Here's some clarification from the video: there was an oversite during editing omitting the quality settings used in games. I'll provide the picture in the community tab, but in text form: Medium: Warframe, Valheim, Shadow of the Tomb Raider, Quake Champions, F1 2020, Doom Eternal, Crysis Remastered, Chernobylite, Apex Legends (mostly Medium, low when there was no medium option) Low: Sons of the Forest, Returnal, RDR2, Hogwarts Legacy, Forza Horizon 5, Cyberpunk, Control, Borderlands 3 High: Path of Exile, Portal 2, Fallout 4, DayZ, Custom: No Man's Sky (Standard), Metro (Very High), Horizon Zero Dawn (Original), GTA5 (PS4 settings), CS:GO (prosettings.net recommended), As for UMA - CryoUtilities recommends a 4GB UMA Buffer, so that is what is used.
It's not a silver bullet for every game, but it's free performance and the gains to 1% lows are quite dramatic for certain titles. Thanks for the independent analysis, been waiting for more creators to provide their perspective on the tool.
I have to disagree with you on one thing. The 1% lows. Those minor improvements are incredibly noticeable. The more stable your frame rates the smoother the gameplay feels even at relatively low frame rates. I cannot stress enough how much of a difference a stable frame rate makes in gameplay. Honestly it can actually have more of an effect than max frame rates.
Cryoutilities turned Midnight Suns from a crashy mess into a playable game for me. It might not help every game, but it's worth keeping it in the toolbox.
Interesting - I've been playing MS without CryoUtilities (but *with* the 4Gb VRAM tweak) and it's perfectly playable at the default settings. I've had one crash in about 40 hours of gameplay.
@@Ashpolt I wish I had steamdeck. But I have to sell my old laptop b4 I buy steamdeck (money issue :D) I feel like in old days when I read about game Ill get in 2 weeks or so(now I watch yt videos about deck :D)
Good video! For the Steam Deck however I'd add a 40fps threshold (might not change your conclusion however). Doesn't make a difference docked of course, but because you can change the refresh rate on the Steam Deck screen, a lot of people shoot for 40fps at 40Hrz. As a stable 40fps on a 40Hrz display is a noticeable better experience over 30fps, but easier to hit than a full 60fps. If I was getting mid 30s and the utility bumped it up to over 40, I'd consider that worth it on a lot of games.
i feel like reaching slightly higher fps is a bigger deal than people think, cause then lowering the hz to 40, will result in stable 40hz vrs more unstable, plus then more preformance for the battery life, but idk
I totally agree that using thresholds is a way more useful metric ...except, in the case of 1% FPS. A small jump of 24-to-27 doesn't simply mean a higher average in this case, but it implies that there are _fewer hitches_. It's *not* that the spikes are just-as-frequent but less severe. They are just as noticeable, but can be _much_ less common. I've been playing Persona 5 Royal nowadays, and the improvement is incredible. The average was already a solid 60 in many scenes, but it suffered from a great deal of spikes. Even in places you wouldn't expect. CryoUtilities improved on it so, _so_ much.
Oh, and if there's a performance issue even with CryoUtilities: I believe the main source of issues is related to [in Persona 5 Royal] NPC Crowd Density, that's where the bottleneck begins. You can confirm that on the Main Menus and the Train Station. Everything else is silky smooth 60fps.
I've had CryoUtilities fix more games than break. In fact, none that I've tried have been worse for it (though it sounds like Halo Infinite may be one case). Without it, God of War used to crash and Midnight Suns wasn't playable. It also seems to help some other very demanding AAA games have fewer dips.
The difference in M&B Bannerlord is absolute night and day, very noticable and not at all insignificant. I stresstested it on med/high settings, with the troop count maxed out (!) at 1000 and started a custom battle. It was somehow stable at 50fps, dropping down to no lower than 30fps at the absolute peak of my stresstest. I originally had these numbers when playing captain pvp at much lower troopcounts, i also noticed everything loaded in MUCH quicker, even the server browser went from taking +/- 10 seconds to refresh to 1 or 2.
Based on these tests, YES! It deserves the Hype. Come on! A free tool that can noticeable increase performance on 50% of your library. I use on a non-steam version of "Bioshock Infinite" and God of War... and the difference on severe stuttering are night and day.
@@TheTerk12:03 can you please help so the colistco protocol returnal and a plague tale requirem does that apply to these newer harder to run bad PC ports
This is a Year later after the tool came out and I am using the Steam Deck OLED. I fully understand all of the points brought up in the video and have no disagreements. I also belive that the performance in the original video was slightly over exatrurated but, coming from a background in IT i think the utility is more geared towards further optimization with the way the Kernel uses the storage. I have no where near the amount of games that you have but i have very few AAA titles and i see the same exact numbers and reading that you do in your testing. But with my other games that are more indie titles the performance that was stated in the video announcing the utility was way more on par. Since these games are not nearly as optimized as the AAA games are. Wonderful video and like i said fully agree with all of the points but i do think it is a tool that should be used and not ignored since it does optimize the decks OS far more than steam does now.
Imo you should add a line for 40fps which is a fantastic experience on the steam deck after changing the refresh rate to 40hz. Otherwise fantastic video, I'd like to see you compare TDP performance this extensively. I did a little testing with rdr2 and it showed that above 11w TDP there was less than 1fps gained for each extra watt added.
@@OrbitalShell steam deck allows you to adjust the screens refresh rate all the way down to 40. If a game can’t maintain 60 FPS, it’s a great option and way better than 30.
I pay my deck at 40hz & 40fps a bit better frame stability, battery. Only game I say I can’t play at 40fps is overwatch which I have to have it at 60fps but I barely play it because the controls just dont feel great in that game
One thing you forgot to look at is how many games managed to go from averages/1 percent lows below 30 to 40 fps. As the 'golden 40' is a well loved threshold to cross for Deck users. I agree that you will be running at a set framerate. So going from 70 to 74 fps doesn't matter if you're capping at 60 fps anyway. So what I'd like to know is if the games that gain the most from CryoUtils get better battery life. Because you'd imagine that if something can go from 49 to 59 fps and you run it at 30 fps that extra headroom means that the Deck needs to work less. Might be another 15 min or 30 min of battery life.
I've found the utility more useful in games with extremely dense and frequent asset streaming, like Dead Space. These games tend to lean more on the storage and various types of memory within the system, as well as the interconnects between CPU, GPU and SSD - in other words, overall system latency is a major factor in these scenarios and small efficiency tweaks in memory handling can shine there. You'll see games like this more commonly as the Xbox One and PS4 reach end of life. I've also seen gains of 5-10 average FPS in Breath of the Wild for reasons I haven't been able to puzzle out.
This video is an unnecessarily adversarial perspective on what the CryoUtilities provide and does not do justice to the testing methodology that CryoByte33 outlines in his various testing videos. You didn't even include game settings as a parameter in your outlined testing methodology, where CryoUtils will benefit different settings at different amounts. Big thumbs down on this video, you should do better.
Trust me, this was the tamed version. As for the settings, I'll pin that comment and post the picture to the community tab. That was an oversight during editing, I put in the wrong chart
All I can say is that CryoUtilities saved my Steam Deck from an RMA. It fixed the crashing on God of War that made me think my Deck was defective. The crashing was caused by a memory leak that CryoUtilities fixed.
@TheTerk Did you change your vram to 4GB in BIOS? It is necessary, as the developer advised. Tthis piece of software is a WIN WIN for everyone using it. You can revert settings to stock within a few clicks if it negativelly affects a specific game. Also, think about it. There are NO negatives. Only Positives... if theres a negative, revert the settings... BOOM.
“Can you notice a 43 to 48 fps difference” actually, yes because when you set your screen frequency to 45hz, if it’s above 45 constantly it looks perfectly smooth, where as 41 looks a bit choppy, I think you have to consider people like me LOL
You missed some of the vital games that enable it playable because of cryo. these are the ff but not all.. Marvel Midnight Suns, WWE 2k22, Emulation specifically switch games.
@@TheTerk I think this video was better this way, since you prioritized neither games it fixes nor games it breaks, it feels like a randomized controlled trial.
It's nice what you analyzed, but the creator of the tool is doing a lot of hard work, for free and it's no oversell. It's someone that is telling you every step of the way what he is doing.
CryoUtilities, Overclocking, undervolting and increasing the TDP over stock together makes a big difference. I have my deck running 4000ghz CPU, 2000ghz GPU and TPD of 22. also undervoted -.40 on both cpu and gpu.
I did this exact setup and it throttles, cpu gets way too hot @ 101-104 degrees Celsius. Had to lower TDP from 22 to 17 to keep it under control in mid 90s. Frequency drops to 3.8ish and 1.7 respectively. The SD needs a thermal solution to run at 4 & 2 reliably all day everyday.
@@sourcey6620 first you have to make sure you have the back plate fan mod, which is putting holes in your back cover between the valve logo and the 2 screw holes. this will help reduce temps by about 7-8C by itself. second you want to make sure you undervolt which will also help lower the temps as its not pushing as much power at the requested speeds. Third you have to makes sure that the frequencies for both are not set to run at peak all the time but rather set the max over ride instead. it will kind of try to do a juggling act to stay at target tdp while not locking at that frequency. Forth you have to make sure you adjust your Tdp control from with in the bios set 35000 for TDP and then both fast and slow ppt limit to 25000. then you can adjust with in powertools with out it ramping out of control. if you have any other questions just let me know. Maybe i should make a video or 2 about it at some point saying virtually no one has yet.. not sure why.
I like it especially for reducing frame dips on games like Kingdom Come Deliverance. It was nearly unplayable in the cities. This helped me a loooot to smooth out the experience. Tried out Cyberpunk today. Here it’s similar. Less frame dips. That’s all I’ve ever wanted. 🤩
I went in REvil2 from medium low at 25-30 fps to almost max at 60 fps stable after I installed cryoutilities also my wii and GCube games run stable at their target fps it does makes a very big difference
Great job explaining what cryo already explained in his video. Cool comparisons, but it just confirmed its better have it than not have it…. Oh unless you like halo…….LOL
@@TheTerk Odyssey ran terrible for me and honestly I was quite shocked and dissapointed at that first moment. With cryoutils it ran like a dream, 60fps with a drop to 55 on the Lost Kingdom. BotW, I changed to the WiiU version which runs steady 30fps (also easy to setup the gyro) npc distance on 1.25, grass distance on 1.25 and more-grass-mod (can remember the exact name). Kirby and the forgottenland got a 5 to 10 fps boost for me. Same for Mario Kart 8 and SSBU. The last two already did oke when capping GPU to 1000mhz but with CryoUtils they both have notabele less stutters and fps is more steady (specially in SSBU that is a life saver;). I have Dread but not played it yet, does not seem like a demanding game. Prime Remastered runs great but I've never tried it without CryoUtils.
What abt stutters and micro stutters, and those random deeps in FPS, CryoUtilities fixes most of them, so not all the time is abt FPS, it's abt smooth gameplay, and better CPU and GPU usage, not to forget Temps and better SSD life, all I got is better gameplay with those setting. The UMA doesn't work all the time like Assassins crees valhalla i had to put down to 1gb now Steam deck uses Ram And Vram better, but Cryo did an excellent job 👏🏻
Gonna be honest with you. Kind of felt like you had your mind made up going in. At the end of the day a free net gain is a free net gain. Free higher lows are free higher lows. Demonstrating that different games see different levels of improvement is rather obvious. Not sure why that point needed to be made. I will say that the info was presented well. I just got the feeling like you were setting up for a “gotcha moment” from the beginning. No one program is a “miracle fix”.
In my intro, I stated my previous experience with the tool and wanted to validate various users' anecdotal evidence. The point to be made is the tool doesn't make massive improvements, isn't the ultimate tool, and is beyond over-hyped. If you read the subreddit, CryoUtilities is sold as a "miracle fix" across the board and is not necessarily true. A free net gain of 0.1 sure is free, but is it really a gain? Feel free to interpret the data how you want, I collected the data and came to my own conclusions.
I appreciate 1% low improvement, but the quantity of games showing "massive" gains doesn't support the claims. Even then, going from 19 to 24 fps in 1% low isn't game changing. Same goes for 0.1%
so Terk you seem rather intelligent but the main thing causing an improvement from his tool is turning down swappiness. swappiness by default is set to 100% which makes the device use the swap file constantly while his tool turns it down to 1% (its not a new tweak by any means) making the system use real ram far more than using the swap. the rest is mostly placebos especially the swap file size
Ehh, I've found, at least on Windows and in Hogwarts Legacy, that Swap space of 2G or less is not enough. The game (on my A1) would crash in Hogsmeade. Changing swap size to 8GB fixed that, again with windows and a different device. On Steam Deck and SteamOS, swap size needs to be increased a bit, but the game doesn't crash due to OS overhead being slightly lower than Windows. By default, swappiness is 60 in ARCH, so Valve did change that.
@@TheTerk from my testing on steamos the swap is barely even used (even at 1% it would use the swap much more if needed) because the 15.2GB I have available is enough for everything I've tested. even if I did need a swap (I actually have mine disabled with no issue) though it is much slower than ram so while it can allow games to run that otherwise couldn't its too slow to actually improve performance.
@@deathblade200 I think the fact that the consequences for when it's outside "everything you tested" are reason enough to just increase it. God of war straight up crashes periodically, and increasing swap fixes that.
@@iurigrang I played God of War perfectly fine without a swap file and 256MB vram. guess I need to specify next time that I tried some of the most demanding games which also includes games such as Cyberpunk and even Elden Ring
Did you increase the V-Ram in the BIOS before the tests? As the creator of the tool explicitly says to do that in tandem with the recommended settings of the tool.
Honestly? Going by your own methodology, it fixes as many games as it breaks (though that's not entirely true if we look at the comment section), it boosts average performance by a statistically noticeable degree in 7 out of 27 games (much more than 5% in some rarer cases), and affects only 1 negatively (again, if we look only at -5% or more, to ignore negligible gains and losses). It also boosts 1% lows by 5% or more (sometimes much, much more) in 12 games, and lowers performance in the same regard in only 2... with one of them being a game the tool's creator highlighted as problematic, so I think we can safely cross that one out, as changing the VRAM settings would yield different results (why even include that one?). Whether something is noticeable or not is purely subjective, especially since the Steam Deck often rides a tight line between playable and unplayable in new AAA games. But looking at raw data - your own raw data - I'd say this definitely looks like a must-have tool, especially at the pricepoint of... well, free. Sure, "massive performance boosts" is probably an unncecessary overstatement (though it's often tight line, as I've said), but it's definitely a beneficial piece of software. If this was a new driver update for a desktop GPU, 99% of people would take these gains in a heartbeat. Long story short, if you wanted to prove Cryo wrong, I don't think your own data matches your conclusion. Because of this, and considering the slightly adversarial tone, it's a bit hard not to treat this video as pointless bait, no matter how hard and time-consuming it probably was to make.
I think you're getting offended for the wrong reasons. You're 3rd paragraph sounds a lot like the conclusion of the video - CU2.0 highlights issues with SteamOS that SHOULD be fixed in the OS. This video isn't meant to "prove Cryo wrong," and I never stated or claimed that (0:41). The video was to prove that the hype behind it is overly exaggerated, which you agree with. As for subjectivity, that's why I presented the data and my opinions about it. You're free to disagree if a 2FPS increase in 1% lows in Control is not noticeable [reported as 5.4% improvement]. And you're free to say that a 7.4 FPS improvement in 1% lows in Portal 2 (which gets above 200FPS on average [reported 9.4% improvement]) is significant. I disagree that this tool is a "must-have," but if someone is having issues with performance, it's definitely worth a try (just like trying other Proton versions, etc). As for CU2.0 hurting performance, you'll notice I didn't mention that at all in the video. The only time I do mention it is Doom Eternal, and I don't even really call it a bad thing, the game benefits from 1% lows and is still well above a playable threshold. Thanks for the comment.
@@TheTerk I'm not getting offended at all. I just shared the overall impression your video made on me together with some of your replies in the comments section. I'm not going to argue whether an FPS increase from 41 to 43 is noticeable, but is it useful and preferable over stock settings in a game with 1% lows hanging just above the valued "40" threshold? Definitely. And considering there are several games with much more noticeable improvements, and the "hanging just above 30/40" or "nearly at 30/40" scenarios are going to happen often on the Steam Deck, I'd say a free tool that helps you squeeze these additional FPS is definitely a must-have if you want to play AAA+ games on your SD in 2023. But alas, no point going in circles. Have a good day.
A few things to note is that CryoUtilities doesn't cost anything, it's easy to set up and doesn't take too long _(About 5-10 minutes depending how fast you are),_ you only have to do it once so it'd stay there regardless of whether you shutdown or update your Deck and its changes can easily be undone, so theirs barely any downs sides to actually wanting to try it unless if you don't feel the need to or just aren't bothered. Also improving the 1% lows is a big thing no matter how minimal the difference seems as you could be rocking 60fps but having your lows being dog water makes for your experience to feel awful. CryoUtilities also fixes some stuttering issues in games, so even though that isn't a direct boost in fps it is great for performance in general! And a bit of icing on the cake, CryoUtilities also makes it pretty easy for you to delete unused game data from already deleted games in bulk or individually for just that nice bit of quality of life yknow?
My experience with this tool, I just got Mass Effect legendary edition and it kept crashing. As soon as I got Cryoutilities, it started working perfectly.
Obviously running RDR2 at 4 gig vram which throws off testing there. Also of note the tool can be used in conjunction with game settings to get you over the hump to 30, 40, or 60 fps average.
As expected, the only change that really matters is a swap size increase. Reddit performance claims should not be taken at face value. There aren't any dramatic changes in the cryoutils that would bring improved performance
@@TheTerk I believe FS22, I have the Steam deck, amd I have a bit of experience with linux knowledge (maintaining some dedicated webservers). First things I do is add some swap file to the system. This help whenever the system hits it's RAM limit. That way the server doesn't freeze up or completely crash. In some cases it is not beneficial to let the system use swap though. Same might apply here. Personally I did try the CryoUtilities with FS22. I didn't notice to much of a difference but it seemed to be stuttering less. I don't believe FPS was any different. So your video might be correct. I think it just highly depends on the game. And might not be worth the "trouble". For those that do want a specific game to run a little better I would suggest giving it a try. Otherwise just leave the system as is. I am sure Valve knows what's the best "config" for playing games on the Steam Deck? 😜
