https://www.youtube.com/live/dv7XR5RduLg?si=mRKwpEZ-t2SYVA3r
Creator of the tech tool showed on a live stream that the performance improvement is negligible
No performance benefits with this tool and 4gb of VRAM bios tweak also doesn’t improve performance like it used to.
Mostly due to the 3,5,5 bios update it seams. Still a good tool for saving space but I’m going to skip when my oled arrives
Popping in to give my 2 cents, as the person who hosted the stream and developed the tool.
TL;DR: The results in the stream were largely good, with 1 game having slight negatives and the other 2 games being better with CU2. Saying “it’s dead” is an extreme overreaction, especially considering I tested games that have historically benefitted very litte from CU2. The entire reason I chose those games is because I wanted to make sure that the “worst” ones would still not suffer, further testing with games that should actually benefit is necessary.
Here are the exact numbers, in the order I tested them:
Red Dead 2 got better performance with CU, even WITH the 200/1040 bug on the CU runs:
Default Preset had a substantially better minimum, and was much smoother:
Without CU:
Min - 22.0081
Max - 63.9321
Average - 49.8511
With CU + 1GB VRAM:
Min - 28.6349
Max - 63.2811
Average - 50.6368
“Prefer Quality” Preset was slightly better across the board:
Without CU:
Min - 19.3353
Max - 59.8151
Average - 46.3316
With CU + 1GB VRAM:
Min - 21.7967
Max - 59.4665
Average - 47.0571
Cyberpunk 2077 had minor losses, but it seems that it is at least partially the fault of having less RAM with 4GB of VRAM, needs more testing:
Steam Deck Preset lost about 1FPS:
Without CU:
48.61
With CU + 4GB VRAM:
47.88
High Preset also lost about 1 FPS:
Without CU:
43.87
With CU + 4GB VRAM:
42.88
Horizon Zero Dawn had gains nearly across the board:
Original Preset was within margin of error:
Without CU:
9/50/97
With CU + 4GB VRAM:
8/51/99
Ultimate Preset had gains in both highs and lows:
Without CU:
9/36/70
With CU + 4GB VRAM:
12/36/76
Just want to say as a fellow underappreciated dev; You rock and I appreciate the amount of effort you make, for free, with the simple purpose of improving your own gameplay and then sharing with us to do the same.
I am happy to see the benchmarks you provide alongside and have even in the early days validated some (Horizon Zero Dawn) having a significant impact (on my LCD Deck) on long term stutter and allowing me to enjoy the title entirely on deck. There’s literally nothing stopping people from playing stock, recording mango data, and then trying the tool themselves to compare.
Please ignore the detractors and keep it up.
you must not be a very good dev if you fall for this then.
Multiple reports are coming in for Rule #1 violations for this thread. Bans will be handed out based on the rule-breaking comment(s) and not who instigated a flame war or who thinks they are right or wrong. Sides won’t be chosen. Respectfully, please observe Rule #1 Be kind or get yeeted. This is the first and only warning that will be given. Thank you.
Your post or comment was removed because it was deemed either unkind/toxic/harassing/insulting/offensive/trolling and/or inappropriate thereby breaking Rule #1 of r/steamdeck. We want this sub to feel welcoming to anyone and everyone who comes here. Discussion and debate are encouraged but name-calling, harassment, being rude to others, generally toxic behavior, and slurs will not be tolerated.
This rule violation has resulted in removal of your content, and could result in a ban from the sub and/or a report to Reddit.
Bottom line - Be kind or get yeeted.
Thank you!
TLDR: very minor difference due to greatly lowering swappiness making a page file basically go unused (which yes the page file will decrease performance) while pretending other factors are in play and guaranteed to get worse performance the longer you play due to bad tweaks such as turning off ram defragmentation. meanwhile acting like a swap file can improve performance (impossible) and that other old build.prop tweaks are having any effect. I could provide some real tweaks such as turning off background services, preventing journald from writing to the disk, unmounting debugfs and tracefs, etc.
Show us the numbers to back it up.
And yet you refuse to provide these things or any proof of your own
Thanks for the post! CU made it so I could actually play through the Marvel Xcom style game. Before using CU it would crash every time I went to sleep in the game when on the Steam Deck. I was then able to beat the whole game on the SD thanks you CU. Thanks for all your hard work, I have not uninstalled it because I have not had any issues while using it and also because the Marvel xcom game was such a good experience, until I have an issue I will leave this installed.
Thanks for the post and all the hard work you do!
I love how the misinformative post has more likes than the comment from the dev.
Unfortunately that is how any social media works these days. The “engagement” comes from any negative repercussion compared to the facts
We fixed that already
Thank you for all the work you do for this community. As soon as I ordered my OLED i was waiting on word for you on how CU holds up. Glad to hear it’s still worth having!
Also I don’t know if this would be the correct place to put this, but I did have a scary glitch occur. I just reinstalled the latest version of cryoutilities on my steam deck OLED, i then hit recommended settings and after that I turned it off from desktop and booted recovery mode to change the vram to 4gb. I rebooted back up, went back into desktop to double check that it worked, and went back to gaming mode (all pretty fast, i think before the deck really had time to boot up and settle in? Is that a thing?). As i switched to gaming mode the screen went black and these glitchy vertical green lines started creeping across my screen from right to left, while i could still move the sticks and get menu sounds. I ended up holding the power button down til it shut off and rebooted and everything seems fine? But I really thought i borked the whole thing for a minute there lol. Maybe this isn’t actually cryoutilities related but in case it is I thought i should tell you! That said I’ve turned it on and off multiple times later today just during regular usage and it seems fine now?
Firstly, thank you for watching the content, and for using CU!
Secondly, it sounds like either GameScope or the Steam client failed to start properly on your way back into game mode. Nothing CU does (or the 4GB of VRAM) should cause that, so I think you’ll be fine as long as it doesn’t continue happening :)
I’d like to add onto this that anecdotally CryoUtilities did resolve a crashing issue for me in Warhammer 3 Total War. I’m on my new oled deck, latest firmware and stable release update, and was having persistent crashes after battles. I installed CU after watching your livestream, set it to the recommended settings, and since have finished several battles in the same campaign with no crashing.
I had a similar experience with CU fixing crashes in modded Rimworld and Cities Skylines. I suspect the swap file size change is the biggest factor since all 3 games would crash during loading. Even if the performance fixes were somewhat nullified by valve’s changes, it’s still great to be able to easily increase the swap file size to deal with crashes like these.
Anyway thanks for this incredible tool and I’d encourage anyone with loading issues (especially with the 3 games I mentioned) to give it a shot.
I’ll make a note of TW3, thank you for letting me know! I’m just happy that it’s still helping people :D
Quick question:
How is Total War 3 on deck? does it maintain at least 30 fps in battles/looks decent? How are controls? What Proton version?
Awesome thanks for posting the results. It was hard to keep track of them in the stream. Keep up the good work
Thank you for the comment, pay no heed to the downvotes. I appreciate you watching the stream and for reading my comment here!
Hey its you! Just wanted to say thanks for everything you’re doing for the community, you’re awesome man
Thank you so much, it’s astounding to me that I’ve reached as many people as I have. I’ll keep doing my best to help :)
I’m still using the original 500 and whatever gb steam deck and probably won’t be upgrading. CU is still relevant to me 😅
Thank you for your hard work sir.
Dude you are a huge credit to the community, do keep testing and creating tools because most of us are paying attention and supporting you! We need more people like you 👍
Thank you so much for the kind words, I don’t plan on stopping any time soon!
The first thing I did when I moved my 1TB over from the OG SD to the OLED is make sure Cryoutils still had all its tweaks applied. It made a big difference on the OG and I can’t see it being detrimental on 3.5. I’ve seen people saying they think it’s not needed as much anymore, but stuff like OP saying it’s dead, well, it’s just clickbait unfortunately. I’m glad you chimed in, and I look forward to seeing more extensive testing.
Thanks for using CU on your new Deck as well, I’m happy that you’ve had good luck with it!
I’m hoping to start more extensive testing tomorrow, now that the weekend is over, stay tuned!
Keep the good work.
Thank you so much, have a great day!