10Gbps routers consume more power than a Fridge!

[tsammyc]

I still can't see the justification to go WiFi 7. Not sure what is real world throughput when you're at a reasonable distance from the AP. Get reasonable speeds on uncluttered DFS channels

There isn't really a justification, until you go beyond 1Gbaps BB. Almost all WiFi 6 routers have 1Gbps ports, so it will be like nothing happened when you get faster BB. WiFi 7 can also be cheap - like buying Archer BE230s from Carousell for $100 and getting 2Gbps WiFi speed.

I gave away my WiFi 6 Asus AX3000 to a friend and set it up for him. My iPhone 16 Pro and Mac get the full 1Gbps on WiFi almost everywhere in his 800 sq ft flat even through one brick wall. 2 walls or more and it will fall to about 500Gbps.

 

[alex22]

this one at home, so can! whoever pays the electricity bills can be judge, jury and executioner.

Then how to apportion Usave?

 

[watzup_ken]

The first thing I noticed when I looked at my new router was the power brick that is rated 65W that is the same as a laptop. The outgoing router on the other hand that is 2.5G capable is rated for 24W I believe. I do agree that networking solution is getting very power hungry.

 

[bert64]

I still can't see the justification to go WiFi 7. Not sure what is real world throughput when you're at a reasonable distance from the AP. Get reasonable speeds on uncluttered DFS channels

The individual benefit for early adopters is small, the real benefit from any new tech is once everyone is using it - then you have much better sharing of the available wireless spectrum.

If you have a user running first-gen 802.11a kit on a 5hgz channel then they can basically hog the entire channel to get 54mbps (or usually even less) of throughput which will drag down the performance of users with modern equipment too. If you have a user with more modern equipment they will use a very small timeslice for the same throughput, leaving the rest available for other users.

Congestion matters a lot in Singapore. Someone who lives on a farm in the American countryside doesn't have to worry about congestion.

It's only in very niche cases that you even need 1gbps, if you have equipment capable of it try graphing your traffic usage sometime. You can stream multiple 8K videos with 100mbps, browsing websites is bursty and even a 100mb connection can download web content faster than your browser can render it. The only thing that really benefits is large downloads, and then only if you're actually sitting and waiting for it to complete rather than just letting it run in the background.

This is why ISPs push these higher spec packages, because they know you won't use more traffic than you would have with 100mbps but by giving you a higher burst rate they can charge more without it actually costing them any more.

 

[shiki_lee]

The first thing I noticed when I looked at my new router was the power brick that is rated 65W that is the same as a laptop. The outgoing router on the other hand that is 2.5G capable is rated for 24W I believe. I do agree that networking solution is getting very power hungry.

this 65w is power rating it can support does not mean your device suck that much of power. Alot power brick don't have PFC, efficiency will be low no doubt.

 

[shiki_lee]

There isn't really a justification, until you go beyond 1Gbaps BB. Almost all WiFi 6 routers have 1Gbps ports, so it will be like nothing happened when you get faster BB. WiFi 7 can also be cheap - like buying Archer BE230s from Carousell for $100 and getting 2Gbps WiFi speed.

I gave away my WiFi 6 Asus AX3000 to a friend and set it up for him. My iPhone 16 Pro and Mac get the full 1Gbps on WiFi almost everywhere in his 800 sq ft flat even through one brick wall. 2 walls or more and it will fall to about 500Gbps.

Asus AX3000 wifi can reach Full 1Gbps ???

 

[tsammyc]

Asus AX3000 wifi can reach Full 1Gbps ???

About 900+Mbps on Open Speed Test. My clients are connecting at Wifi 6 5Ghz 160Mhz, which is 2,402 Mbps. AX3000 is limited by its 1Gbps LAN port.

 

[BBCWatcher]

Also bear in mind that the more heat your networking equipment generates due to higher power draw the harder your air conditioner has to work (and the more power it draws) to transfer that excess heat outside. If you're properly measuring power consumption you should be taking these thermal effects into account, too. Unless you never use your air conditioner.

Faster isn't always better. Or even always faster.

 

[shiki_lee]

Also bear in mind that the more heat your networking equipment generates due to higher power draw the harder your air conditioner has to work (and the more power it draws) to transfer that excess heat outside. If you're properly measuring power consumption you should be taking these thermal effects into account, too. Unless you never use your air conditioner.

Faster isn't always better. Or even always faster.

Agreed, we should plan to put all ont, routers, switches all outdoor racking, only access point indoors to the room via STP. Don't need to waste energy to cool it.

 

[BBCWatcher]

Agreed, we should plan to put all ont, routers, switches all outdoor racking, only access point indoors to the room via STP. Don't need to waste energy to cool it.

You don't actually need Wi-Fi access points on the "cold" side of most homes.

On top of all this, if you're more frequently buying new equipment you're generating more e-waste. Every piece of equipment involves environmental impact throughout its supply chain and service life. If you buy more energy efficient equipment with longer lifecycles then you reduce these impacts. For example, an energy efficient wireless router that can run OpenWRT (and with reasonable RAM and flash storage capacities) should provide many more years of service than the typical junk ISPs provide and consumers buy. MikroTik gear also tends to be good in this respect. They support their equipment for a very long time, and they pay attention to energy consumption given the markets they cater to.

 

[1 Ghz]

Chelsio T520 also supports ASPM, but to get it working requires some cooperation from the motherboard.

I have that card with ASPM functioning on a Z390 board, but not X670E.

Tested the ConnectX-4 Lx EN MCX4121A-ACAT and ASPM works with the x670E.

Funnily enough, the integtrated NIC (Marvell AQC113CS) doesn't support ASPM ...

 

[wkweksl]

Tested the ConnectX-4 Lx EN MCX4121A-ACAT and ASPM works with the x670E.

Can check if it idles below C3?

 

[1 Ghz]

Can check if it idles below C3?

PC is using SSDs that don't support ASPM. Integrated NIC is Marvell AQC113CS that doesn't support ASPM.

No ConnectX4

ConnectX4

 

[wkweksl]

Wow nice, thanks for sharing! Better value proposition over the x710-da2.

 

[sglandscape]

PC is using SSDs that don't support ASPM. Integrated NIC is Marvell AQC113CS that doesn't support ASPM.

No ConnectX4

ConnectX4

Thank you for sharing this. When the time comes, I'll upgrade the CX3 to CX4, but my CX3 may not fail anytime soon

 

[overratedplot]

I had a power issue where an SSD suddenly used an additional 5W and did some further testing related to power usage. I'd like to correct that my CX3 also doesn't support ASPM. It turns out that the PC does go into C7/C10 after I updated the firmware, regardless of the CX3 ASPM status. Maybe this was coincidental . If I physically remove the CX3, total consumption goes down by about 5W, so that doesn't seem outrageous (it's a dual SFP module). Disabling ASPM increases total power usage from 26W to 40W, so that's pretty substantial.

 

[xiaofan]

Some discussions of Ubiquiti Unifi WiFi 6 and WiFi 7 APs being too hot.
https://forums.hardwarezone.com.sg/threads/wifi-8-is-coming.7081496/page-2

Typo, what i meant is with the heat WiFi 7 devices are outputting now....

Hopefully it will better. With my experience on U7 Pro XGS on 85 celcius, i now scare liao.

Even on Omada EAP 773 is on 45 celcius, EAP 783 on 55 celcius is quite high temp already.

Although i do admit one cannot deny that the heat output by those equipment which involves analog signal will have a high heat output

High heat = high energy loss also, so this is something that needs to be addressed. High power, low efficiency is bad for us.

This is also why I chose to stick with U6 Pro. As the hardware iterates, the efficiency gains come later. Also, as adoption goes up, costs come down.

So looks like the fanless design of Ubiquiti U7 Pro XGS is not so suitable for the weather in Singapore...

They claim to have learned a lot from the U7 Enterprise or other higher end gears to improve thermal.

Guess that is all marketing...

Actually there are also many reports over the internet claiming U6 Pro is hot ...

Yup. Their APs mostly are hot one even their WiFi 6 APs. Unsure why as well.

And yes, I agree high heat means high energy loss. Their U7 Pro XGS idle at about 13 to 14w, which I feel is decent for WiFi 7 APs but their APs are incredibly hot. I am unsure whether they sacrifice cooling for the look of the APs or they just didn't bother to optimize their circuitry design better. To me it is too hot for my liking, and it melted the 3M tape I used to mount. That was the trigger point for me to just get rid of the set.

Even their USW XG PRO 8 POE also idles at 75 celcius. I do admit network equipment runs hot la. Even my Omada set runs hot, but still it is like 45 celcius range and definitely not at the scorching range of Ubiquiti where u place your fingers for more than 3 seconds, u will burn your fingers.

 

[xiaofan]

From the first post:

The Dongknows website actually measure power consumption of the routers he tests over a 24 hour period. So I compiled all this tested routers, divide the results by 24 (since he measures over 24 hours) to get the W that they consume and reproduce that in a table below:

Asus BE88U 14.2W
Asus BE86U 11.7W
Asus BE96U 15.6W
Asus BE98Pro 23.3W
Asus BT8 11.3W
Asus BT10 15W
Asus BT16 22W
TP-Link BE230 8.1W
TP-Link BE550 12.9W
TP-Link BE800 23.5W
TP-Link GE800 24.4W
Deco BE25 8.1W
Deco BE85 20.2W
Deco BE95 22.3W
Netgear RS200 6.7W
Netgear RS300 10.4W
Netgear RS700S 15.6W
Eero Max 7 18.8W
Velop Pro 7 9.8W

Despite the high power consumption of the Asus and TP-Link flagship WiFi 7 routers, I seldom read about overheating issues, unlike Ubiquiti Unifi WiFi 7 APs.

I guess in the end size does matter.
https://forums.hardwarezone.com.sg/threads/wifi-8-is-coming.7081496/page-3

So far I cannot find good reports of overheating for Asus BCM4916 CPU based WiFi 7 router, including the big ROG GT-BE98 (quad band, dual 10G ports) and the "smallest" RT-BE86U (dual band, single 10G port).

So size does matter for thermal.

Same for the Tower Shape flagship TP-Link Archer WiFi 7 routers like Archer BE900/BE800 (except the SFP+ port) and Archer BE805.

Again size does matter.

Yup. Size indeed matters. You see the size of UCG Cloud fiber vs ER8411 Omada, so drastic difference.

Also, the USX XG PRO 8 POE vs SX3206HPP. ubiquiti one consists of 8 POE++ 10G ports, 2 SFP+ 10G ports while the latter only 4 POE++ 10G ports, 2 SFP+ 10G ports.... Form factor more or less the same.... So you will know which is hot af

I myself is attracted by the form factor of Ubiquiti tbh. So compact compared to Omada, till I experience the heat they produced. Totally change my view on them

 

[xiaofan]

Some data from Ubiquiti for the Dimension and Power Consumption Data (exclude PoE output if they have PoE function).

Need to get the Ubiquiti gear users to provide the typical power consumption data as the website only provides max power consumption data.

I will also try to collect some data from the Internet.

1) UCG, one of the most popular wired gateway
https://store.ui.com/us/en/category...ctions/cloud-gateway-fiber/products/ucg-fiber
Dimensions: 212.8 x 127.6 x 30 mm (8.3 x 5 x 0.5 ")
Max. Power Consumption: 29.4W (Excluding PoE output)

2) U7 Pro XGS
Dimensions: Ø215 x 32.5 mm (Ø8.5 x 1.3")
Max. Power Consumption: 29W

3) U7 Pro XG
Dimensions: Ø206 x 32.5 mm (Ø8.1 x 1.3")
Max. Power Consumption: 22W

4) U7 Pro Max
Dimensions: Ø206 x 46 mm (Ø8.1 x 1.8")
Max. Power Consumption: 25W

5) U7 Pro
Dimensions: Ø206 x 46 mm (Ø8.1 x 1.8")
Max. Power Consumption: 21W

6) U7 LR
Dimensions: Ø175.7 x 43 mm (Ø6.9 x 1.7")
Max. Power Consumption: 14W

7) U7 Lite
Dimensions: Ø171.5 x 33 mm (Ø6.8 x 1.3")
Max. Power Consumption: 13W

8) U6 Pro
Dimensions: Ø197 x 35 mm (Ø7.8 x 1.4")
Max. Power Consumption: 13W

9) U6+
Dimensions: Ø160 x 33 mm (Ø6.3 x 1.3")
Max. Power Consumption: 9W

 

[xiaofan]

Some data from Ubiquiti for the Dimension and Power Consumption Data (exclude PoE output if they have PoE function).

Need to get the Ubiquiti gear users to provide the typical power consumption data as the website only provides max power consumption data.

I will also try to collect some data from the Internet.

2) U7 Pro XGS
Dimensions: Ø215 x 32.5 mm (Ø8.5 x 1.3")
Max. Power Consumption: 29W

From Dongknows' test data.

U7 Pro XGS -- about 15W average
https://dongknows.com/ubiquiti-u7-pro-xgs-review/

Finally, with no internal fan, the U7 Pro XGS ran completely silent during my trial; however, it remained quite hot, similar to the U7 Pro Max. I generally couldn’t rest my finger on its underside for more than 10 seconds.