I Tried Every Windows Optimization for FPS (+ Benchmarks)
ฝัง
- เผยแพร่เมื่อ 7 ก.พ. 2025
- In this video, we're diving deep into some of the most popular Windows optimization myths that gamers and optimizer gurus swear by. If you've ever wondered if tweaking certain Windows settings can truly give you a noticeable FPS boost, this video is for you! We’ll go through each setting, explain what it does, and reveal whether it actually impacts your gaming performance or if it’s just placebo.
Here’s what we’ll cover:
► Fullscreen Optimizations: Enabled vs Disabled
► Game Mode: On vs Off, does it really help?
► NVIDIA Reflex: Latency reduction or placebo?
► Fullscreen vs Borderless: Which gives better FPS?
► Power Plan Tweaks: Does "High Performance" really make a difference?
► Windows Visual Effects: Should you disable them?
► NVIDIA Control Panel Settings: What actually impacts FPS?
► Windows Game Bar: Can disabling it improve performance?
► Parked vs Unparked CPUs: Are you losing FPS?
► G-Sync: Should you enable it for competitive gaming?
► Optimizations for Windowed Games: What works and what doesn’t?
► Hardware Accelerated GPU Scheduling: A hidden FPS booster?
► Timer Resolution: Should you change it for better frame times?
► NVCleanstall: How does it optimize NVIDIA drivers?
► The %temp% Folder: Can clearing it boost FPS?
► Hyperthreading: Is it helping or hindering gaming performance?
► Efficiency Cores: Impact on gaming FPS?
► XMP: Does it affect gaming performance?
► Hardware Prefetcher: Should it be disabled?
Join our Discord
/ discord
🔗 Programs Used
CapFrameX: www.capframex....
UnparkCPU: coderbag.com/p...
LatencyMon: www.resplenden...
🖥️ Bench System
Intel i5 - 14600KF
NVidia RTX 4070
32GB DDR5 RAM
1TB SAMSUNG 980 PRO NVME
📃 Games
Fortnite
Valorant
Counter-Strike 2
Watch as we test these settings on real games, with benchmarks and comparisons to give you the truth behind every tweak. Stop wasting time on myths and make sure you're only using settings that truly optimize your Windows games.
🔔 Don’t forget to like, subscribe, and hit the notification bell for more myth-busting content!
Music Credit:
LesFreeMusic - Girl Dreams / girl-dreams
props for shutting down all those pseudo-optimizations and showing what really works
Just doing what had to be done a long time ago, thanks for watching!
not all PCs care equal, what works for one may not work for another and there are valid reasons for that
@@1godzukiI think 99% of the optimisation videos are bs, this one however shows gains that most can benefit from. That being said, I will agree that sometimes bizarre "fixes" work on windows, like the one time I was getting 2-3 fps in an early 2000 pc game from my childhood that I wanted to re-play . I looked everywhere for answers and found none until one day, on a whim I tried it and it was running at 60 fps. It took some time for me to realise that I had my ps4 controller plugged in... For some reason that made the game run at 60, a game that didn't even support controllers xD
@@1godzukiYou are not a bright individual
@@PixelDew1337 Back in the day I had a wireless razer mouse (idk the name, but it was the first one they sold) and if it was plugged in (the dongle, or the mouse itself) the computer would hang on the bios at boot and wouldn't boot.
U should test the difference between a regular windows install and a debloated windows install using a tool like Chris Titus utils
Brilliant idea! I'll be on it ASAP, ty
I'm using AtlasOS, wonder if that also changes anything 🤔
@@peterpanda7475 I have tried AtlasOS briefly in the past, just to see what it's all about, a comparison video (Custom OS vs Windows 11/10) will be coming soon so stay tuned! 👏
@@FrameSyncLabsthat’s great, thank you, but those custom ISOs are a little bit scarier to run compared to the Windows 10/11 LTSC versions or those from Chris Titus
For what I've tested there is no measurable difference, from optimized win 10 to standard win 11, really the only notable difference is disabling defender and couple of services.
regarding overclocking I would recommend everyone tightening their ram subtimings, very easy to do if you follow the guide as well as very noticeable improvements
THIS! On ddr5 if you have Hynix ram just tightening tRFC to 120-160ns depending on A/M-die and increasing tREFI to 50k+can increase performance in games by 10%+ if cpu-limited.
Yup. First find the highest STABLE frequency first(that means intense memory tests that test 97-98% of your capacity), than tighten down timings to get the lowest latency ram.
This will help kits that can clock higher but don't run as tight. 3200 c14 has identical performance to 3400 c15 and 3600 c16. Raising frequency or lowering timings both reduce latency. In these cases 3400 and 3600 is overall better than 3200 because of same latency but higher bandwidth.
Reduced latency is what boosts single core performance the most. Multi core doesn't see near as large a gain from latency as it does from increased bandwidth.
I'd say anything that is like 0.5% to 2% could literally be margin of error as well. It's pretty hard to test exactly the same every single time in game. How you move the camera could affect it even if you try to use the same spot
I've tested each multiple times to try and avoid that, but obviously you could be right as every run is different and a 1% difference is tiny
Except that there is a difference. And 1% on top of 1% on top of 1.5% on top of 0.5%... all that stacks up to real, actually large differences. So I don't subscribe to the margin of error crap when it comes to benchmarks. A 0.7% difference every time is still a 0.7% difference which matters once stacked, imo!
@@hdwblade you don't understand what you're talking about. tests have a margin of error.
Sure, if every time the test was run, there was a 1% difference, that would mean something.
But a margin of error could be up to 3 or 4%. The numbers change slightly each run.
@@freshfrij0les except consistently, across all reviewers, their benchmarks stay the same. Most do multiple runs and average those runs, and it always ends up with the better components slightly faster - consistently. So no, I stand by my point. The average of 3 runs being 0.7% faster does matter because they could do it 100 times and get the same result.
Important note about hyperthreading and e-cores: It's highly depended on the software/game as to whether you'll see a gain or a ~loss~. Can even depend on which processor you have as lower core count/less capable chips will suffer more with them disabled. Some high end chips with multiple core complexes can see reduces performance because the number of used cores is splitting across them incurring a latency penalty (or using the wrong cache on X3D models). And more games seem to have a gain with them on than ones that gain with it off (hence why they are on by default). So at bare minimum you should test each game you play and decide from there. If you're not sure then it's generally better to just leave it on as even windows and the multiple programs & services that you might run in the background would benefit from being on.
Well said, the games I tested are Valorant, Fortnite, CS2, and can confirm they benefit from it being off on an 14th gen intel i5 running Windows 11 24h2. Anything different (as in game, or cpu), please test it and see how it works for you, or better yet, use Process Lasso and only set specific games to use physical cores while having HT/SMT enabled for other apps/games - next time I'll mention it in the video to help make things clear. Thank you Carl!
Name one game that benefits from E-Cores on? I only play CS and don't give a shit about any other game but afaik there is only two games that got boosted by APO/E-cores and that was R6 and Metro "which no one plays" and as shown by the 7800X3D and 9800X3D core count is basically irrelevant, it's the performance of the cores paired with a big enough L3 cache that's relevant.
Even on my 12600K HT and E-Cores off increased gaming performance significantly so I don't know what the f you're gaming on that has E-Cores that actually increase performance? Maybe it's their mobile low power i3-1315U? 🤣
@@usrname1y Civ 6, Cities Skylines 2, and Rainbow 6 Siege all have reduce perf with e-cores disabled. Total War 3 Kingdoms and The Witcher 3 can have lower performance with them off on lower CPU models but higher or the same performance on higher models. Some games have similar average FPS but lower 1% lows with them turned off. So as mentioned... it's always good to test. If the games you play get an increase then by all means. It's just not a win across the board or they'd just not include them on the CPU.
If you only play cs why bother to write a comment?
Does this have any benefit in COD O.O
Quick tip for Nvidia users, if the person recording the tutorial opens the Nvidia control panel you can close the video
Please elaborate
@@ADepressedGuy-l1p Unless it's to make sure you're running on your Nvidia GPU (i.e. on laptops), most changes do effectively nothing and you are better off leaving them on default settings. Things like texture filtering and antialiasing are configured in your games so it is pointless to do it in your driver.
So glad I didn't open the control panel in this video haha. However, it's there for a reason, right?
Not all nvidia control panel settings are bad, some actually do help if you know what you're doing.
At the same time, I do feel you, because as soon as you see the control panel pop up, you know it's gonna be the same settings from 10 years ago that someone just copy-pasted over to their vid, oh well.
@@FrameSyncLabs indeed, most recent games dont even accept the changes you do in there, also i love when they tell you do drag the slider to the magical "performance mode"
@@ADepressedGuy-l1p Ray said pretty much all there's to it, also "performance mode" does not get you more fps, it only makes you gpu boost clock always be maxed out
Hold on, at 1:22 are you using an LDAT to measure latency, or an fps measuring software?
Reflex as I understand it won't decrease the latency you see in fps measuring software, but it will decrease the "Click - to - photon" latency, i.e. the total system latency.
3kliksphilip has a good video (or two) on reflex. But essentially, you need an LDAT to accurately measure latency with things like Nvidia reflex.
If you are using an ldat, and you're still not seeing latency improvements with reflex, that's very interesting. that absolutely would warrant further investigation!
Hey, I was using Nvidia FrameView to measure it, but we'll have to revisit that since now I have an actual Reflex Latency Analyzer tool, which works the same way as an LDAT with the exception that it's actually integrated into the display. This will be done when Reflex 2 gets released, so we can also compare the two :)
Dang, earned a sub, great quality, thought you had atleast 100k
Glad you liked it and ty for subbing ❤
Those thumbnails may only be true when switching to high performance mode in a gaming laptop that had some kind of power saving feature still enabled, but most are still fake results. You should've also tested Resizable BAR ON vs OFF but other than that, I like that you kept the video short (
Thanks! I decided to leave resizable bar on vs off for another video in which we'll test it in more graphically demanding games, I got you! :)
(I actually did test it in cs2/fortnite/valorant before, but didnt make a difference for my system in those games, in case you're wondering)
Even then the only way you're getting a 500% uplift on a gaming laptop if is your GPU literally off and your CPU is heavily downclocking itself like if it's on eco mode. Most balanced options still run the GPU and CPU at decent speeds just not max but that wouldn't result in 500% extra FPS. It's still heavily misleading in my opinion because no laptop is going to default eco mode while plugged in unless the user knows the settings and does it manually.
if you set up timer resolution correctly it can insanely decrease input delay
Hey, I'll make a separate video about this where we'll test the end-to-end latency with and without timer resolution :)
@@FrameSyncLabs try the amits timer resolution that’s what i did and it made my sleep times so much better
I'm in love with Process Lasso. The affinity tool to separate my 5900x CCD's into a gaming half and a everything-else half (within reason of system sensitive components, of course) really changes the game.
I couldn't imagine the benefits from a 7950x3d or 9950x3d where you have a 3d v-cache chiplet that is a higher speed and better bin of a 7800x3d, but you BIOS select a frequency CCD preference for windows scheduling, and then Lasso yourself onto the V-cache chiplet for games and game launchers (though I'd imagine you'd likely have to do this without the frequency BIOS setting for some launchers - as to avoid anti-cheat shittery).
If only I had the need, the budget, and the desire to swap to AM5 to test this niche-ness! A cheeky like and sub for the hope that you'll satiate that curiosity for me 😉
Of course I will satiate your curiosity, matter of fact, I already ordered a 7800x3d + a B650 motherboard and it's due to arrive 10th (11th?) of Jan! I haven't used Ryzen since 2020, so that'll surely be fun and your tips are appreciated!
Process lasso can be used to selectively disable hyper threading/smt on games that don't like it while keeping it enabled for those that benefit.
I'm gonna preface this by saying that I'm a computer scientist (mostly into software stuff these days but I studied some CPU stuff too in uni back then) and I'd say this video is better than 95% of what's out there, BUT with some caveats.
You measure an average of +2.2% FPS on Reflex, which is honestly an all right boost, but then you conclude that your "spray is bad in CS and Valorant" with it so you don't use it? Why is the "spray bad"? Do you notice stuttering? Also, what's the variance? Could the 2.2% be due to randomness in the values? (In math we'd say it's "statistically significant" if it's not due to randomness alone). Along these lines, at 1:56 we observe a difference of only 0.6% avg FPS but then it's viable to turn it off when beforehand you'd recommend we don't use reflex at 2.2%. I know I'm splitting hairs here, but when you claim you aren't a guru, you gotta be careful with making strong statements about stuff, or else an obnoxious engineer might hop into your comments section and raise these questions (lol).
Regarding hyperthreading: it lets the computer run two "imaginary" (logical) cores within one physical core, so more logical cores but worse performance for each. That means it's HIGHLY dependent on the program you're using. You'd have to show the specific games, and make a clear distinction between the ones that do multithreading well and poorly. Older games surely will use this feature a lot worse or not at all. Yes, you might get a boost in some games running in a single core, but you NEED to test this in a game by game basis. When I studied my last CPU engineering course our teacher always, ALWAYS repeated that you never know how a program will perform on a specific CPU (or circuit) till you tested it. Benchmarking programs is a very complicated science. That being said, I'm glad you brought hyperthreading up because in many games it can be a pretty massive boost as they tend to use a few threads only.
In regards to the prefetcher settings... I'm confused about it. Prefetching is INTEGRAL to modern computing, and built into circuitry to a very low level. I'm mindblown you can "disable" it. I recall designing prefetching subroutines when studying CPU logics and they were very integral to modern CPUs. I wonder if this setting is referring to some flag to make the prefetching more aggressive or it's just specific to some level of caché or w/e, because it sounds strange to me you can just turn off something so core (pun maybe intended?) to CPUs. If you find out what's it about, please let me know.
Anyway, I'll be keeping an eye on your channel. It's nice to see new TH-camrs grow so fast like that, it took me a while to start getting the avg view count you have on your four first videos (in fact my last month's videos have been pretty meh), and it took me MONTHS to get my first one over 60k, and this one looks like it will go past that and more. We will watch your career with great interest :)
Hey, obnoxious engineer (you're not!), First I want to say I measured a 2.2% FPS boost with Reflex OFF (not On), hence why the conclusion was to turn it off if /all/ you want is higher FPS, and this video is all about benchmarking what gives the biggest boost to try and compare to the promises of other videos claiming they'll give you a +500 FPS boost. I didn't make it clear and I apologise for it!
About the weird spray feeling in Valorant and CS2, at least on my system, I would describe it as micro stutters and a doo-doo hit reg (as in, you know you're hitting the player, but your bullets are missing). Regarding the 2.2%, I tried to minimize run variance by running the benchmark 3 times.
I mentioned in the video the games tested are Fortnite, Valorant and CS2 and Hyperthreading disabled did give me a boost in those games, I'd always recommend you test it yourself if you play a different game or have a wildly different system than mine as results will most likely differ. I also went ahead and ordered a Ryzen 7 9800x3d yesterday and it arrives on the 10th/11th of January, just for the sake of making another video specifically for AMD systems and see how these same settings affect it.
About Hardware Prefetcher, I assume it doesn't really turn it off /completely/, but more like sets a level instead. BIOS settings and their help strings are not always 100% accurate, but I'll definitely dig more into it in future videos!
I get that I forgot to mention some things (and make others more clear), but it's one of my first videos and my main idea was to make it short and to-the-point, as I'm aware people want to know the "WHAT's" more than the "WHY's" (as in, WHAT gives me more fps, not WHY). However, I'll be revisiting these settings in separate videos, so we can understand why do they work, how do they work and what's the best for your specific system and game.
Once again - thanks for your kind comment, you're really knowledgeable.
@@FrameSyncLabs Best of luck!
hyperthreading works by letting one 'thread' on a core do its own set of instructions while the other 'thread' is waiting on i/o, its a technique that in theory should allow for better uptime on core usage for compute tasks. this generally sucks for gaming because this type of parallelization doesn't really compute whatever you're rendering faster in practice but does incur latency and also each thread will tend to clear out the other's l1/l2 caches
Use process lasso to disable HT on a per game basis.
For your use case-video editing, streaming, playing videos, playing soccer games, and multitasking with multiple Chrome tabs-you want to optimize for multi-threaded performance and system responsiveness. Based on this, here's what I recommend:
Recommended BIOS Settings for Your Workloads
1. Efficiency Cores: Keep Enabled
Why: The E-cores on your Intel i7-12700H are designed to handle background tasks and light workloads efficiently. Disabling them would reduce your CPU's ability to multitask effectively, which is critical for your workflow (e.g., streaming, video editing, and having multiple Chrome tabs open).
Action: Leave E-cores enabled.
2. XMP/DOCP: Enable
Why: Enabling XMP/DOCP ensures your RAM runs at its maximum rated speed, which improves performance in memory-intensive tasks like video editing, streaming, and multitasking.
Action: Enable XMP/DOCP in the BIOS.
3. Hardware Prefetcher: Keep Enabled
Why: Hardware Prefetcher improves performance by preloading data into the CPU cache, which is beneficial for multitasking and applications like video editing and gaming.
Action: Leave Hardware Prefetcher enabled.
4. Hyperthreading: Keep Enabled
Why: Hyper-Threading allows your CPU to handle more threads simultaneously, which is crucial for video editing, streaming, and multitasking. Disabling it would significantly reduce performance in these workloads.
Action: Leave Hyper-Threading enabled.
5. Hardware Accelerated GPU Scheduling: Enable
Why: This feature improves GPU performance in some games and applications, including video editing software that leverages GPU acceleration.
Action: Enable Hardware-Accelerated GPU Scheduling in Windows (not in BIOS).
Smells like gpt
@@TheXanderage yeah it is indeed. i try to fit the settings to my daily use, with a little help of gpt 😊
@@SV.NguyễnHoàngAn Care, he is a good lier
brother, this channel is GOLD, no snake oil and no bullshit. you just went straight to the point, love you man good luck on the journey and rest assure ill stick to this channel
Much love bro, appreciate it
If you have completely broken your PC, sure you will get 300% boost and you will be at the same level as a newb.
All i did on my system was using chris titus to delete all those apps and turning off background apps + unneccesary windows effects and now its all running fine
@@Memmmeeboiiiiiiiii It really doesn't take a lot to have a stable and well performing system, good job :)
Reflex should be on. It makes the render queue much better in timing. Input Latency is in that case more important than 13fps more on 500+ fps
@@De-M-oN This comparison was just to determine what gives more raw FPS for the sake of seeing if we can really get "+500 FPS" as some youtube optimizations claim. I'll have to make a separate video about Reflex where we'll talk about how it works, the frame times, latency and more - thank you
Reflex only makes things worse if your gpu isn’t even maxed out
3:57 The graphs show that it's better on, but you're saying to disable, which is the correct one? :)
I have some observations to add:
I would argue that Game Mode can be more helpful on than off. Ex. when using Borderless window mode (especially with a lot of recent titles going for this mode by default rather that exclusive fullscreen), it suspends the desktop compositor from running, which can help with stutter or minimum fps. This in combination with FSO, by leveraging the DXGI flip model, can be very good pair of settings to keep on, or to enable both.
The problem with testing this is that it requires very specific scenarios to prove they are working, which most of the time is difficult to do.
HAGS, for example, helps with the CPU-GPU balance. Ex. when the testing scenario involves a graphics intensive situation, where many rendering tasks are needed (which rely on the CPU for scheduling), HAGS moves that workload to the GPU, which can perform that faster. It can probably hurt performance when using a very low-end GPU though.
I think it would be more helpful for everyone to know how things are working as well, to help them decide whether to enable or not some of these settings, besides the recommendation.
Thanks for your input! I'll definitely make a separate video explaining how everything works, but the point of this video was to test the most recommended tweaks by "optimization" channels which target games like Fortnite,Valorant and CS2. Some of them recommended specific settings to be OFF, while others said keep it ON, and thats how this video idea came to life - by testing the same common tweaks and show how they perform in these specific games. I'd never say these tweaks work the same for ALL games or systems, because thats impossible to determine without the viewer testing it for themselves. I do see your point and I forgot to mention some important things in the video so I'll write a comment when I get home and pin it to clear things up, thanks again! :)
HAGS is only good in CPU bound scenarios when you want to offload some of the work to your GPU, it can be game dependent if you are not entirely CPU bottle-necked
great content bro keep this up
Thanks, will do!
Really love the work. Did freethys optimization guide a while back, and was happy with the results. I’m on blurbuster forums quite a bit but I never saw these bios settings. Have tried them now but I get tiny micro stutters. Have a 13700k and 3080. Will mess around with the settings and see which is causing it. Love th channel. Subbed.
@@KennyKennTV Thanks for the sub!
Leaving hyperthreading on and setting all your processes to run on those cores while your game only runs on your physical cores via process lasso is better than outright disabling them, because you're still only using physical cores in game but they have barley any other processes on them.
Only way it would be worse is if you had very bad thermals/cooling, and disabling it gave you a massive temp boost which means much higher clocks, but my 7800x3D is efficient + I own a good cooler.
@@Hybred Well said, I outright disabled HT for the sake of this test, but its a thousand times better if you disable threads/ecores from Process Lasso which we'll talk more about in part 2 of this video. Thanks for pointing that out!
@FrameSyncLabs yes but I suppose this is a good solution for people who want something they can set and forget. Lasso will require adjusting each game.
Although one thing I discovered is whatever affinity you have set for Steam.exe, the game will match it if its launched through steam, so I have it set to my physical cores but the steamwebhelper and steamservice set to my hyperthreads.
Another discovery I had is the svchost.exe process titled (DcomLaunch) will set the affinity of any app to match itself, so make sure that affinity is set to "None"
1 minute in and so far. I recommend everyone listen to this video. It aligns with exactly what I have been using for the past year after spending an ungodly amount of time figuring it out. Thank fuck this isn’t sending my ocd into overdrive and I can relax
Thank you, glad I could help your OCD too haha
u should almost never disable hyper threading
If you really dont want to disable hyperthreading - thats fine. You can literally set your game to run only on physical cores using Process Lasso and see if it benefits from it! I'll make a separate video regarding this very soon. Thank you
@@FrameSyncLabs A in-depth Process Lasso guide would be very much appreciated.
you should almost always disable hyper threading actually. Nowadays it's just a marketing technology. General consumer doesn't need >3 cores at all. Only rare games utilize multiple cores and even then it still could sometimes be better without HT.
The only downside is disabling HT disables Sleep Mode
@@ATom-jm2jw I have high multi core utilization when gaming. More games than you'd think do utilize multiple cores, but typically not fully. The main benefit is that people very rarely have only their game running. A browser, discord, teamspeak, etc are almost always open for me, plus others sometimes as well.
Unless you have a 6C/12T CPU or stream / video edit, it's not needed anymore.
Timer resolution i only ever used on fortnite, and it wasnt for fps it was for input lag and surprisingly to this day i still feel the difference
@@lihpat I'll make a separate video on timer resolution and actually test if it decreases input lag. Even if it does, I highly doubt its gonna be anything significant. Default timer res when playing media/games is 1ms, with the actual app it can go down to only 0.5ms. :)
@ fair enough, idk what it adds for u but my friend using keyboard didnt notice anything while i would use controller and always noticed. And also my default would sometimes be higher than 1ms when i opened the app
@@FrameSyncLabs you are very wrong on timer res, some precise input games benefits from it insanely, like geometry dash
I think you may have just solved alot of issues for me that i have had with my system for YEARS, i will be back here after trying these tomorrow and will edit my comment and speak on some results. Cant wait for the video you are planning on making for intels p and e cores etc.
subbed! 🎉
Would love to know how it went for you, thanks for the sub!
any words?
Liarrr
I have an AORUS laptop with an i5 12500H, and the only thing that worked for me and significantly improved my FPS performance was reserving the P-cores for gaming and running Windows on the E-cores, all managed by Process Lasso. I went from about 240-280 FPS in Valorant to over 360. By the way, I have a 360Hz monitor. I tried all these modified Windows videos, scripts, etc., and with just this program, I solved everything regarding optimization.
you should do one for amd cards
yes sir 🙏🏻
DUDE HOLY FKKIN COW MAN, DISABLING CPU HYPERTHREADING BOOSTED MY PERFRMACE SOO MUCH LIKE WTF
This video is full of errors, you mentioned that NVIDIA Reflex does not do anything and lowers FPS, this is is the intended effect, and it does reduce latency, but the higher your FPS the less noticeable the improvement is. I'm not sure how G-Sync being enabled managed to give you extra performance since most of the time it actually gives you less. You gaining FPS with hyperthreading disabled confuses me, it might because of the E-Cores on Intel processors? You mentioned that "disabling" hardware prefetcher gave you extra performance but the graph on screen shows you gained performance with it enabled.
Overall I enjoyed the video but in future, please double-check before uploading because calling out the guides with their random "optimizations" and then commenting on them yourself in your own video incorrectly is equally as bad, maybe even worse as people might think to trust you more. I'd also appreciate a bit more details in the video in regards to how your testing was performed, what game or software, what hardware, how many times did you test to make sure the result is consistent rather than an odd error.
In regards to other stuff in the video, letting Windows manage core parking is the way to go, game mode was probably an outlier since it's been changed and the only thing game mode does in Windows now is disable notifications while in Fullscreen games. I recommend keeping the NVIDIA settings on default rather than using performance since a 0.6% improvement is not worth the downside of some games looking worse and also potentially having graphical problems. The MPO bug is only an issue in the new 24H2 Windows 11 update. Thank you for calling out Timer Resolution and the Windows power plan being set to High Performance in other guides. You also said to use Borderless earlier in the video then said to use Fullscreen at the end.
PS: Do not recommend CCleaner, it's problematic these days, just use the built-in Disk Cleanup or Storage Sense functionality in Windows 10 and 11.
Hey, that was a good read for sure! However, I just wanna make a few things clear by saying that absolutely none of these settings (except for the BIOS ones) are MY recommendations, it's merely just a conclusion of what gives the biggest FPS boost based on the benchmark, so we can compare it to the videos promising you +500 FPS if you tweak your NVidia control panel settings and by turning reflex to on+boost for example.
Also you're totally right about the higher FPS you get, the less noticable the latency improvement is (which in my case was -0.1ms latency with it ON and/or ON+BOOST), but what really struck me (and there's other videos/posts on the topic) is that Reflex /does/ mess up the spray shooting feeling in Valorant and CSGO especially, hence why I personally have it OFF (more FPS and shooting feels better, while the tradeoff is only +0.1ms latency, I can live with that)
Regarding G-Sync, I was honestly shocked to see it made such a difference, I will re-test and confirm if that's the case.
Hyperthreading always gave me a boost for games like Valorant, CS2, Fortnite and similar FPS games, on Intel that is, it seems there's no connection to the E-Cores as I've tested these separately and it still works the same. I'm yet to get my hands on a Ryzen chip so we can test if SMT behaves the same way. Other people in the comments have reported a similar FPS boost by disabling HT on Intel cpus too.
About Hardware Prefetcher, I simply mislabelled the results, so ON is OFF, and OFF is ON - I swapped them accidentally, totally an oopsie on my side, and I apologize for it.
The MPO bug has been a thing since driver 461.09 (2021, but maybe it was reintroduced in 24h2) and it's on NVidias official website, they even have a registry fix for it.
About the details, I added the games we tested and the system specifications right before the benchmarks start, what I forgot, though, is the Windows version (24h2) and the NVidia driver version (566.14), I also ran the benchmarks 3 times each to try and avoid margin errors, then aggregated them. I'll be including more details on that in the upcoming videos of similar nature.
I'm glad you enjoyed the rest of the video and I aplogize if I didn't explain everything clearly, but that will change in future videos. You live - you learn, this is also one of my first videos so I missed some fairly important details (I even gathered all complaints from the comments and saved them up in a txt file to avoid such things happening in the future).
P.S: The CCleaner part was a joke and I wouldnt recommend it after what happened in the past. The same way cleaning your %temp% folder for an fps boost is a joke so I thought it'd fit in lol, should've clarified :D once again ty for the insights
@@FrameSyncLabsthis guy us bugging, haters gonna hate, keep doing your thing man, you deserve a lot more subs 😁
Great video, i hope more people see this
Thank you, much love
The graph doesn't allign with what you narrated 4:14
You meant enabling it will gain performance right?
@@ThePhenoix504FI3D1 That's an oopsie on my side hahah! I labelled them wrong - Hardware Prefetcher OFF gains performance. Good catch, ty!
@@FrameSyncLabs no problem man, great video overall
@@FrameSyncLabs uuh
its not just about fps, you should have shown frametime graphs.
high fps ain't shit if the ft is all over the place
NVidia Reflex is for lower input latency IF! the Game is maxing out your gpu. This is known as render Queue buffering.
But if the game is cpu limited (i guess like in your case) it could have potentially tiny bit negative impact due to less fps.
--> Conclusion: turn on Mvidia Reflex or Amd anti lag only if gpu is maxed out!!!
Thanks for this, will definitely make the upcoming videos longer to be able to explain these type of things.
The point of this video was just to simply:
- Check if the optimizations that these channels recommend do anything on a new-ish system in 2025.
- Check if it makes a difference in the games they target, which are mostly Fortnite, Valorant and CS2.
I only recommend turning Reflex off on a similar system to mine (and if you play these specific games) as it changed the way shooting feels for me for the better (there's other videos on that topic too), only with roughly +0.1ms latency trade-off.
No just leave it on the minus fps is too miniscule to even make any difference. Reflex also reduces input latency for dlss frame gen vsync and other features
Nvidia reflex works well even if there is a slight performance hit, your overall latency will be reduced to as if you had set a manual consistently achievable frame cap. Now you can enjoy the better latency with more frames
when comparing hardware prefetcher, you mislabeled the graphs or made a mistake. It looks like this option set to ON gives higher fps
My bad, I did mislabel them indeed - sorry! OFF is the one that gives better performance, just swapped them by accident.
Yes i noticed that to.
@@JespeRCompeX Wont happen again haha :)
you can prob insert a caption in the video to correct the error in YT studio
@@ddallmann1373 I'm still new to this whole editing thing, and holy, is that possible? I'll check in the morning and correct it - thank you!
Holy shit I was not expecting that jump with hyperthreading, thanks man
I got you man, you're more than welcome! Just be careful as it might actually reduce performance in demanding AAA titles. I'll make a video on how to set up Process Lasso soon, so you can disable hyperthreading only for specific games from there :)
looking forward to it! :)
what about Amd?
Most of these optimizations apply to AMD too, with a few exceptions like:
- Hyperthreading which is the equivalent of AMD's simultaneous multithreading (SMT in BIOS)
- G Sync is similar to FreeSync
- Intel XMP is AMD's DOCP or A-XMP
- Then we have overclocking, which obviously will be different for every CPU
However, I can't tell if these optimizations will affect an AMD (Ryzen) CPU the same as an Intel one, so what I can do for you is buy a Ryzen processor and make another video in the future testing these same settings on an AMD system 👍
Its really worked man that's amazing
Glad I could help, congrats
You did not turn off windows ram compression - the major FPS killer in the games.
Ooo that's interesting! I will definitely do some research and try it so we can talk about it in part 2 of this video, thank you!
Explain yourself young man...
@@FrameSyncLabs Yes that kills tons of fps. On a few games shockingly many fps
@@slendertroll92 its true, it must be off at all costs
Thank you for the video. I have done some of these optimisations on "old reliable" PC to squeeze out every last drop of performance but in a couple of days I will be building a new one because parts arrived, so this video is very helpful on what to do, or not :D
That's exciting 🙌 Best of luck (or skill!) Hope it turns out well! What's the specs?
@FrameSyncLabs Thank you. To keep it short, I've managed to get a Ryzen 9800X3D, B650 Tomahawk, 32GB 6000 cl32, 2TB m.2 and a Sapphire Nitro+ 7900XTX and a 860W platinum PSU. There's a bunch of other things but I don't want to make a wall of text xD
@@RandomGeneratorOutOfPlace Lets goo! Managing to get your hands on a 9800x3d nowadays is lucky as hell 😁Enjoy your new build when you finish it and congrats!
@@FrameSyncLabs Tyvm :D
Intel(R)Core(TM) i3-2130cpu @3.40GHz lenovo enhanced experience 2.0 are my specs bad and would these optimizations help
Man you are the only optimizer i trust, because even khorvie techs tweaks reduced my fps
Thank you so much
4:03? turn it off or on the graphs showing that the performance is better with it on??
@@ABN-DARA99 I mislabeled it, sorry. Off gives better performance :)
@@FrameSyncLabs oh okays good to know
testing now
thanks
with nvidia reflex you mean the low latancy mode in the control panel/app or you mean in the games that support nvidia reflex?
The nvidia control panel setting isn't called Reflex, my bad for not clarifying, I meant the In-Game setting
@@FrameSyncLabs thank you👍
@@FrameSyncLabsso you didnt test the low latency setting in the control panel?
@@zen4714 that is the same thing for games that do not natively support reflex.
Great video, for cs2 did u turn e cores off in bios and didnt change any settings ingame or launch options?
Hey, no launch options were used and the e cores were disabled in the BIOS :)
@@FrameSyncLabs thank you!!
Careful with disabling SMT. It depends on CPU and game. Many times, if not most, it will be detrimental.
The point here was to show how it performs in the post popular multiplayer games (valorant, csgo and fortnite), which are also the games targeted by the youtube optimization community. These games dont care about multithreading and will perform better with SMT/HT turned off. I mentioned in the video that if you're doing something like rendering/video editing, its good to keep it on.
question
can you do some tweaks in taskmanager?
1:29 but doesn't playing windowed increase latency?
ya in some games
Nothing noticeable, my latency increased by 0.1 ms switching from fullscreen to windowed borderless and this could be a benchmark error so dont worry too much about it
@@FrameSyncLabs windowed is real bad if u r using integrated graphics or low end system. Back on 2015 i was using intel hd4000 to play csgo and it gave me around -10fps hit, it put me down to 40 fps
@@axandraalex5869 That pc was meant to play solitaire not csgo 😂
If it's working correctly on your win10/11 install - there's almost no difference at all between them.
I don't think so with W10/11. They swapped out the old WDM method with a swap chain model that should be identical to pretty much any DX12/Vulkan fullscreen mode (which in reality is a borderless window since they don't do exclusive fullscreen anymore).
Game mode: mainly for multitaskers, will be beneficial if you do a lot of shi at once, will be better having it off if theres no big difference
FSO: latency reasons, better disabling it in properties
Borderless? You never use borderless if youre serious about your fps gaming due to, again, input lag
Balanced and generally power plans are useful on laptops, not on pc. Wont limit your cpu, you can set your curves in drivers if you really like to (or disable aggressive boost)
good video tho, those clickbait videos are iqless af:D
@@dexyydev Well said, borderless doesnt add any latency with "Optimizations for Windowed Games" ON in Win11 though, for Win10 - definitely stay away from it!
iqless LOL, have a heart
My brother in christ, you ran these tests on a CPU released not even 2 full years ago, on an RTX 4070 graphics card and with 32GB of RAM
I get that sometimes people clickbait, but this sort of small tweaks are for low-mid end hardware that's already strugling to run the new insanely bloated windows versions. They need to squeeze as much as they can, anything can make a difference, I've been there
I get your point, but then again, the performance increase will be relative to the increase in this video. No way on earth anyone is going from 60fps to 539fps just by tweaking their Windows settings.
I'm more than happy that these work(ed) for people with low-end PCs, but the purpose of this video was to show HOW MUCH it helps, in percentages, so people can have realistic expectations instead of hoping for high 3-digit FPS numbers just because they got baited by a random youtuber that told them to change their nvidia control panel settings.
I also really appreciate your comment, we gotta show love for our low-end pc brothers (and sisters). I'll try to buy an older PC in the future so we can test and see how it affects their systems - thank you! :)
@FrameSyncLabs of course, the numbers must be quite an exageration. But going from say 16 to 24 from a single setting really starts adding up on the low end scene once you put many tweaks together
About exsclusive fullscreen and borderless - maybe on modern pc there is no difference but on older specs like mine ( ryzen 5 1600, GTX 1060 6gb, 16gb ram) there is a BIG difference, trust me. It depend on a game tho, but most games work much better with fullscreen. For example, Im playing days gone right now, on borderless i have lot of drops from 60 fps to 50, after switch to fullscreen, i've got solid 60 fps on high settings. And it looks the same for most game i've played on this pc
Hey, there will always be a difference between Fullscreen vs Borderless in Windows 10. If you're using Windows 11 with "Optimizations for Windowed Games" ON, there's barely, if any, latency increase or FPS difference.
Try atlasOS, it makes a difference
We'll try the most popular custom OS in an upcoming video, such as Atlas and Revi etc - thank you!
Nope it does not make much difference from an optimised normal windows. Also using custom OS made by shady people is a no go, they can insert a RAT whenever they want.
@@nameless.0190atlas os is open source.
Nvidia reflex is supposed to make your fps drop a bit by nature. In short, its syncing the gpu and cpu so that the render queue latency is shortened. Even though you're getting less frames, youre receiving the frames sooner; therefore lower latency despite lower frames.
In theory that's true, but it seems to make no real difference on high FPS numbers (latency-wise, saw a 0.1-0.2ms improvement with 650fps avg). Obviously, in demanding AAA titles, people still see up to 30-40ms improvement which is INSANE and turning Nvidia Reflex ON is a must. I also can't wait for Reflex 2 with predictive rendering and frame warp, now that's exciting :)
Quality stuff. No scam. Dropped a like and subbed to the channel,
Much appreciated, thanks for the sub!
Yo, using the task manager, go to details, and then right click in your proccess, affinity and select the P cores only, and put higher priority. The e-cores would handle the rest
@@GrulbGL That's another good way of doing it instead of disabling hyperthreading/e-cores, or if you don't want to run an extra app (Process Lasso) on your PC - thanks for mentioning it
NIce vid. I found that using MSI mode causes stutters in some games so I don't use it.
Glad you liked it! There's no "one fits all" tweak, so testing it for yourself was a good call, thanks for the info!
Good results but how do you achieve each step you showed? Specifically the ones that are in windows
truly no BS video and well done aslo. Lowering layency guide would be much appreciated 🙏
@@oblivionfrag8035 Thank you! Will definitely make one in the very near future :)
You are really awesome because you explained everything in a concise way.
Personally checking"✅ Disable Fullscreen optimization" boosted my fps by 40 units in CS2, also reflex on +10 fps, disabling core 0 and 1 using lasso process gave me like 10 fps. The rest is kinda pure placebo (at least for me) or no effect. Windows 11, laptop user, i5 11400H, Rtx 3050, 8gb ram. Most of these Windows tweaks depend on the computer and SO you use. If you are a Ryzen x3D user the impact will not be so great.
I also use riva to cap my fps.
Thank go a debunk no bs guide we all appreciate you for this fr Fr
Huh, eveyone else seems to recommend turning game mode on since there isnt a performance hit but in games like rdr2 people have measured a decent performance improvement
I dont think having something like game mode mess with your processes and how your PC works in general is a good thing, ever. Especially when it gives you 0 info on what is it doing in the first place. Use Process Lasso "ProBalance" and exclude your games instead :)
great video and I recommend you to do a video on the new nvidia app
Awesome content, thank you for the warnings and quick explanations, really put my morning coffee on the next level
Nothing better than a morning coffee with some good ol' optimization videos, glad you enjoy it, and thank you
Thanks for this guide! I have an older MoBo from 2016 (X99a Gaming Pro Carbon) CPU (i6950x) which is overclocked to 4.0ghz All (10) Core with quad channel DDR4 3200 and a Radeon 7800 XT. I have tried many (and I mean many) of these tuning/tweaking/overclocking/undervolting/registry/bios etc tips and tricks...and just like anything else some work better than others on different Hardware and Software (I am stuck on Win 10) versions.
The biggest (new) benefit I got from this video personally is the "Hardware Pre-Fetcher" setting in the BIOS. Turning this off yielded me 10% increase in single core and 5% in multi!
The main thing I disagree with is Hyper Threading, but I know this used to be the case with older games....and you are testing with some pretty old games. I do not play Fortnite or CS2 or Valorant, and in my testing on a variety of games I see mostly increases in performance with HT Enabled. Also on my machine I was not able to achieve even a single millivolt of or meghertz of overclocking ability...and the decrease in thermals although noticable, was negligable.
Some other (perhaps more advanced) things you are not covering here that made a difference for me are DPC/ISR Latencies...CPU/GPU MSI Mode/Affinity etc. Stuff that creators like Khorvie covers.
After all the tweaking and finding out what works best for my system I can play i.e. Starfield (yah I agree it sucks but it's hardware intensive) @1440p Ultra (NO Upscaling) max settings and get 60fps. In Indiana Jones on Supreme settings (NO Upscaling) i get 70fps average (52fps1%lows) in the Jungle and 110+fps average (70+fps 1 %lows) indoors. It's not amazing but it works well enough!
Hey, I'm glad to hear something from this video worked for you! :)
Hyperthreading will definitely help when specific games utilize multiple threads properly, others that don't will perform better with it off, that's why I'd recommend everyone tests it for themselves in their game of choice and see what works best.
I've made a video about DPC and ISR latencies recently, there will be videos on MSI mode and affinities in the near future, I've got lots of ideas on my list and I'm also taking ideas from viewers - so thanks for suggesting this!
The results mentioned look pretty good and stable to me. Good job and keep it up!
Btw if you ever find a game "optimization" tutorial showing you to edit something in regedit don't EVER do that. It will brick your PC after a while
Dang, now this is what i call a quality content. Thank you for testing all of these
Great and informative video. Also I recommend you add some graphical intensive games to have a more complete benchmark.
Thanks, will do!
Disabling SMT might help slightly in older or poorly optimized games that heavily favor single-threaded performance. However, modern CPUs can handle both single and multi-threading fine, disabling SMT is unlikely to provide a major boost.
Could you maybe show us how to change those bios settings on windows 11?
A 100%! I will make a video on how to actually apply those settings soon :)
I've had a bad laptop for so many years so i used to follow these optimization guides and none of them really worked, thanks for making this video it helps people like me and others in the same situation. ❤
You're very welcome! :)
1:30 there's no way borderless window doesn't add latency, sometimes my games change to windowed when alt tabing and i can feel instantly when moving the camera that i have noticeably higher latency/input lag
same here, I always play on fullscreen and then I don't have to distract myself anyway with tabbing out every 5 sec
Hey, are you using Windows 10 or 11 with Optimizations for Windowed Games?
@@FrameSyncLabs Win 11, as far as I remember I didn't disable it so it's probably on
@ I still hold onto Win 10
It is really dependent on if you're running old games or not. The difference between windowed and full screen on a modern game due to graphics API is most likely going to be down to margin of error or unique hardware quirks. But if you're running old games, especially if you don't use things like dgVoodoo to (potentially) upgrade the game to a modern API, you will see better performance with Fullscreen due to the games inner workings not playing kind with windowed mode compared to full screen. Windows 7 and below may also naturally gain FPS due to lacking modern OS changes too. I've seen very little difference in games these days except on old games once I went from 7 to updated 10/11.
There is also the side benefit like someone else said, you alt tab less :P
Maybe my math doesn't math, but I count 9 things that improve FPS somewhere from 0.3-3%. What happens if you do them all at once instead of one at a time? How much improvement are you getting? EDIT: Nevermind, I should have just watched the video.
Haha, thanks for the comment though
Where do i find how to turn off Hyper Threading and turn off e-cores? Great vid!
Hey, thanks! You can disable hyperthreading in the BIOS, prettt sure its somewhere under "Advanced CPU Settings" or similar. You can also use the search bar to find it easily. Same goes for e-cores if your processor has them, some don't.
@ wow quick response King thanks. I found hyper threading, testing now. Will do e cores after.
@ thanks so much, but I can’t find where do turn off e cores in my bios. i13900k and 4090 do u think i have them?
@ruthfull3261 Your CPU might not have e cores, do a quick google search to find out :)
@ ok thank you so much my game already feels better! i am looking, i have 1 socket 24 cores 24 logical processors.
This might be a big ask, but is it possible to do the same but for the NVIDIA Control Panel ? I'm really curious to see what settings actually affect performance.
Solid video!
I'm working on the new video right now, we'll talk about the best settings and best driver(s) in it, so stay tuned and thank you!
@@FrameSyncLabs Just watched that video. You're the man 😁
Your content is refreshing and straight to the point. Keep it up!
Awesome video and thank you for filtering all the crap. Do you have a video on how to change the settings in the bios that actually made a big difference?
Not yet, this video was merely a benchmark of common "tweaks", there are multiple other videos in the works like:
- How to set the correct BIOS settings
- What nvidia driver is the best
- What nvidia control panel settings are the best
... and plenty more about the things we talked about in this video, I'm trying my best to get them ready as soon as possible!
Thanks so much, appreciated
Thank you for debunking all of these, however if you could make the "real" optimization guide more detailed that would be absolutely amazing!
Great video, seriously
Working on it! Thank you
Thanks for theses stats! It is valuable, i appreciate if you can do a video comparing Linux vs Windows gaming performance
I'll try to learn more about Linux as I havent dipped into it yet, but will definitely add it to my list of upcoming videos - thanks!
You shouldn't disable hyper threading blindly, it's basically leaves you with out HW cores, which is not ideal
This was just for the sake of the benchmarks for this video, if a game benefits from Hyperthreading disabled - I'd always recommend disabling it through Process Lasso only for that specific game instead of system-wide from the BIOS :)
thanks for the skibidi video man! im all about getting the maximum out of your rig! really good videoooooooooooooooooooooooooooooooooooooooooooooooooooo
hello to you from Russia, I liked your video so much that I want to ask you to do more and more! You did a great job, I tried more than once to find tweaks or other methods to improve FPS in games! Oh yes, these 1000 fps are these professionals, I have done so many times I caught a broken system that again and again I had to change Windows again, if something goes wrong it’s Google translator’s fault XD
algo que me fez ganhar mais desempenho em jogo, foi ir no gerenciamento de tarefas, escolher o jogo, ir para detalhes e definir prioridade como alta e definir afinidade e desativar CPU 0
This is good stuff. I've subscribed and will check out your older videos too. Worth pointing out that this is just one system. It could be that a system using an AMD/Intel GPU or AMD CPU may respond differently to these settings. Obviously, like you mention in the video, it's unlikely that you will see the 500%+ gains the clickbait videos claim, but some settings may have different gains/losses based on different hardware. I doubt it's within budget right now for your channel to look at current gen hardware but perhaps if you start making some money from the channel then you could buy some older hardware and see what differences these settings can make to those systems. Bleeding edge hardware is great and does draw big view numbers, but budget/patient gamers are a huge viewership.
You don't even know how much I appreciate that comment, you're speaking the truth.
I'm aware every system might respond differently, however, this is not an optimization video, but merely a benchmark with conclusions of how these optimizations are holding up in 2025 on a recent(ish) system using the most popular shooter games as an example.
I did include the system specs as to avoid confusion, but I have to admit I missed some important details which will be there in upcoming videos! For the time being, if anyone has a question regarding these tweaks or their hardware, I'll be answering every comment to help clear things up. I'd still recommend to /always/ test such optimizations for yourself to ensure they work the same way on your system, even if it's closely similar to mine.
Especially with X3D chips on the rise, I'd love to get one (currently out of stock) and build another system, so we can see how it performs on recent AMD cpus. It definitely isn't within budget, but I'm saving up so we can get it ASAP when it comes back in stock. If the channel grows and makes some money, that's even better as we'll be able to test on an even wider array of PCs ranging from low end to high end, fingers crossed and thanks once again for the kind comment
@@FrameSyncLabs Thanks for the reply, and good luck with the channel! :D
tweaking possibillities are near endless with nvidia registry shananigans beeing discovered etc. tbf a super debloated optimized windows install with thousands of tweaks post install will get you a shit ton of more frames but also is a shit ton of work hence people who have the knowledge of optimizing systems in that extreme manner will charge you a good sum for it
Great vid mate!
3:56 benchmark says on is better but you said disabling it
@@ChopyTweaks My bad, I mislabeled the results for Hardware Prefetcher, sorry. Off gives better performance! :)
That Nvidia Reflex thing. Is that hit reg issue only valorant and csgo or all games? My main fps is CoD and even the #1 ranked player have On+Boost. What about Reflex 2?
I will make a video for Reflex 2 veeeery soon! For now its only a Valorant/CS2 issue when you already play on high FPS, haven't heard anything bad about Reflex and COD :)
Excellent no frills video. Really appreciate my time being respected!
Always! Thanks for watching
Should you disable hardware prefetcher? Since it has better fps then off but you said in the video to turn it off?
Hey, my bad, I actually swapped the labels accidentally. As said in the video - OFF gives better performance :)
@ okay that’s what I thought thanks for the speedy response👍
great and simple explanation. thanks for the video. but some game engines use HT. for example, I know that source, the game engine of cs2, needs HT. I am very confused whether it needs to be turned off or on. (I5-13400F)
Thank you! I've got an FPS increase in all the games tested (CS2, Valorant, Fortnite) on my system, but it might work differently on others. I suggest you give it a try by disabling it and test to see if it gives you a performance increase your 13th gen Intel :)
You know what I do? I turn off everything. Just leaving the necessary Windows background processes. When doing this. My laptop idles at 1% with 7% to 6% RAM usage which is about 2.5 GBs out of 32. My laptop runs really great now. It never crashes, or freezes. Sure it's not the most powerful thing, so if I push it, it chugs along but never crashes anything. I also removed all the bloatware so I'm sure that helps too.
Sometimes that might be the best thing to do, depending on the games you play. I'll also make a video about clean windows install vs debloated windows, I'm REALLY curious what kind of difference does it make for gaming. And if there isn't any, hey, at least you cleaned your PC from all the crap! :)
@@FrameSyncLabs It has helped. Games don't chug as hard when the settings are cranked, and they would sometimes freeze before I decided to clean my laptop. No more does this happen. Also I have gotten about a 10 to 20 fps boost in most games. Not crazy, but my laptop was never top of the line to begin with.
@@Group-935 10-20 FPS boost on a laptop is HUGE anyway. Good job on that!
@FrameSyncLabs Yeah, I guess so. Thanks. It also helps that anytime I use it, it is plugged into the power supply. Oh! And the biggest performance boost I got was, this is silly... Using it like a desktop. I turned off the internal display, hooked up an HDMI cable from it to a monitor, and disabled the CPU's integrated graphics. That's how I have gotten untold performance out of my laptop.
good video , i was really curious if those videos really works
hardware prefetcher graphs looks different from your conclusion unless you swapped it in the graph
I did mislabel it by mistake, sorry. Off gives better performance :)
thanks man, I won't look at so many useless videos to increase my fps.
Hey, glad I could save you some time :)
Reflex reduces input latency when using adaptive sync. Should not be off for all cases
Thank you very much for this, my laptop is very weak and now I can play my favourite game with more ease.
@@CometInOrbit So glad I can help, enjoy! :)
What resolution were these teated on. Do these make more of a difference on 1440p vs 1080p?
This test was ran on 1080p, but I'm pretty sure there will be a difference on 1440p too in these games :)
The channel that needed me, keep up the good work, I'll sub you right away, could you have done fps tests with these methods too, cpu 0 on vs off, also process lasso program how much impact does it have, also windowsxlite which is one of the best for debload, thank you.
I'll make a separate video soon specifically for Process Lasso where we'll test all the settings - thank you!
great job and i agree with you. Now do one for latency
Thanks! Latency optimizations are on the list of upcoming videos, shouldn't take long! :)