Interesting video. I'd like to see how Doom eternal holds up across all 3. Very fast paced, and is one of the few games I can think of that the developer specifically focused on reducing latency on Stadia, publicly, after launch
Oh great choice. I find destiny 2 is playable with the stadia controller but doom 2016 the latency is very noticeable. Almost unplayable lol. I live in Seattle near a stadia server and have amazing internet.
I get 4 ms to GeForce now i am in Montreal Canada. So remove 40 sec from your test and its the results i get :) (they opened a new server for east Canada) so my experience with GeForce now 3080 tier on par with your local hardware impressive.
Insightful comparisons. Any idea how consistent your internet access latency was while you were testing? Was it during a busy time of day for internet use - or middle of the night when your neighbors aren't streaming, there's less congestion, and you get more consistent internet latency? Do you want to share who your ISP is? fiber or cable?
Extraction is bugged on GFN and it won't stream at 120fps despite showing 120fps in whatever ubisoft overlay and despite 120fps stream settings. It is as if vsync was ON on system level, the same goes for No Man's Sky. There are few games that are "bugged", maybe because there's some conflict with GFN's virtual display set at 60hz, or some other thing. But to this day they haven't fixed it. Basically, you are getting only 60fps over 120fps stream. Easily confirmed on 120hz+ monitor if you record a 120fps video. There's also a built in overlay ctrl alt f7 that shows the timeframe graph. So extraction isn't the best game to cover GFN input lag over 120fps stream.
Makes the result even more interesting, there's a big difference in latency when running the game all low and max. Also ran the fps counter on the ubisoft ui itself and it was giving me readings over 100 fps, you can actually see it in the video on the top right. Ill pass this info along to nvidia though to see what they say
@@AnytimeAnywhereGaming Nvidia knows about it, I made a reddit thread about this weirdness, shared it with GFN discord mods and guys at CGX channel. If you're enthusiastic about cloud and tech you'll be tempted to look into it even more. Take a loot at ctrl+alt+f7 graph in top left in different games for starters. Or capture locally 120fps video and go frame by frame to confirm 120fpa streaming. There's no denying that if game is rendered at very high fps, even if afterwards is limited to 16ms frame time for 60fps, the input lag will be lower compared to the game running at 60fps and streamed at 60fps. When it comes to r6 extraction i immediately noticed that it's not as fluid as other games, despite showing 120fps+ in game, then I started to dig into it and easily confirmed "60fps stream vsync".
Hi! Do you know how bit rate effects latency? I have a gigabit connection, so if 50 mbps is not a problem do I have any latency reasons to play on like 30-40 mbps? For me latency is over quality.
Latency will primarily be effected by your distance to the server your connecting to, if you go to your settings you can manually connect to a different one but it should auto connect you to the nearest server
@@AnytimeAnywhereGaming I have done it, my question is if there is any latency difference between 50 mbps bitrate setting and a lower one like 30 mbps.
@@AnytimeAnywhereGaming I ve a genuine doubt. If this cloud gaming is so good is there any point in buying new GPUs??? I mean its far better to invest in a strong internet connection and good monitor rather than building a rig with say 3080 and i5 etc which would cost like around 2000 USD naa. Is GeForce shooting in their foot by allowing this cloud Gaming? Cuz their own Hardware is gonna sit dusty at stores right!!
@@1989rs500 GPU's still have a market and will have one for quite some time, remember that geforce now has its limitations, not ever major title is on the service yet or arrives day one. Additionally new and stronger gpus will probably hit markets before they hit gfn servers as they did with the 3080 tier. There's also the argument to be made for work related pc's, gfn allows you to play games but not work on programs like premiere and such either.
@@AnytimeAnywhereGaming Well consider me a casual gamer. If I need to build a new RTX PC its gonna cost me around 2000 USD. Then I ve to opt for an internet thats around 15 USD/month Then I ve to buy games from Steam/GOG/Game Pass etc Then when I play the PSU is sucking a hell lot of power too adding to power charge. Now consider a cloud gamer. 15 USD/ month + GFN 3080 Tier = 17 USD + a 300 USD 2k monitor+ lower power charge. I mean its just around 1/8 times the expense of building a PC with 8/10 performance of a native RTX 3080 pC. Its a no brainer for majority of casual gamers. Nvidia will lose market n hardware space.. But they might make up for it n digital space. The sole losers will be hardware vendors and PC builders and retailers.
heard latancy and ping wont be resolved untill the conccton between geforce now server farm and your local isp hub is upgraded to fiber optic. . . who knows when that will be
This is misleading because different internet providers will lead to varying results. The video doesn't say where the person is located and what ISP they are using, and what node they are using to travel the data stream across to the datacenter. Some household has worse connections than others, and Nvidia is trying to do the same gamble as Stadia, in a different way, but I also do not see this succeeding because of one factor. Declining network health as a result of more and more people trying to use the datacenter at the same period causing it to be strained and all the nodes in the surrounding network starting to become strained from increased packet activities. You don't see ISP in Canada replacing all their old copper lines with fiber optic, or upgrading all their old server hardware with ones that are higher performing. You are jumping from hops to hops with your packets hoping it's the shortest number of hops needed to reach the datacenter.
And damn, I wouldn't be able to play at 40ms ping. If I'm connecting to 20ms server I am whining already from being used to 10ms. Though, this was prior to 3080 sub.
Wow your connection is really bad. I'm on a pretty average connection for my country... Fibre to the home with 1000/1000 and wired ethernet. I'm at around 9-12ms on the 3080 tier depending on time of day, playing from a server in our neighboring country.
Latency is my biggest gripe with cloud gaming. 30+ latency is too high for me. In some games anything higher than that starts to become frustrating and unplayable. You will always be at a disadvantage to people playing natively.
You should have done Stadia comparison 1440p 60fps using stadia enhanced. GFN is 1440p max you should have set them the same it would have been a more fair comparison it's slower to decompress 4K than it is 2K? I bet the numbers would have been closer. And then if you factor in using the stadia controller connected to a 5G Wi-Fi it reduces the latency even a little bit more I will be willing to bet that Stadia at 1440p 60fps with the stadia controller connected to 5G Wi-Fi is equal to our less than GFN 3080.. ..
Agreed, however gfn 3080 tier is pretty freaking fast and runs games at much better settings then stadia. From everything I have seen they are on par with latency but gfn has much more pros(besides costing a good amount of cash lol)
@@Masta_E I wasn't making the argument to disqualify the power of GeForce now 3080 as I use it myself daily the point of my message was running Stadia at 4K increases the lag time because it takes longer to compress & decompress a 4K signal versus a 1440p signal and since this is a lag test video and not based on quality . It's important that they are both equal for the test to be accurate .✌️
Here's a request for a even better comparison can you do the Stadia free service versus GFN free service I'd be willing to bet that Stadia free 1080p 60 will be substantially better than GFN's free service because without the 3080 and the AMD thread ripper that you get for the $200 a year subscription the lag is substantially increased on GFN it would be great to see those numbers as well. The lack of lag is one of stadia's greatest strengths especially on the free service it's lower than the pro service more than likely.. Thanks.
Can you confirm that you were using the Stadia controller when you were testing with Stadia? Did you try measuring the latency on Stadia when it is set to 1080p mode? You should have mentioned that the GeForce Noww service costs you $200 per year whereas Stadia is free forever for 1080p once you have paid for the game. This is the most important point that you could have made in a comparison video like this.
1. Currently figuring out the best way to do this, its a bit more complicated when using a controller to do this sort of testing do to how the ldat reads the input, it could introduce more variables. 2. Isn't so much a stadia vs gfn thing just more of my use case and real world scenario that I wanted to compare but I get where your coming from and ill take it into account for the next video for sure.
Stadia free forever is great, but its one reason its not thriving. That's why Stadia free tier is not mentioned anywhere at signup. Stadia is trying to force new users to Pro. Users have to go through loopholes to access free. It's obvious Stadia is pushing users away from it. Getting the big games come at a price. That's why Xcloud and GFN have superior libraries. Maybe he should mention that too.
@@AnytimeAnywhereGaming it would be cool if you were able to test the gamepad through WiFi, as I've noticed it's faster than plugged in locally. So, I'm pretty sure Stadia can do better in some scenarios (where you use your controller directly connected to the server)
@@Manuel-rl6um He replied above saying that he did not use the Stadia controller in his testing, which is unfortunate in a latency comparison video because that is one of the major reasons that the Stadia controller exists, i.e. to reduce latency supposedly.
Interesting video.
I'd like to see how Doom eternal holds up across all 3.
Very fast paced, and is one of the few games I can think of that the developer specifically focused on reducing latency on Stadia, publicly, after launch
Oh great choice. I find destiny 2 is playable with the stadia controller but doom 2016 the latency is very noticeable. Almost unplayable lol. I live in Seattle near a stadia server and have amazing internet.
I get 4 ms to GeForce now i am in Montreal Canada. So remove 40 sec from your test and its the results i get :) (they opened a new server for east Canada) so my experience with GeForce now 3080 tier on par with your local hardware impressive.
that sounds amazing hopefully well see more servers in canada
J’habite en Gaspésie intéressant je vais prob avoir 15-20 ms
Insightful comparisons. Any idea how consistent your internet access latency was while you were testing? Was it during a busy time of day for internet use - or middle of the night when your neighbors aren't streaming, there's less congestion, and you get more consistent internet latency? Do you want to share who your ISP is? fiber or cable?
Extraction is bugged on GFN and it won't stream at 120fps despite showing 120fps in whatever ubisoft overlay and despite 120fps stream settings. It is as if vsync was ON on system level, the same goes for No Man's Sky. There are few games that are "bugged", maybe because there's some conflict with GFN's virtual display set at 60hz, or some other thing. But to this day they haven't fixed it. Basically, you are getting only 60fps over 120fps stream. Easily confirmed on 120hz+ monitor if you record a 120fps video. There's also a built in overlay ctrl alt f7 that shows the timeframe graph. So extraction isn't the best game to cover GFN input lag over 120fps stream.
Makes the result even more interesting, there's a big difference in latency when running the game all low and max. Also ran the fps counter on the ubisoft ui itself and it was giving me readings over 100 fps, you can actually see it in the video on the top right. Ill pass this info along to nvidia though to see what they say
@@AnytimeAnywhereGaming Nvidia knows about it, I made a reddit thread about this weirdness, shared it with GFN discord mods and guys at CGX channel. If you're enthusiastic about cloud and tech you'll be tempted to look into it even more. Take a loot at ctrl+alt+f7 graph in top left in different games for starters. Or capture locally 120fps video and go frame by frame to confirm 120fpa streaming.
There's no denying that if game is rendered at very high fps, even if afterwards is limited to 16ms frame time for 60fps, the input lag will be lower compared to the game running at 60fps and streamed at 60fps.
When it comes to r6 extraction i immediately noticed that it's not as fluid as other games, despite showing 120fps+ in game, then I started to dig into it and easily confirmed "60fps stream vsync".
Hi! Do you know how bit rate effects latency? I have a gigabit connection, so if 50 mbps is not a problem do I have any latency reasons to play on like 30-40 mbps? For me latency is over quality.
Latency will primarily be effected by your distance to the server your connecting to, if you go to your settings you can manually connect to a different one but it should auto connect you to the nearest server
@@AnytimeAnywhereGaming I have done it, my question is if there is any latency difference between 50 mbps bitrate setting and a lower one like 30 mbps.
@@dtamas No there isnt any difference
Best cloud gaming app you recommend?
Can you test Hunt Showdown please?
it would be nice to know what the test result of the free geforce now trial are
I'd like to see Cyberpunk 2077 tested. I don't notice any latency while playing the game, but I do in the menu's.
will do!
@@AnytimeAnywhereGaming Nice. Thanks mate.
@@AnytimeAnywhereGaming
I ve a genuine doubt. If this cloud gaming is so good is there any point in buying new GPUs???
I mean its far better to invest in a strong internet connection and good monitor rather than building a rig with say 3080 and i5 etc which would cost like around 2000 USD naa.
Is GeForce shooting in their foot by allowing this cloud Gaming? Cuz their own Hardware is gonna sit dusty at stores right!!
@@1989rs500 GPU's still have a market and will have one for quite some time, remember that geforce now has its limitations, not ever major title is on the service yet or arrives day one. Additionally new and stronger gpus will probably hit markets before they hit gfn servers as they did with the 3080 tier. There's also the argument to be made for work related pc's, gfn allows you to play games but not work on programs like premiere and such either.
@@AnytimeAnywhereGaming
Well consider me a casual gamer. If I need to build a new RTX PC its gonna cost me around 2000 USD.
Then I ve to opt for an internet thats around 15 USD/month
Then I ve to buy games from Steam/GOG/Game Pass etc
Then when I play the PSU is sucking a hell lot of power too adding to power charge.
Now consider a cloud gamer.
15 USD/ month + GFN 3080 Tier = 17 USD + a 300 USD 2k monitor+ lower power charge.
I mean its just around 1/8 times the expense of building a PC with 8/10 performance of a native RTX 3080 pC.
Its a no brainer for majority of casual gamers. Nvidia will lose market n hardware space.. But they might make up for it n digital space.
The sole losers will be hardware vendors and PC builders and retailers.
Should try Luna. I still think that has less lantency well for me it is
What games do you want to see me latency test? Regardless of platform or service, let me know your suggestions below!
Doom Eternal and The Division 2
Apex Legends and Rust
Doom, and counterstrike.
HOW ABOUT COMPRESSION i get a lot of compression is that normal? the game runs fast and great bu i hvae compression
heard latancy and ping wont be resolved untill the conccton between geforce now server farm and your local isp hub is upgraded to fiber optic. . . who knows when that will be
At least you entered your ping, thanks.
This is misleading because different internet providers will lead to varying results. The video doesn't say where the person is located and what ISP they are using, and what node they are using to travel the data stream across to the datacenter. Some household has worse connections than others, and Nvidia is trying to do the same gamble as Stadia, in a different way, but I also do not see this succeeding because of one factor.
Declining network health as a result of more and more people trying to use the datacenter at the same period causing it to be strained and all the nodes in the surrounding network starting to become strained from increased packet activities.
You don't see ISP in Canada replacing all their old copper lines with fiber optic, or upgrading all their old server hardware with ones that are higher performing.
You are jumping from hops to hops with your packets hoping it's the shortest number of hops needed to reach the datacenter.
And damn, I wouldn't be able to play at 40ms ping. If I'm connecting to 20ms server I am whining already from being used to 10ms. Though, this was prior to 3080 sub.
I would like see some single player games like Ubisoft games if your able to do so
Wow your connection is really bad. I'm on a pretty average connection for my country... Fibre to the home with 1000/1000 and wired ethernet. I'm at around 9-12ms on the 3080 tier depending on time of day, playing from a server in our neighboring country.
Latency is my biggest gripe with cloud gaming. 30+ latency is too high for me. In some games anything higher than that starts to become frustrating and unplayable. You will always be at a disadvantage to people playing natively.
LDAT results should not be compared to Ping lol. You have probably 100ms of latency playing on a local machine, the ping just says 30.
You should have done Stadia comparison 1440p 60fps using stadia enhanced. GFN is 1440p max you should have set them the same it would have been a more fair comparison it's slower to decompress 4K than it is 2K? I bet the numbers would have been closer.
And then if you factor in using the stadia controller connected to a 5G Wi-Fi it reduces the latency even a little bit more I will be willing to bet that Stadia at 1440p 60fps with the stadia controller connected to 5G Wi-Fi is equal to our less than GFN 3080.. ..
Agreed, however gfn 3080 tier is pretty freaking fast and runs games at much better settings then stadia. From everything I have seen they are on par with latency but gfn has much more pros(besides costing a good amount of cash lol)
@@Masta_E I wasn't making the argument to disqualify the power of GeForce now 3080 as I use it myself daily the point of my message was running Stadia at 4K increases the lag time because it takes longer to compress & decompress a 4K signal versus a 1440p signal and since this is a lag test video and not based on quality . It's important that they are both equal for the test to be accurate .✌️
Here's a request for a even better comparison can you do the Stadia free service versus GFN free service I'd be willing to bet that Stadia free 1080p 60 will be substantially better than GFN's free service because without the 3080 and the AMD thread ripper that you get for the $200 a year subscription the lag is substantially increased on GFN it would be great to see those numbers as well.
The lack of lag is one of stadia's greatest strengths especially on the free service it's lower than the pro service more than likely..
Thanks.
Can you confirm that you were using the Stadia controller when you were testing with Stadia?
Did you try measuring the latency on Stadia when it is set to 1080p mode?
You should have mentioned that the GeForce Noww service costs you $200 per year whereas Stadia is free forever for 1080p once you have paid for the game. This is the most important point that you could have made in a comparison video like this.
1. Currently figuring out the best way to do this, its a bit more complicated when using a controller to do this sort of testing do to how the ldat reads the input, it could introduce more variables.
2. Isn't so much a stadia vs gfn thing just more of my use case and real world scenario that I wanted to compare but I get where your coming from and ill take it into account for the next video for sure.
Stadia free forever is great, but its one reason its not thriving. That's why Stadia free tier is not mentioned anywhere at signup. Stadia is trying to force new users to Pro. Users have to go through loopholes to access free. It's obvious Stadia is pushing users away from it. Getting the big games come at a price. That's why Xcloud and GFN have superior libraries. Maybe he should mention that too.
@@AnytimeAnywhereGaming it would be cool if you were able to test the gamepad through WiFi, as I've noticed it's faster than plugged in locally. So, I'm pretty sure Stadia can do better in some scenarios (where you use your controller directly connected to the server)
@@Manuel-rl6um He replied above saying that he did not use the Stadia controller in his testing, which is unfortunate in a latency comparison video because that is one of the major reasons that the Stadia controller exists, i.e. to reduce latency supposedly.
Stadia to other countries like India please