Do NVIDIA GPUs REALLY lose performance over time? GeForce 2023 vs 2024 Drivers Benchmarked
Vložit
- čas přidán 5. 08. 2024
- Get 30% discount with my "SKAG" code!
Windows 11 Pro ($22): www.gvgmall.com/gvg/SKAG1
Windows 10 Pro ($15): www.gvgmall.com/gvg/SKAG
Windows 10 Home ($14): www.gvgmall.com/gvg/SKAGH
Windows 11 Home($20): www.gvgmall.com/gvg/SKAG1H
Office 2019 ($46): www.gvgmall.com/gvg/SKAG9
Hello Dear Viewers :D
In this video I am testing the FIRST Drivers for the RTX 4070Ti cards (528.02) versus the latest ones available when I tested (555.99) to see if NVIDIA GPUs do in fact lose performance over time as stated by a lot of people.
00:00 - Intro (Old vs New GeForce Drivers, Methodology, etc.)
01:47 - Cyberpunk 2077
02:11 - The Witcher 3
02:26 - Counter Strike 2
02:53 - Forza Horizon 5
03:11 - Fortnite
03:46 - PUBG: Battlegrounds
04:11 - Dying Light 2 (Raster + RT)
04:50 - Microsoft Flight Simulator 2020
05:16 - Metro Exodus: Enhanced Edition
05:50 - Dragon's Dogma 2
06:10 - Starfield
06:39 - The Last of Us: Part 1
07:05 - Alan Wake 2
07:51 - 14 Games Average
08:34 - Final Thoughts (Performance, driver tweaks, etc.)
11:25 - Channel Members
11:45 - More Videos
Hope you enjoy the video. Comment with your thoughts about it. Don't forget to hit like, subscribe and share the video if you can.
If you want, support me on:
Patreon: / ancientgameplays
Paypal: www.paypal.me/AncientGameplays
Social Networks:
Discord: / discord
Instagram: / agpisco
Facebook: / ancientgameplays
AMD AM4 PC USED for CPU testing:
CPU Cooler: Mars Gaming ML360
MB: Asus X570 Strix-F
RAM: 2 x 8GB - Patriot Viper 4400Mhz CL19 (3600MHz CL16)
SSD (Nvme): 500GB + 1TB
PSU: BitFenix Whisper 650W+ Gold
CASE: Bitfenix Nova Mesh SE TG
INTEL 10/11th Gen PC USED for CPU testing:
CPU Cooler: Artic Freezer Penta 33
MB: MSI Z490 Tomahawk
RAM: 2 x 8GB - Patriot Viper 4400Mhz CL19 (3200 MHz CL14)
SSD: Kingston V300 120Gb
HDD: Seagate Barracuda 2Tb 7200 rpm
PSU: Seasonic SIII 650W Bronze
CASE: Aerocool One Frost Edition
INTEL 12th Gen PC USED for CPU testing:
CPU Cooler: Mars Gaming ML360
MB: Gigabyte Z690 Aorus ELITE (DDR4)
RAM: 2 x 8GB - Patriot Viper 4400Mhz CL19 (3600 MHz CL16)
SSD: 240GB
SSD (Nvme): 500GB
HDD: Seagate Barracuda 2TB 7200 rpm
PSU: Aerocool X-Strike 800W 80+ Silver Semi Modular
CASE: Phanteks P400A White
AMD AM5 PC USED for CPU testing:
CPU Cooler: Enermax Aquafusion 360
MB: Asrock B650E Taichi
RAM: 2 x 16GB - Flare X5 6000MHz CL32 (6000 MHz CL30)
SSD (Nvme): 1TB
PSU: Enermax Revolution D.F.X 1050W (80+ Gold)
CASE: Enermax StarryKnight SK30
My Main PC (used for GPU testing):
CPU: AMD Ryzen 7 7700X (5.4GHz Static Overclock)
CPU Cooler: Noctua NH-D15S
MB: MSI X670E Carbon Wifi
RAM: 2 x 16GB - Flare X5 6000MHz CL32 (6000 MHz CL30)
SSD (Nvme): 500GB + 2TB + 2TB + 4TB
PSU: Corsair RM750X (White) 750W 80+ Gold
CASE: Thermaltake View 51 TG
#nvidiadrivers #nvidiagpu #ancientgameplays - Věda a technologie
AMD Adrenalin 2022 vs 2024: czcams.com/video/OQ0h1lVaWHo/video.html
AMD vs NVIDIA CPU Drivers Overhead: czcams.com/video/crij8C4S9FM/video.html
RX 7600XT vs 7700XT vs 4060Ti 16GB: czcams.com/video/sEF1irzKJcI/video.html
brother was reflex messing with CS:2?
Nope, disabled@@quintrapnell3605
@@AncientGameplays CS2 on DX12 ? Is this a typo ? I believe the game runs on DX11.
@@tvelichkov Yeap, have to fix that. thanks!
I think for these findings to actually be meaningful it should be with older generation GPUs like the 1-2k series or even 3k series. Where Nvidia actually has an incentive to lower performance. There's not reason they would lower the performance of their new lineup, at least not until the new series come out.
Yes, that could be done, but I wanted to see with the newer generation, and it seems that it isn't the case. Will be doing a new video with the 4080 Super or something after a while of the RTX 5000 series drop
@@AncientGameplays Can you try do a video with the GTX models because this seems to be the case and relevant for a lot of people.
@@v.9876 I could try if I had one
They don't decrease the performance of older GPUs when new ones release, when games get released they are intentionally optimized only for the latest generation, making the performance gap bigger than it should be. Same story with phones.
@@AncientGameplays Cheapest 1060 6GB on Ebay is 94$ CAD + shipping, if you think it would be worth it, if not, heh, no biggie.
Amd drivers ✅
Geforce drivers ✅
Arc drivers next?
No doubt the results will show massive increases in performance. HUB has done this.
Intel will just have Fabio showing 300% gains with graphics stretching across the screen. Their gains have been massive at times.
would love to see the drivers for the arc a530m on the lenovo loq 15
Fabio i think that what people say is that Nvidia loses performance in time is with older GPUs, the review should has made with rtx3000 and rtx 2000 GPUs , nobody says that Nvidia drivers of a current generation loses performance....
Will test that later too
@@AncientGameplays Faça isso com RTX 3090 se tiver ou semelhante. 😎
Well, my 2060 Super and 3080 haven't lost any performance... Got benchmarks from when I got both new and tested again, and it's within margin of error.
@@lilpain1997the only performance it may lose is due to having the cooling system to be cleaned up
Everyone saying he should try it on older hardware, the fact that even on newer hardware some games lost performance, just means older hardware will be even worse. And also, if you go too far back newer drivers won't be supported.
Congrats on the house Fabio, well deserved my friend.
Thank you my friend!
right and kinda wrong i can go far enough back to a 1900's game that new devs from a game brought back and play in it with modern gpu and cpu drivers in it and get around 200 500 frames in it while modern games i get near 200 frames but i get only around 100 or so no matter what now
@@Cole_Zulkowski reading this made my brain explode.
@@bluej511 good? any ho i use a rtx 3060 in a laptop and a rx 6800h to run my games current drivers now and almost every game runs like my 2020 gaming laptop could which is to say that laptop was using a gtx gpu with a far weaker 6 to 5 core cpu in it and i can just playing dead island 2 assassin creed fallout destiny 2 and 1 borderlands halo far cry etc my framerate goes between each game from stable 100 frames to almost all the time 10 to 1 frames which is horrible plus stutters as well which makes it even worse so modern gaming is harder. to do unless you have a rtx 4090 and a beast cpu to run with it but if not expect framerate in games now to be decent but not great unless you do go back to older games that still support modern gpu and cpu in it and have it give you 200 300+ frames but other than that gaming is a chore to get your gpu and cpu to run properly in a game now let alone to get the framerate you want
@@Cole_Zulkowski In the 1900's Red Dead Redemption run live in real time with actual multiplayer and multicultural whole Earth experience for everyone without the need of game console or gpu. Porsche's were EV and Diesel engine ran exclusively on vegetable oil. But I'd rather render my game in 0's to 1500's settings where all Forests were Unreal and Land Servers where less populated and more sea and continents were still intact to build your own Civilization... 1900's to 1950's Battlefield were also a little too much, and also the Viral trending topics without proper antiviral developers a real hassle to cope... It ran at faster FPS though even if you walked more than any new launched 3rd person game...
I can vouch for AMD drivers maturing positively but I cannot wait to see if my suspicions about Nvidia are correct. Thanks for doing this.
I would like to see a comparison between AMD and Nvidia driver changes all on the same chart.
It would be interesting to see if you bought a card a year+ ago of similar price, how are things now.
Here is a good question for you ancient gameplays, do setting the shader cache size in the nvidia control panel in manage 3D settings to 10GB or 100GB better instead of leaving it on driver default?
Not only those GPU's loose performance but also their own tech is not available for old same gen .(like DLSS for RTX cards , they will force you to buy a new GPU because your last gen card is not compatible with newer DLSS shiexxx)!
@@riannair7101 thats how capitalism works
@@arx117 that capitalism rule you mentioned not apply to AMD fine wine!
@@riannair7101 This is true, AMD stopped innovating their GPUs
dlss work every rtx card frame gen need 4000 series. and thats why nvidia fg is better becouse of dedicated hardware.
@@powerplay074 Say that to rtx series 2xxx and 3xxx...but i sense in you a nshitia sheep!🤣
Very nice investigation Fabio👍👍 , I always had that concern, I have the same card with March 2023 drivers, I guess I will only update them if really necessary 😁😁
You could have tested the RTX 3050 on the old 2022 drivers vs 2024 drivers. That would be interesting.
Dude was dreaming he was on a cruise. It was so immersive he had to do the ship sound 🤣
🤣🤣🤌 vr 2.0
Dreams are better than any VR ;)
I love these intros🤣🤣😁
Can you check Gaming vs. Studio drivers? Considering Nvidia disables DSR factors for some monitors/tvs with the gaming drivers, but not for the studio ones.
Hello AncientGameplays I’ve seen you made a video about lossless scaling and I’ve been seeing some people having trouble using it with the RX 7600 causing shutdowns on their pc and I just wanted to know if you are going to make a video talking about the topic. I think it would be very interesting to see your thoughts.
Hey fabio quick question. How do i prevent msi afterburner from changing my gpu settings in the adrenaline software settings. Everytime i open afterburner my fan curves and undervolt reset in adrenaline software
just use Afterburner at stock settings and it will follow the adrenalin ones
@@AncientGameplays h̶o̶w̶ ̶d̶o̶ ̶i̶ ̶d̶o̶ ̶t̶h̶a̶t̶?̶ ̶d̶o̶ ̶i̶ ̶n̶e̶e̶d̶ ̶t̶o̶ ̶r̶e̶i̶n̶s̶t̶a̶l̶l̶ ̶t̶h̶e̶ ̶s̶o̶f̶t̶w̶a̶r̶e̶?̶
EDIT: Thanks figured it out mate! Also insane work on the video
Close afterburner then in its folder in program files delete what is in profiles and reopen afterburner. Fabio is probably correct
@AncientGameplays @v.9876 THANK YOU GUYS SO MUCH! goats!
I think the question about Nvidia drivers over time slow down gpus was because Pascal series were seeing massive performance drops compared to newer generation gpus in many modern games on more demanding engines ….for example texture mapping etc in newer DX12 engines.
I remember my 8600 GT got slower after installing new driver....this happened in 2010
Since then i migrare to AMD
you should check older cards, not current tier one. so like 2080 launch drivers, one from 3xxx era and newest one
or you mean like people with 1080ti using the gpu for years? Another AMDumb owned.
Will try 💪
The thing is, the fact on trying it on newer hardware and the drivers slowing down the cards or speeding them up, it is basically useless to testing because the older hardware would either be same or worse
Everyone says that the 537.58 drivers are the most stable, so that's the one I run for my rtx 3070, wondering if that is true 🤔
Interesting test, I never thought about this before
Very cool video and interesting data. Thank you.
Happy for your new place man!
Thanks!
Nice vid Fabio - I believe CS and similar games would be similar without Rebar on the newest drivers as well - technically most competitive games, or games that needs a ton of CPU power benefits from rebar OFF - MW3/WZ3 is a great example as well
Good point!
I sold ngrydia RTX two years ago because drivers was freaking me out .. all the time must search for stable drivers and better perfomance... download half year or one year old driver for stability and perfomance... sold.. and buy RX7800 and next 7900gre.. best deal ever.. no problems with drivers... pure power no upscaling... and silent gaming...
I'm buyng a rtx 4060 just to compare with 7600
Diference betwen 4060 and 7600 is the ray tracing performance?
Watch my comparison video with those cards 💪
i have i5 10400f and gtx 1650 super and i want to buy i5 12600kf and rtx 4070 is there going to be a huge bottleneck ?
hey man great video, I've been watching your videos since I had a RX580 and now I changed my whole pc to 3070 and i5 13600k. but I have problem in some games like elden ring. my problem is fps drops, im playing on 1080p res and it expected to have no problems but idk why i have some wierd fps drops its like it goes from 60 to 56. in doesnt seams a lot but in my screen it feels like almost freezing for a second. my gpu driver is 556.12 and I dont know which version of nvidia driver is better for my gpu. I spent lots of money on this new pc but im not getting what i paid for and its just so sad for me, I will appreciate if you could help me
would also be cool to test older windows versions and make sure the other drivers for other devices are also older
My 6800xt saw a huge performance uplift with newer drivers back then in many games, but doesn't see big performance gains since, My 3090ti well I haven't seen too big or any increase other than some select games like Forza Motorsport and Diablo 4.
for the 7900xtx can I use 2300 for the minimum to get a more stable performance or is 500 for the minimum frequency the better option?
Use 2300 minimum only for extreme cpu bottleneck scenarios
Thank you for such a test
I hope these big companies watch techtubers benchmark videos
Good to know, great work 👍
Excellent work Fabio, funny that the card lost a bit of performance in some games, but that may be to improve stability.
My suggestion for next driver testing: How the RTX 3000 and RX 6000 aged after the launch of each current generation? Are AMD and Nvidia Supporting the previous generation to a good level?
I guess I can do that :D
Good luck in building the new Studio! Spend quality time with GF too 😀
Congrats on the house, keep up the great work 👍
Thank you!
@AncienGameplays congrats on the new house, may you create beautiful loving memories there while reaching 1million subs
Thank you for the kind words
Sorry if I disturb you here, I ha E a 7999 gré and my memory goes at 92 degree Celsius.... Is too much or not, sometime in some games from 239 fps I go down to 9 /20 fps... Have you any suggestions. Thx in advance for your help
92C is completely fine, dont worry.
Maybe your vrms are overheating. Undervolt it
@@AncientGameplays have u made tutorial for undervolt if remember correctly...right ?
@@AncientGameplays i have used the automateed undervolt at 1025, do yiu think is enough ?
Such a friendly demeanor contrasted with a T-shirt that says "F*** off." 😂 Great video, btw, and love to see you with Moore's Law is Dead.
Thanks, and thanks haha
Nice testing. But surely it would be nice if you could test a card like 3070, to see how the previous gen is aging up.
True, should have done that
can we have rt results for 2nd gen rt cores from both nvidia and amd (rtx 3080 vs 7800xt or 3090 vs 7900xt) would love to see how amd's 2nd gen rt cores compare to nvidia ones.
you can find that improvements on digital foundry's Video's When matched for Shader count/R.O.P count. Although not being about to be equal but, with little difference in Tensor cores count on Nvidia's here is how it currently is playing out
Nvidia's currently gained 6% from RTX 2000 to RTX 3,000. They also gained the same 6% increase from RTX 3000 to RTX 4000.
While AMD's has increased from around 8-11% from RDNA2 to RNDA3.
They have both have increased rasterizing more from like 50% to 100% each generation. The efficiency of RT cores is currently linked to rasterization rendering. The increased rasterization has help RT gain a lot more from previous generation cards over the actual RT improvements.
@@kevinerbs2778 Thank u for your time providing such detailed information, Truly appreciated
I think you found out why 1P lows on cs2 were so low. Thanks.
🇵🇹🇵🇹🇵🇹
Is IT possibile to do this test with games 5-10 yrs old ?
I'll try
@@AncientGameplays thank You :3
I would question how much performance change people think they are seeing is the actual hardware.
GPUs, especially NVIDIA, run pretty hot; thermal pastes dry out faster under high temps, so could the older cards be starting to throttle due to this; it would be interesting to see a comparison of older, heavily used cards from when it was launched, against a brand new unused card of the same model (new-old stock or latest manufacturing date?).
All using the same driver release.
Nah, nvidia gpus just run hot woth poor models usually
Thank you, Fabio. NVIDIA's wine doesn't get much better over time, but that was just the latest generation of video cards. Interesting, how does the wine taste for the still on the market RTX 2000 Series? I have doubts it gets better. Thank you for the Witcher 3 comparison. Due to horrible stuttering on RTX306012G in this game, I switched to the RX6750XT and RX7800XT.
Have a 3070 Ti and had heard about the intentional performance degradation through new drivers so I'm still running an early 2023 version.
Just update mate, much better overall
My sincere congrats with a new house, Bro !
Thank you!
The talk about this is about older generation GPUs. So like test 2070 super or one of the other 20xx series gpus that are not sold or made new anymore.
The talk back in the day was even about newer gens, will test the older ones
@@AncientGameplays Well I first heard about it as a concept with iphone and how discontinued ones would get slowdowns from updating the OS. In GPU terms I remember my friend that had a GTX 770 cursing the 10xx series release and new drivers. But I dismissed it as a skill issue on his part.
Congrats on the new house!
Thank you!
Wouldn’t it make sense to test this on a 2000 series card
If I had one
Where can I get those old drivers?
You just download them on nvidia website
Ah this looks like a watch that will interest me and what I think AMD did and probably still do is limit certain releases potential driver side to match an Nvidia tier then unlock that driver limitation later. That's what they did with the RX series that at first were close to or even slower than the Nvidia counterpart but after a certain driver release get 15%+ boost which is a lot.
Nah, AMD is just better at hardware and NVIDIA at software, so as they increase the software optimization their hardware gets considerably better
@@AncientGameplays Makes sense as they seem to lack on the software side but it is what people were thinking when the boost happened.
Very useful stuff👍🏼👍🏼👍🏼👍🏼
Thanks
@@AncientGameplays 😍😍
Current gem cards were never the issue. What people have been complaining is that Nvidia will drop optimization for older gen that are still supported.
So, instead of being about 4000s, this video should be about RTX 2000 and 3000.
Will be testing the older ones soon!
congrats on the new house :D
Thank you :D
you should include firestrike benchmarks next. Thats where i find the most difference when I had an Nvidia card, when i use the latest drivers at the time my 1080ti would be 1000+ points lower than it should be
You need to done this with rtx3070, rtx2070, and gtx 1070
When’s the 2024 adrenaline settings?
Some very interesting new feature is coming soon I guess, so I'll wait to see that
Thanks for the testing Fabio, good to see it compared to AMD’s “Finewine Technology”..
Also, congratulations to you and your partner on your new house.. that’s very exciting and looking forward to seeing the new studio.
Thank you Adam
I have an Idea for AMD ,Nvidia and Intel. What about making an application that have access to all drivers versions and automatically choose the best driver for the game that you are going to play and before you the game launches it choosed and switching to that driver automatically and then the game launches, and making it have an offline method by choosing the best version for the game locally (from the versions that you installed).
That's almost impossible
Congratulations, got a new house🙌🙌👍👍
Thank you!
nvidia reddit recommended driver is usually 537.58
But...nvidia doesn't have bad drivers...
When the 6700 xt launched Techpowerup and Hardware Unboxed had it as 4 percent slower than the RTX 3070 at 1440p. In their newest reviews they have the 3070 being 9 (hbu) or 12(tpu) percent faster than the 6700xt at 1440p on average. Obviously the new reviews use newer games. You see similar results across the rdna2 vs ampere stack. This is raster only native performance.
I don't think the difference is caused by drivers. Just one architecture was better designed for current generation games.
From my testes the 3070 is still overall faster unless your using it in vram limited scenarios like ultra settings
Maybe AMD has an increase due to the fact that initially the drivers were crude, but now they have been corrected? But Nvidia's drivers were good from the start, and they simply cannot be improved? for example, fsr at amd also gets better, only after many years. is it the same as good wine?
Excelente contenido. Mi humilde criterio es que( de manera general) los drivers de Nvidea desde que son lanzados están muy bien optimizados. Al contrario de AMD, por eso la diferencia cuando ha pasado una generación. 🙏
I thought he would say sadly there is advantage in performance overtime with new drivers
Hahaha xD
Pretty sure all hardware is going to get slower over time regardless of anything, software updates can only do so much.
always wondered if nvidia does anything shady, thanks for testing
to me i feel like they do and its because for what i think is they are slowly giving up on supporting gamers and are not going out of their way to fix as many issues and or make areas of improvement better anymore because of well one reason i see is they care more on the super computers and ai side of things which is what they are focusing a lot more on and why probably for me framerate sucks performance sucks and so forth
I wonder if is it easier to optimise drivers for nvidia than amd or they just have a massive team with infinite budget and they can deliver the maximum performance on day one. 🤔
Yesnt😊
Possibly, bigger ecosystem. Amd has been improving that recently
@@AncientGameplays I mean it is not that bad, because its nice to see my card improving xd
I honestly think the "fine wine" thing has nothing to do with drivers but rather with how the companies operate. When Nvidia releases a new architecture, they partner up with A LOT of game developers in terms of "sponsorships" to utilize key features and the new architecture of cards. AMD on the other hand has generally released more long-lived architecture in conjunction with console releases, like GCN-series and RDNA2+ series. However, they don't have as much resources to pour into the same initial "optimization marketing strategy" as Nvidia. Ultimately at release you get the impression that Nvidia cards are REALLY powerful while AMD is generally "slightly behind", and this is where our performance perception of cards usually take place (when most reviewers are benchmarking them and when they are in the media spotlight). But over time, Nvidias focus shift towards new cards/architecture, and those initial tailored optimizations on the older nvidia architecture is lost on newer titles (as that marketing window has passed). while most game developers still optimize well for the older AMD-based architecture residing in consoles. Resulting in new games starting to become more optimized for older AMD hardware compared to older Nvidia hardware. That way older AMD cards are starting to run ahead in newer titles, and that's where the "fine wine" comes in... Not in driver optimizations, but rather that developers have been focusing on consoles rather than older nvidia architecture.
Not with rdna2 and rdna3 case. Sam made huge differences
@@AncientGameplays I don't need convincing with RDNA2 and RDNA3. :) I think they have preformed brilliantly from the get-go. And that's not even considering market price difference between Nvidia and AMD 6-12months down the line. My RX6800 cost the same as an RTX3060Ti in Sweden at purchase. Nvidia doesn't even have the upper hand at RT considering the massive leap in raw GPU grunt AMD offered, for the same price.don't need convincing with RDNA2 and RDNA3. :) I think they have preformed brilliantly from the get-go. And that's not even considering market price difference between Nvidia and AMD 6-12months down the line. My RX6800 cost the same as an RTX3060Ti in Sweden at purchase. Nvidia doesn't even have the upper hand at RT considering the massive leap in raw GPU grunt AMD offered, for the same price.
I still think the "game optimizations" will hold true over time however. That the optimization effort Nvidia puts into game development partnerships will eventually dwindle. Meanwhile RDNA2+ optimizations will remain as long as current gen consoles are relevant.
@@PixelShadeconsole does not always mean AMD same architecture on pc can benefit from it even if AMD want that to be true. A few years ago (this is during 8th gen era) war thunder dev said their game perform very well on console like PS4 but on pc AMD GCN end up having more issues than nvidia because nvidia actively engage with developer about issues (drivers) than AMD. part of the reason AMD only focus on triple A games or their sponsored tittle on pc else their effort with game developer are minimal.
the real question sure the current newest cards are gonna work very good its specificcally those older gpus to manipulate the user to upgrade to the newest cards
Will try that too
What about the gtx’s i think many of us have gtx on their build
Will come
I have never heard that Nvidia gets slower over time. Only that AMD driver updates have a bigger impact, compared to Nvidia's. Probably because Nvidia being the market leader by a huge margin gets the benefit of games being optimized for their GPUs first, where as AMD always has to catch up. So if an AMD card on launch is as good or better as an Nvidia card, that means you can expect that in the long run AMD will come out on top. AMD is for people playing the long game, Nvidia is for people that wanna get the best perf right away.
Actually it was peoven years ago that nvidia cards did get slower, newer ones? Maybe not
I bet older gen cards (10/20 series) are definitely slower with never drivers.
So the drivers play a big part in the 1% low fames. This shows drivers can be better optimized for smoother gameplay.
You made a mistake choosing newest rtx, fabio. It should be tested with Rtx 2000/3000, not with 4000. I don't know how it looks with rtx 2000+ but i can confirm this "loss of performance" on nvidia gpus. I bought gtx 980 in 2014 and somehow it was 10-15% slower in 2018. Then I replaced it with gtx 1080 and it also got weaker after time. I noticed it in 2020 (so 2 years later) while playing the witcher 3. I had save with the same spot where I finished playing this game and my fps dropped from 95-100fps to ~80fps. I tested more games and it was really slower. Sometimes 5%, sometimes 10% to even 20% fps drop...
I will test an rtx 3000 as soon as the rtx 5000 drop then
@@AncientGameplays Good to hear. We will be waiting :)
The AMD fine wine is used for older cards. You should compare generation old cards from both companies. I mean AMD 6000 series vs. Nvidia 3000 series cards
well, my previous video with the RX 7900XT disagrees
@AncientGameplays 7900 is a current-gen card. I mean AMD 6000 series vs. Nvidia 3000 series old vs. new driver's
@@moto6981 Yes, I know. I meant that my previous video shows a perf uplift even on newer gens
I think you should have tested the 3000 series and not the latest card. Of course latest cards will get better performance otherwise what is the point of buying? I'm more concerned about older cards or previous generations get reduced performance over time through drivers. You know Nvidia. That would make more sense actually. 🤷🏻♀️
yeah, it seems like you didn't watch all tests
I wonder why in older games the oerfoemance is worse... Kinda weird tho
not weird at all I guess...
According to most recent Steam hardware profiles , Nvidia cards are used by 74% of gamers. AMD is 16% (Intel takes the slack) and a few years ago they were only 10%. Most new games are optimized for Nvidia cards. It takes a few months for AMD to dial-in. My 6600 XT was released in summer 2021 and all the reviews from then use stats that were obliterated a long time ago. Better drivers optimized for games, plus the upgrades to FSR made big differences. I never judge an AMD review by its first stats.
Yes, but the time... While you're waiting, new cards will be released.. The train will leave😂
Good conclusion from bad results.
The point is old drivers vs new ones not for newest gen and newest game titles.
I have experienced same issues as shown here in CS2, but it happened in some other old titles.
Its common that people with GTX graphics cards don't play Alan Wake or Cyberpunk but rather online competitive games..
Finally those online competitive games are most popular games in the world.
The rumors you trying to dismantle are experienced by many.
Geforce drivers has some bad releases which could hurt very old games on really old game engines by dozens of percents.
👉There was such good rule either stick with drivers released at date of purchase of gpu or update one by one and stick with last good performing ones depending on own preferred /most favourite game.
It was 257 driver for my old (previous) GTX1050👍
But I agree😉 that once it moved to playing mostly DX11 game, even newest Geforce drivers didn't affect my performance at all, rather it had positive effect on stability.
/to mention games I saw negative impact was really pure hardcore death match Quake 1 arena on DX or Open GL client, but newest drivers didn't hurt performance in relatively new, regularly updated DX11 World of warships/
We get at least one good stable Nvidia driver every year. Just have to stick to it until the next decent driver comes around.
Bruh hit me with the fanta sea out of nowhere 😂😂😂
🤣🤣
Mate you should have done the test using a previous generation card. Obviously they aren't going to slack off on optimization for the current gen.
I will do another test soon 💪
@@AncientGameplays Looking forward to it mate. You're one of the few channels who do in depth testing on less looked over stuff like this. Keep up the good work man.
Hmm, I think the idea was nice but the execution wasn't, I don't see a reason why NVIDIA would purposefully lower their current generation cards performance, it makes no sense when you think about it, this should be done with GTX 1080Ti, RTX 2080/Ti and even RTX 3080/90, picking old drivers corresponding to these cards year of release.
I mean, we did get some lower 1% lows. I will though make a new video with the 3000 series as soon as the 5000 series arrive
the best way to test that is use old cards of the same performance like 1060 vs rx 580 on modern games. and those cards are cheap to get.
That's a crap if Nvidia want to charge more while not making this better than AMD then this shit makes no sense
Brother can you try doing a video using the alternative nvidia drivers without telemetry and nvdia geforce center? Thank me later i think it will be an i tresting video
already did that: czcams.com/video/-zDcGcXRP9Y/video.html
I think many don't understand what Nvidia driver gimp means, nvidia never make slower their current newest gpu's, they make slower their previous gen gpu's when they release new ones and they don't make them slower by forcefully making drivers worse, not at all they just drop support and make no effort in optimizing stuff for previous gen gpu's. Good example nvidia lies about framegen not working on rtx 20 and 30 series and AMD proved them wrong by making their framegen equivalent working on them. Or good example when RTX 2080 was released it was barely on par with GTX 1080 Ti in all games from that era, today it's usually GTX 1080 TI is more on par with RTX 2060 Super and can't even touch 2080. If newer game has same engine from time when old gpu was released it will usually run well, if it's updated engine it will run like **** , good example GTA V and RDR2 , same engine, but red dead had upgraded one and it 1080 Ti is barely slower than 2080 Ti in GTA V while rdr2 ran really poorly. Methods of testing this myth from techtubers are flawed, many still don't understand what gpu users mean by nvidia driver gimp. In one sentence it can be exmplained "Ur good with your gpu for atleast one generation after it, however one generation more and it's basically fully ignored by manufacturer, he won't make older game run poorly, however he will do no effort in making game run good in newest tittles" Simple
I can only speak from my personal experience with this, but I will give you one example, I had 2x 780 in SLI back in the day and 2x 7950 in crossfire. The 7950 can still be used today in many games. The 780's are basically a useless paperweight. It was around that generation I stopped buying brand new Nvidia gear.
true
Nice ❤❤❤
too recent cards ... i had gtx760 and when i bought it it was between r9 280 and 280x ... when i was selling it after 3/4 years and jumping to RX470 it was between R7 260x and 270 so this is what people mean when talking about geforce losing power .
Interesting
Where are you from, you have a very cool accent
Portugal here :D
you have to wait for the release of the 50 series card drivers to do this comparison, that's where Nvidia hits its old graphics cards
Maybe
well , if old to new Nvidia drivers does not give noticeble performence uplift that does not mean they are bad, it maybe mean even older drivers are optimised to thier best to began with, and if AMD released less quality/performence drivers before and now they are optimising them , it does not mean AMD is better, its maybe means they started to figure out how to make their drivers in thier full form.
but no doupt free performence are always welcome.
it's funny that you made this video today because I noticed recently that games like GTA V were currently working very poorly on nvidia
windows is a wild card. On Linux the stuff just get better all time, with time.
I read these results as: nvidias drivers are more optimised when the game actually launches.
I have a full AMD pc, so im not really biased towards nvidia, but I would much prefer having all the performances immediately, instead of getting them along the way 😅
The positive is that amd knows where to optimize, so hopefully they know what to prioritize to make the card more valuable (AI upscaling, driver stability, secondly RT)
Pubg case is very sensitive. Pubg should be tested with directx11 not enhanced. Maybe then can be most accurate result.
It was tested several times as I said, also, dx11e is the standard the game puts you in