone 20 in away from the device the noises will sound louder to you in your headphones or speakers then perhaps in real use that's because obviously we don't control the whole pipeline here we don't know what you're using for your speakers your audio device and we have to control the audio by boosting it to get it to a similar volume to my voice to make sure you can hear it on everything from a weaker phone speaker uh to powerful headphon so for this demonstration what we're trying to show is the types of noise the volume of them as we play it through the speakers or the headphones will be about equal you'll need the DBA plots for that comparison but for this listen to the types of noise and decide which annoys you the most it's up to you what you like the best however the deck definitely needs work on the fan curve overall for battery life testing we start off with dead cells using vsync and using the native resolution this is one of the few types of games that is playable at a higher resolution on these devices so we've left that as a variable because it's potentially one of the advantages and it's worth knowing holistically how much it affects device battery life the deck olad remains K here with an impressive 8.3 hour battery life from its larger battery and its modest resolution the deck LCD is next also impressive and after that is the legion go with its low performing buggy and stuttery power saving mode we wouldn't necessarily count this one as viable in most instances the Rog Ally is next at 3.5 hours with with its 120 HZ vsync configuration down at 2.8 hours the difference between those two demonstrates that an FPS cap would lengthen battery life even though the restored power saving profile led to abysmal performance in all of our game tests it only gained 16% battery life for the Go versus the original power saving profile which had consistently better performance the original power saving profile is therefore overall impressive with its higher resolution display not robbing it of performance and allowing it to run near the Ally which has a lower res resolution here's a 3D game we didn't use any frame rate caps for f123 but none of the devices can really render this Benchmark significantly above their respective refresh rates anyway and that's also part of the test resolution is equalized the restored power saving profile did last significantly longer than the others in this one it's at 67% ahead of the original 1A profile but remember that it also brings the average frame rate in this test down from 77 to 30 FPS with the low struggling The Deco Leed the Alli Z1 extreme with performance and the go using its original outof box power sa profile each had roughly similar average frame rates in this test and as a result of that the go comes out looking strong with a battery life improvement over the Ally of 40% with 1 a at 2.1 to 1.5 or 133% with 2B it appears to have some advantages in the battery life department and the deck LCD also gets left behind a bit here as for the controllers they last an extremely long time we'd need to do a ton of testing to account for all the various combinations of Rumble which would wear it down maybe faster inputs and FPS mode and we didn't do that what we did do was try to open a game and strap the joysticks down until the controllers died but after 4 hours the controllers had barely dipped below 80% of continuous non-stop input from just the joysticks Lenovo lists the capacity of the controllers at 900 Mah hours each and they claim 11 hours of battery life it's rare that you would need to worry about them time to get into charge rate the go has a 49.2 WTH battery about the same as the 50 50 wat hours on the deck olet the Ally and the deck LCD both have 40w hour batteries even with the extra capacity the go is the fastest charging device we tested by significant margin the results shown here were measured with the devices fully shut down and the go pulled 67.8 Watts at the wall for 35 minutes with the charger rated at 65 Watts it then dropped off sharply and completed charging that's in contrast to the decks which spent a significant portion of their charging times on the last few percentage points of charge we showed that in the deck OLED if you want a bunch of charts about that in total the go took 82 minutes to reach 100% charge while the OLED deck took 205 for Lenovo its result is impressively close to the 80 minute 0 to 100% charge speed claimed on the GH spec sheet remember however that charging too fast can be damaging to a battery's long-term endurance we currently don't have the expertise to judge if this is too fast or not however Lenovo has charge speed controls in BIOS if this is something you're concerned about or if you leave it plugged in all the time we didn't experiment with testing these for the this review this chart shows the percentage charged and is done while the device is on and idle unlike the last one having the go on at all even idling significantly increases its total charge time for example the spec sheet says that the go should hit 70% charge after 30 minutes but with the device on and logging it takes a full hour this is the only handheld we've tested that maxes out its AC adapter through charging alone and so this Behavior shows more notably each of the goh's profiles reached 100% charge within 100 110 minutes as opposed to the 82 minutes with the device powered off our main takeaway from this chart is that the go charges unusually quickly under any circumstances for example it can reach 80% charge after 67 minutes even if it's on and idling during the charging process we don't know whether that will impact battery longevity without longer term testing Lenovo does advertise however that the battery is to cell on the spec sheet which may help explain the speed finally some thermal tests for the legion go to look at the temperature remember as well that there are few other variables like power and noise but looking at the torture testing while plugged in and at 100% brightness with performance TDP had the legion goh CPU temperature at 76° C in a 20c ambient the noise levels were about 20 DBA here the closest comparison would be the deck LCD at a louder 24.1 DBA which was 78° C it's roughly the same considering that we don't have the same software for the steam decks as in Linux the go and performance TDP also had a 75° GPU and 74° drive the deck has some significant Drive temperature advantages however there are two factors to consider here one is that it's different measurement software in Linux and two is that it could be exposing a different Drive sensor there are multiple the Ally is completely like for like with softare though so for that one it ran at 77° T di 76 GPU and 75 for the drive it's about the same as the performance TDP Legion which is in theory a similar power level in total the CPU package power measurement is not perfect it's not the full picture but it's kind of the best thing we have without wiring in and soldering a bunch of stuff to the board Peak temperature for the go was about 86° and held for about a 10-second period finally the legion Go's case has a hole in it over the charging port so our first thought was that this is a useful feature our second was that we didn't want to blast 65 watts into a lithium battery wrapped in insulating foam and fabric we attached thermal probes to the device on the top and the bottom of it and the device started slightly below the ambient temperature of 22° C at the time of testing and we saw a peak external temperature is just over 37° that's a significant amount of warming and higher ambient within the case but it's not a worrying Peak temperature especially given that the temperature leveled off as the charging rate dropped since the go charges so quickly when it's powered off we think it's unlikely that the hottest period of charging will last more than an hour uh however we would not recommend putting an already hot go into its case and then immediately charging it especially in warmer areas but this seems like less of an issue than we'd feared as we get deeper into this generation of handhelds it's clear that something has changed they are starting to experiment and do different things and because the manufacturers haven't even perfected once again the basics like the ergonomics that's exciting that means there's a ton of room for Innovation and it's not just a bunch of copycats although there are still plenty of them the different companies are trying to innovate in different ways and we like that so uh the core Hardware itself is starting to get syy in some ways so the legion go and the Ally use the same Chip and A lot of these other handhelds use the basically a Z1 extreme 7 84u and soon the 8840 is coming out too the older custom Apu powering valves steam deck and the Intel Ultra 7 155h coming out in the MSI claw are the exceptions here we'll look at that soon too but the Z1 based handhelds are common and it's the specific features between them that matter with the 78 40u as well so the experience and those specialized features are what set them apart most of the devices we've reviewed so far are relatively large but compared against each other they're a similar form factor that's why we bought some smaller ones to review soon that's separate for gaming the go and the Allies share the same weakness which is that frame time pacing remains weaker for the drivers in combination with Windows in some games this is familiar to the 2009 era of AMD gpus where pure average FPS throughput took priority over pacing we'd imagine that it'll get fixed as more people talk about it and it's not in every game so that's the most important thing it's sometimes the gaming results stack of this way in first place it's the Alli Z1 extreme with its 30 wat turbo profile that's only accessible on wall power which partially defeats the purpose of a handheld and second place it's the go with its out of the box 30 wat performance profile that only functions correctly on wall power and therefore has the same issue following that there's the Go's performance profile with restored defaults on battery power meanwhile the Allies battery powered performance profile that goes out of box power saing profile and the decoled are on approximately equal footing in most cases albe it with more stable frame times on the deck deck these represent the unaltered out-of thebox performance for each of these devices focusing on the allly the go first so the Z1 extreme is used in both the legion go and the Rog allies Z1 Extreme as a result of that the primary differentiating factor between these two comes down to actually Asus and Lenovo not just AMD this time this comes down to their execution of different Power profiles and the execution of the firmware that drives everything uh and as a result of that right now we can say that Asus in that regard has done a better job than Lenovo Lenovo has a lot more ground to make up here they need to push more updates the Ally for example can automatically switch between profiles based on if a charger is detected Lenovo absolutely needs to do work on profiles and firmware fortunately these are things that can ship to existing owners if they patch them they also need to do some work on the fan tuning of the hesis as we discussed in the Acoustics section uh however for latency the legion go is actually a leader here it's doing well most of the time it depends which profile you use but especially when isolating against PC latency to analyze the controls in the screen the legion is a leader in latency the screen size of the Go's panel is more beneficial than its resolution the Allies touchcreen feels tiny and hard to navigate in comparison to the go while the Go's larger screen and sturdy kickstand make it practical to stick it on a desk and use it at a distance which may be a plus side for you especially this is true with the detachable controllers for the legion go most of the specific features like the relatively large screen the sturdy kickstand the attachable controllers are geared towards sticking it on a desk and as a display and using the controllers wirelessly if you only care about handheld use then the Alli Z1 extreme is normally a cheaper option and a smaller one although that cheaper part is in flux right now so uh as we're writing all of this it's been plus or minus 50 bucks between the two so that part will change you'll have to check for yourself when you watch this but uh the Armory crate as much as we dislike it in general it seems to work fine on the Ally as a handheld device and if you really want basically Steam OS features you can turn on Steam Big Picture Mode and windows becomes kind of invisible to you but it's still there doesn't fix everything though against the go though the Asus software is more refined than Lenovo and we'd primarily recommend the go over the Ally if you already know that you want to detach the controllers that'd be even more compelling if each detachable controller could natively and instantly act as its own full controller for multiplayer on one device like on the switch but it doesn't right now and we wouldn't recommend counting on FPS mode to be a game changer but we did like how it was executed and thought it was clever each device has its place currently you can find that place in most of our charts the steam deck is the one we haven't discussed much yet so uh Steam OS is its biggest advantage that it has where it has mastered the frame time pacing in places where Windows has kind of dropped the ball in conjunction with the driver set from AMD the interface is also good for Steam OS again Steam Big Picture Mode's available to kind of soak some of that difference but it can't fix the frame time pacing uh and that's going to be it for this one so we we do think each of these has different strengths we've uh spent a huge amount of time defining them thank you for watching you can subscribe for more check back for the tear down of the Legion go and additional thoughts on it and otherwise go to store. Gamers access.net we spent thousands of dollars testing this one device for this review uh not counting the cost of the device itself so grab one of our hoodies we just launched these you can also get a soldering mat a mod mat for your PC builds or one of our other things like our coaster packs and our t-shirts it directly helps fund the work we do here and thanks for watching we'll see you all next time

Channel: Gamers Nexus

Views: 374,068

Keywords: gamersnexus, gamers nexus, computer hardware, lenovo legion go review, lenovo legion go benchmarks, legion go review, legion go vs steam deck, legion go vs rog ally, asus rog ally, asus rog ally z1 extreme, steam deck lcd, steam deck oled, steam deck oled review, best handheld gaming pc, handheld gaming pc benchmarks

Id: I37mxlDqLrk

Length: 38min 36sec (2316 seconds)

Published: Sun Mar 03 2024