Another shop goes 100% Linux with CachyOS!

[Sky-Knight]

You do realize that WiFi is a hub right? That slower devices in range, if they are yours or anyone else's slow the entire network?

That's a lot of downside just to not spend $20 on a USB doodad that solves it forever.

 

[Blues]

I had always understood Wireless networks to be capable of the max speed the broadcasting HUB supported and/or the max speed of each devices your description suggest a single device operating at a lower speed limits the speed of all devices. If this is a long standing misconception I am glad to have the clarified though I believe many smart cameras and other smart appliances often use slower receivers so this would mean many home with smart devices are being throttled by those devices. I would say its not a matter of $20 as most adapters I see currently are $40+ to improve this one devices performance and even still the router is at least 5 years old so how much I can improve my speeds is limited there also. I would need to spend at minimum to upgrade the network fully at least $100 in the whole picture not a lot but I am a "If it ain't broke why fix it" mentality person. My setup has worked I haven't had any issues where my speeds were a problem for me until trying to switch to Linux and finding my WiFi card in the PC is not supported under basic precompiled drivers.

 

[britechguy]

I am a "If it ain't broke why fix it" mentality person.

Which saves you, and your clients, a boatload of pain and money.

If you (generic you) are happy with "the way things are" then don't let anyone convince you that you need to change out your hardware.

As I've said many times, excess capacity that you never actually use/exploit, regardless of what, specifically, that capacity involves is a dead asset.

 

[britechguy]

Gemini: [Does a WiFi network use speeds for all devices based upon the throughput capabilities of the slowest device connected?]

Perplexity.ai: [Does a WiFi network use speeds for all devices based upon the throughput capabilities of the slowest device connected?]

Copilot: [Does a WiFi network use speeds for all devices based upon the throughput capabilities of the slowest device connected?]

ChatGPT's answer to this prompt:
Short answer: No — not exactly.

A WiFi network does not permanently drop everyone to the maximum speed of the slowest device. However, slower devices can reduce the overall performance available to others, especially on older WiFi standards.

Here’s how it actually works:

---

1. WiFi is shared airtime, not shared speed​

WiFi works on shared airtime, not equal bandwidth allocation.

Each device gets time to transmit. A slow device:

• Uses older modulation rates
• Takes longer to send the same amount of data
• Occupies the air for more time

So while faster devices can still use high data rates, they get less available airtime because the slow device is "talking" longer.

Think of it like:

Everyone shares a microphone.
A slow talker doesn’t slow your speaking speed —
but they hold the mic longer.

---

2. When does the “slowest device” matter more?​

It depends on WiFi generation:

Older standards (802.11b/g)​

• Mixed networks could drop into protection modes.
• Performance impact was much worse.
• A single 802.11b device could noticeably slow the network.

Modern standards (802.11n/ac/ax — WiFi 4/5/6)​

• Devices connect at their own negotiated rate.
• Routers manage clients independently.
• Impact still exists, but it’s far less dramatic.

---

3. What actually causes slowdowns in real life?​

Often it’s not just “slow device = slow network.” It’s:

• Weak signal devices retransmitting packets
• Far-away devices using very low modulation rates
• Interference
• Too many devices competing for airtime
• Legacy WiFi standards still enabled (like 802.11b)

A device far from the router can hurt performance more than an old device close to it.

---

4. Modern routers use smarter management​

Newer routers may use:

• MU-MIMO
• OFDMA (WiFi 6)
• Band steering
• Airtime fairness

These features reduce the “slow device hogging the network” effect.

---

5. Example scenario​

If you have:

• A laptop connected at 866 Mbps
• A smart plug connected at 6 Mbps

The laptop can still transmit at 866 Mbps.
But when the smart plug transmits, it takes much longer per packet, slightly reducing total network efficiency.

---

Bottom line​

WiFi does not cap everyone to the slowest device’s speed.
But slower devices can reduce overall efficiency because WiFi shares airtime.
Modern WiFi minimizes the impact significantly.

---

If you'd like, tell me:

• Your router model
• WiFi version (if known)
• What devices you're worried about

I can give you a more tailored explanation.

 

[Sky-Knight]

Yay for AI slop and lies...
Modern WAPs have multiple radios, and one radio will drop to support just the beacon of an older device in range. Each radio is a logical network hub, anyone else on that radio will slow down to the older most accepted standard. Now if you have enough spare radios this is fine. But most WAPs are 4 radios or less. How many devices do you have again?

The only way to avoid this is to configure the radio to not respond to older standards.

The AI even mentions this: Legacy WiFi standards still enabled (like 802.11b)

But fails to surface the devastating truth. You do not have a network "Switch" in the air, you have a hub... and because of that, the performance issues suffered go exponential. Spend the $20, get a new wireless device, and go configure your WAP to ignore older standards. You'll thank yourself, and me later. Because having a radio go back in time causes issues... painful ones... in performance, security, and reliability.

 

[britechguy]

The AI even mentions this: Legacy WiFi standards still enabled (like 802.11b)

And how many of us have these enabled in our WiFi routers? I certainly haven't for quite some time, and that was in the out of the box state.

The answer is pretty clear, from multiple sources, and your concern is overblown, to put it mildly.

 

[Sky-Knight]

And how many of us have these enabled in our WiFi routers? I certainly haven't for quite some time, and that was in the out of the box state.

The answer is pretty clear, from multiple sources, and your concern is overblown, to put it mildly.

And back to ignoring you because I'm done with the rage bait.... you cannot handle being wrong... and I've called out your ignorance in the form of AI slop, because it is... exactly that.

"multiple sources" when you reference AI... my goodness... it's almost like you don't have a clue how those models work and are trained. This is one of those "common knowledge" things that is utterly incorrect out in the world. I'm done here, I'll just tell you to go strike up a chat with a qualified RF engineer.

 

[YeOldeStonecat]

I had always understood Wireless networks to be capable of the max speed the broadcasting HUB supported and/or the max speed of each devices your description suggest a single device operating at a lower speed limits the speed of all devices. If this is a long standing misconception I am glad to have the clarified though I believe many smart cameras and other smart appliances often use slower receivers so this would mean many home with smart devices are being throttled by those devices.

Not in modern times. Not even slightly older times.
With features such as...MIMO....and....mixed mode....wireless radios no longer have to downshift "everyone"...to the lowest connection speed. The phrase "a chain is only as strong as its weakest link" is not true here.

So you can have a modern wifi 4 or 6 or 7 wireless network...and if some old wireless N client connects....it will not force all other clients to max out at wireless N speeds. Of course, with the latest wireless APs you'd have to use mixed WPA2/WPA3 mode...to allow legacy clients such as A/B/G/N....instead of requiring WPA3.

Now...technically...on busier radios with a lot of clients...a legacy client or a handful of legacy clients can very slightly slow things down due to them being able to talk only half duplex and...have very slightly longer "airtime" to talk back 'n forth. So while modern clients are still connecting with modern speeds...an AP might, on paper, barely measurable, be slowed down a tiny bit. But it's still the same thing..."airtime"...that a poorly designed wireless network that allows wireless clients to connect at very weak signal strength...has to impact the performance of others due to longer hangtime of airtime.

 

[britechguy]

Not in modern times. Not even slightly older times.

But, of course, we're both wrong about this, as are multiple sources (yes, I check the source's sources when AI searches are involved) that clearly indicate why the assertion that a WiFi network will only be as fast as its slowest participant is wrong.

This is even more true when the slowest participant isn't "chatting frequently," and that is very often the case. A temporary pause for someone that "speaks slowly" is no different in a WiFi network than in human conversation. It doesn't bring the whole conversation to a halt nor does it make all participants speak at the speech rate of the slowest speaker.

But heaven forbid we talk about the facts on the ground as they have been for some time.

 

[YeOldeStonecat]

But, of course, we're both wrong about this, as are multiple sources (yes, I check the source's sources when AI searches are involved) that clearly indicate why the assertion that a WiFi network will only be as fast as its slowest participant is wrong.

This is even more true when the slowest participant isn't "chatting frequently," and that is very often the case. A temporary pause for someone that "speaks slowly" is no different in a WiFi network than in human conversation. It doesn't bring the whole conversation to a halt nor does it make all participants speak at the speech rate of the slowest speaker.

In the "old old days of wireless "we're going back pretty far now"....before MU-MIMO...there was SU-MIMO. SU-MIMO was not named at the time..but it's basically single radio that can only handle each and every wireless client one-by-one. True half duplex. So back in those days of SU-MIMO, a legacy wifi client...bogged everyone down....due to ALL clients being on the same antennas/radio.

MU-MIMO (I'll shorten it to MIMO)...unofficially came out in the early days of pre-wireless N. And officially came out with wifi-5 80211.ac.
MIMO = multiple-input multiple-output. Using multiple antennas...and multiple radios....an AP can separate groups of wireless clients. So say someone has an old old Windows XP laptop with a wireless N wifi adapter....it can latch onto a modern AP...and...have its conversation unique to the network...separated from other wifi clients...and those other wifi clients can connect to other antennas/radios at their higher modern speeds....and never deal with the "speaks slowly" client in any conversation. Well, I suppose we can come up with a situation of....the legacy wireless client is sharing a folder on its hard drive ..and others have to talk to it on its slow leg of the AP. But I'm feeling pretty good saying...these wireless clients will just be hitting the internet, perhaps a networked printer...so...the fact that there is some legacy wireless client on an "N" adapter...means nothing at all to the other 75 wireless clients on...wifi-5 or newer. No slowdown for them at all.

 

[britechguy]

@YeOldeStonecat

Indeed, there are absolutely always edge cases, but that's exactly what they are.

Things have been as you described them (and every AI Chatbot did similarly) for a very long while now. WiFi has not been a "pipeline only as speedy as it's slowest participant" for quite some time now. The issue with it being so was recognized quite early in the progression of the protocol. It's been fixed.

 

[Blues]

I deal in mostly Residential and Home Business/Small Business who use off the shelf retail equipment mostly so based on @britechguy & @YeOldeStonecat it sounds like it is being blown a bit out of proportion and my understanding while flawed doesn't really change the effective facts for my clients or even myself since this is for home personal use

 

[britechguy]

I'm done with the rage bait....

The fact that you consider a very measured response, using multiple sources to be rage bait says more about you than about me.

I was correct, you were incorrect, it's that simple. Posting corrections is part and parcel of what a group like this is about. Accuracy of information is paramount. I always make sure that anything I say is well supported, if it's not a matter of opinion, before I say it. Would that you would do the same.

Using AI Chatbots isn't lazy, either, when they put it all well, and their source material is actually valid, like it was here. I actually said, elsewhere, the other day that using AI Chatbots requires a meta version of "consider the source." In the case of AI output, you must consider the source's sources. I routinely follow my own advice when evaluating AI Chatbot output.