Page 1 of 1

Do Laptops Have The Same Latency as Desktops?

Posted: 09 Nov 2022, 17:33
by Mark Fable
Given similar CPU and GPU performance, would a USB 3.0 port on a laptop yield the same latency as a USB 3.0 port on a desktop when a peripheral is plugged in?

Re: Do Laptops Have The Same Latency as Desktops?

Posted: 16 Nov 2022, 14:28
by imprecise
Generally speaking, there shouldn't be a difference. Specifically speaking, every computer performs differently.

Re: Do Laptops Have The Same Latency as Desktops?

Posted: 17 Nov 2022, 08:33
by andrelip
With a Macbook (@ 280hz VG278QM DP) it was pretty much the same in terms of latency for the FPS that I had (200~300 CSGO).
I even tried in the battery as it's DC to check the "dirty electricity" theory.

Be aware that Notebooks usually have a lot of power savings, and may have power/thermal throttle. The Macbook I had even had VRM issues (2020- i9 9980h and RX5500m).
It ran fine for the first 6 minutes and them it dropped to the lowest clock possible.

The solution I had was to use a thermal pad for the VRM and a combination of MorePowerTool and QuickCPU to avoid crazy turbo-boosts (quickly overheating), then underclocked and set clocks to kind of balance the CPU-GPU usage for the games that I play.

In general I don't recommend notebooks for gaming, unless it was really a heavy and prepared one. You should also check if it has MUX Switch, otherwise, you are prob. going to have more input lag.

Re: Do Laptops Have The Same Latency as Desktops?

Posted: 18 Nov 2022, 18:09
by Chief Blur Buster
andrelip wrote:
17 Nov 2022, 08:33
In general I don't recommend notebooks for gaming, unless it was really a heavy and prepared one. You should also check if it has MUX Switch, otherwise, you are prob. going to have more input lag.
Using a desktop is better, but sometimes a laptop is needed to check off some boxes.

I have a Razer 240Hz gaming laptop in addition to my tower. I need to do 240Hz TestUFO PowerPoints from time to time, but it's nice to run Cyberpunk 2077 anywhere. Or show a specific game behavior.

It's good for convention travel too! DisplayWeek, CES, etc. Attending meetings with a laptop with a built-in 240Hz screen has some important benefits too.

And if you're a student, a gaming laptop can fit a university dorm better than many options.

In addition, if you're bringing stuff to a LAN game, and squeezing a small gaming monitor into your carryon baggage (it's possible) and your laptop in your personal-item backpack, then you can attend a LAN game with your own equipment with zero checked baggage. That can be a perk for some people!

Performance tip: Use a good-reviewed laptop cooling pad and plug in your laptop, to crank the frame rates at performances similar to the "same framerate performing" GPU on a desktop. Laptop GPUs are somewhat downrated, but brute helps: An RTX 2080 Max-Q outperforms a desktop GTX 1080, when the laptop is running off AC power, and you can use DLSS to help RTX ON framerates for solo gaming.

Re: Do Laptops Have The Same Latency as Desktops?

Posted: 13 Aug 2024, 01:26
by Mark Fable
Chief Blur Buster wrote:
18 Nov 2022, 18:09
andrelip wrote:
17 Nov 2022, 08:33
In general I don't recommend notebooks for gaming, unless it was really a heavy and prepared one. You should also check if it has MUX Switch, otherwise, you are prob. going to have more input lag.
Using a desktop is better, but sometimes a laptop is needed to check off some boxes.

I have a Razer 240Hz gaming laptop in addition to my tower. I need to do 240Hz TestUFO PowerPoints from time to time, but it's nice to run Cyberpunk 2077 anywhere. Or show a specific game behavior.

It's good for convention travel too! DisplayWeek, CES, etc. Attending meetings with a laptop with a built-in 240Hz screen has some important benefits too.

And if you're a student, a gaming laptop can fit a university dorm better than many options.

In addition, if you're bringing stuff to a LAN game, and squeezing a small gaming monitor into your carryon baggage (it's possible) and your laptop in your personal-item backpack, then you can attend a LAN game with your own equipment with zero checked baggage. That can be a perk for some people!

Performance tip: Use a good-reviewed laptop cooling pad and plug in your laptop, to crank the frame rates at performances similar to the "same framerate performing" GPU on a desktop. Laptop GPUs are somewhat downrated, but brute helps: An RTX 2080 Max-Q outperforms a desktop GTX 1080, when the laptop is running off AC power, and you can use DLSS to help RTX ON framerates for solo gaming.
Hey, sorry for resurrecting an old thread, but I had two things to clarify that have been on my mind for a while.

Firstly, is it correct to state that a laptop, in fact, does not stand at a disadvantage (latency-wise) when compared to a desktop? I believe this to be true, but I'm just trying to make sure that me gaming on a laptop is not a detrimental decision. I cannot come up with a logical reason as to why a laptop would perform differently to a desktop, other than the expected design-related power limits to avoid overheating in a smaller chassis.

Secondly, when people say that Nvidia Optimus adds input lag, at what stage is this latency added? This is how I understand it to work: When an external display is connected to the iGPU, the the frames rendered by the dGPU have to pass through the iGPU to be displayed, thus causing a bottleneck in the pipeline. If my reasoning is correct, then wouldn't the added latency be accounted for in frame-time measurements? For example, Optimus-disabled might yield 300fps (3.33ms) on average, while Optimus-enabled might only yield 280fps (3.57ms) on average. I'm just trying to make sure that the added latency in this case isn't some arbitrary, hidden figure.

Thanks in advance!

Re: Do Laptops Have The Same Latency as Desktops?

Posted: 13 Aug 2024, 10:10
by Chief Blur Buster
Mark Fable wrote:
13 Aug 2024, 01:26
Secondly, when people say that Nvidia Optimus adds input lag, at what stage is this latency added? This is how I understand it to work: When an external display is connected to the iGPU, the the frames rendered by the dGPU have to pass through the iGPU to be displayed, thus causing a bottleneck in the pipeline. If my reasoning is correct, then wouldn't the added latency be accounted for in frame-time measurements? For example, Optimus-disabled might yield 300fps (3.33ms) on average, while Optimus-enabled might only yield 280fps (3.57ms) on average. I'm just trying to make sure that the added latency in this case isn't some arbitrary, hidden figure.
Optimus style technologies can add some latencies that creates issues.

Latency is a big chain and a big rabbit hole, with tons of links at play, but focussing on a 300fps vs 280fps difference:

(1/280) - (1/300) = 0.000238sec = 0.238ms = 238 microsecond frametime lag difference

Even differences in display pixel response makes a much more massive difference than that.

Even differences in two display models (240Hz Model A versus 240Hz Model B) can be a bigger lag difference.

Yes, many youtubers have shown really bad gaming laptop lag. Maybe avoid that model and/or do some optimizations they didn't do (e.g. AC power + correctly selected additional very good cooling pad blowing in the same direction as the laptop's own fans (not incorrectly selected laptop colling pad).

Heck, most people don't realize that 200fps on a 480Hz OLED vs 240Hz OLED has a bigger lag improvement than this (massive reduction in scanout latency, even for low framerate frames). Most just buy MaxHz only barely above their max framerate, but part of that was because of LCD pixel response refinements of a more mature refresh rate, which can outweigh (e.g. very fast 240Hz LCD can outperform the slowest 360Hz LCD in esports). Just so many factors.

A good gaming laptop with good proper cooling, will only induce a minor latency penalty.
But also BUY A GOOD LAPTOP COOLING PAD <-- IMPORTANT!
And make sure any of your SSD upgrades have some good heatsinking too, etc.
And try to use an external gaming monitor for professinoal competitive gaming.
And always plug in your gaming laptop if you are playing critical esports games.
And uninstall the RGB animating software (or turn off during competitive gaming) as RGB software sometimes adds lag.
And use Performance Mode during competitive gaming.

Remember, laptop LCDs have slower pixel response, high-Hz laptop LCDs are there for convenience, 3ms GtG versus 1ms GtG even manufacturer VESA 10%-90% cutoff for just one color (black to white), still is a big difference for a lot of colors in real world pixel response speeds of the shades of colors in your game.

You don't want extra lag from thermal throttling.

Let's not forget that network latencies just... dominate latency problems. Especially when you throw in netcode quirks and sheninigians (including hidden automatic handicap algorithms for LPBs etc). If you're just an online gamer and not a LAN gamer, a good gaming laptop vs midrange gaming rig is generally a statistical dead heat.

At the end of the day...
A top of the line desktop gaming rig will provide better performance, but a good gaming laptop can outperform a mediocre gaming rig quite easily. Not every esports gamer can afford a $5000 - $10,000 rig, and I've seen many $2000 gaming laptop outperform $2000 gaming towers -- but YMMV. Gaming laptops are quite serviceable midrange competitive machines, if properly cooled & optimized. The nice thing is you get to take it with you, for the portable laptopping you need to do, or some less critical gaming on the go, or easier backpacking for LAN gaming, etc.

They DO have annoying behaviours like automatically switching to Intel GPU when running TestUFO (unless overriden), causing tons of stutters, and misleading laptop performance, especially on battery power. And there's bad models of gaming laptops. And you'll have really short battery lives on most gaming laptops. And sometimes you need to figure out which ports functions best (e.g. using a Thunderbolt hub to connect your gaming mouse, if the built-in USB ports are crappy... no need to throw away a perfectly good gaming laptop that has laggy USB ports; if you can just simply use your Thunderbolt dongle). Gaming laptops ARE a separate rabbit hole of optimization work.

Just know what compromises you will have, and know the right weak links to optimize. The laptop-vs-desktop in some budgets, once properly optimized with the right external accessories.

Re: Do Laptops Have The Same Latency as Desktops?

Posted: 13 Aug 2024, 19:37
by Mark Fable
Chief Blur Buster wrote:
13 Aug 2024, 10:10
Mark Fable wrote:
13 Aug 2024, 01:26
Secondly, when people say that Nvidia Optimus adds input lag, at what stage is this latency added? This is how I understand it to work: When an external display is connected to the iGPU, the the frames rendered by the dGPU have to pass through the iGPU to be displayed, thus causing a bottleneck in the pipeline. If my reasoning is correct, then wouldn't the added latency be accounted for in frame-time measurements? For example, Optimus-disabled might yield 300fps (3.33ms) on average, while Optimus-enabled might only yield 280fps (3.57ms) on average. I'm just trying to make sure that the added latency in this case isn't some arbitrary, hidden figure.
Optimus style technologies can add some latencies that creates issues.
Thanks for the detailed response! There's a lot to unpack here so I'll try to be as specific as possible. Regarding this statement, I have no doubt that Optimus adds latency, but what I wanted to clarify was that the latency added is DIRECTLY reflected by frame-time measurements. Hence, my prior example.

Another example to better elaborate my reasoning: Say Optimus-disabled gives you 400fps (2.5ms) on average, on high graphics settings. However, you decided to enable Optimus and lower your graphics settings, and you ended up with 400fps (2.5ms) on average. In my understanding, your net latency in both instances would be identical. I don't understand why some people make it seem like Optimus adds some hidden latency that can't be directly measured.
A good gaming laptop with good proper cooling, will only induce a minor latency penalty.
What is this "minor latency penalty"? If all else is the same for a given laptop and desktop pair (performance-wise, assuming equivalent performance from the CPU, GPU, RAM, and disk), where would the extra latency on the laptop (connected to AC power) come from?
But also BUY A GOOD LAPTOP COOLING PAD <-- IMPORTANT!
I have used multiple cooling pads that have fans directly facing the intake vent at the bottom of my laptop, and the benefits have been minimal. I found more success in using an air conditioner and a ceiling fan to cool my room, with a laptop stand that doesn't obstruct the intake vent and that raises the laptop high enough to benefit from the cooler room temperature and air flow.
And try to use an external gaming monitor for professinoal competitive gaming.
Definitely! I currently use the Asus VG259QM which has excellent scan-out latency measurements (1.9ms as seen on RTings.com) and response time measurements as well. 280Hz never felt better! I don't game on my laptop's monitor as I always found the screen too small.
And uninstall the RGB animating software (or turn off during competitive gaming) as RGB software sometimes adds lag.
Not sure of this suggestion, because my laptop keyboard (manufactured by SteelSeries) and my external keyboard both have RGB on, and I can't think of any way lag would be added that wouldn't be directly reflected by an increase in average frame-times. Not that I have noticed any, as I understand RGB software to be extremely lightweight...
And use Performance Mode during competitive gaming.
I always do! I also run fans at max speeds while gaming to keep temperatures as low as I can, and to avoid any form of throttling.
Let's not forget that network latencies just... dominate latency problems. Especially when you throw in netcode quirks and sheninigians (including hidden automatic handicap algorithms for LPBs etc). If you're just an online gamer and not a LAN gamer, a good gaming laptop vs midrange gaming rig is generally a statistical dead heat.
I always game with a wired Ethernet connection, and my ping is generally low. Packet upload/download rates are high and stable, and everything runs smooth. No issues here really.
At the end of the day...
A top of the line desktop gaming rig will provide better performance, but a good gaming laptop can outperform a mediocre gaming rig quite easily.
No doubts here! Laptop components are power-limited, by design, to keep them portable and to prevent overheating or excessive power consumption, so they will not be as powerful as equivalent desktop offerings. My only concern was whether laptops were functionally the same as desktops, which would mean that the only added latency on laptops is DIRECTLY due to their performance limitations.
And sometimes you need to figure out which ports functions best (e.g. using a Thunderbolt hub to connect your gaming mouse, if the built-in USB ports are crappy... no need to throw away a perfectly good gaming laptop that has laggy USB ports; if you can just simply use your Thunderbolt dongle).
What exactly do you mean by laggy USB ports? I've never heard of this. My laptop has two USB 3.0 ports and two Thunderbolt ports, and I've never had any issues with any of them, so I'm wondering if I'm missing something here...
Gaming laptops ARE a separate rabbit hole of optimization work.
Maybe I've just been using them for too long, but I don't find it any less tedious to optimize a laptop for its best performance, versus doing the same for a desktop. Much of what you need to do is straightforward, and as long as you're satisfied with the laptop's performance (mine performs extremely well), you can be competitive at any level with it.

Once again, thanks for your help in advance! I just need to separate the facts from all the misinformation flying around nowadays.

Re: Do Laptops Have The Same Latency as Desktops?

Posted: 14 Aug 2024, 07:07
by joseph_from_pilsen
I was playing at laptop and if you connect custom monitor (also owning vg-259QM), analog keyboard, some good gaming wireless mouse, the good laptops after HUGE optimizing didnt have measurable higher latency. But there is one but... You cant buy laptop with 3D processor for similar price like the desktop (except the one ultra overpriced 16 core ryzen laptop for outragious price).
In 2022 i was able to play on laptop at the same level like at desktop as 3D CPUs werent a thing and games werent so CPU hungry. In CSGO or Valorant were gaming laptops at par with desktops if you connected your own perripherials and tuned their settings a lot.
Now i cant (CS2 crap is somehow smooth only at 3D CPUs). If i were forced to buy a portable pc again to travel between locations, i would better build some ITX build and a monitor to every location.

Re: Do Laptops Have The Same Latency as Desktops?

Posted: 29 Aug 2024, 16:10
by Chief Blur Buster
Mark Fable wrote:
13 Aug 2024, 19:37
What is this "minor latency penalty"? If all else is the same for a given laptop and desktop pair (performance-wise, assuming equivalent performance from the CPU, GPU, RAM, and disk), where would the extra latency on the laptop (connected to AC power) come from?
Focussing on this lineitem -- simply because you can't get the same full power of a gaming rig. An RTX 4090 full performance isn't available in a laptop form factor, for example. Most of the lag will be from things like frametime latency. 200fps always has more frametime latency than 300fps does, no matter if laptop or desktop.

It's the same thing as saying a midrange gaming rig will have a minor latency penalty over a higher end gaming rig. However, it may be sufficiently imperceptible enough for your needs (Even in esports). Especially if the dominant latency is the network rather than the framerate, etc. YMMV.

Framerates of certain games can drop like a rock, if they push laptop in ways that makes it suffer much worse than a desktop. When framerates drop like a rock = that's a lag surge too, as frametime is a lag-additive to the total lag number.