understanding modem speed
July 1, 2018 8:59 AM   Subscribe

I have a year-old cable modem/router. It has the latest technology and is a dual band. It has 4 wired ports as well. I have about 60Mbps download speed. I am trying to understand if it makes any difference in reaching that limit whether the devices are connected wirelessly (usually only my phone) or wired. The wired devices include the PC, PS4, Roku. These devices are rarely on at the same time.
posted by jtexman1 to Technology (6 answers total)
 
At your scale, functionally, no. Your download speed will be the bottleneck well before your equipment.
To put another way, when you're using multiple devices at once, the 60Mbps will limit how fast everything goes--not how the devices are connected.
posted by matrixclown at 9:28 AM on July 1, 2018


I have about 60Mbps download speed.

Per what? Have you tested this, or is this your ISP's claim?

I am trying to understand if it makes any difference in reaching that limit whether the devices are connected wirelessly (usually only my phone) or wired

There is ordinarily no incentive for an ISP to give a shit what is happening on the other side of the modem.
posted by pompomtom at 9:28 AM on July 1, 2018


Sorry for the sit. Yes I have tested the speed independently.
posted by jtexman1 at 9:37 AM on July 1, 2018


For some purposes, latency and packet loss will be a more important part of your experience than bandwidth (and latency and packet loss will both have an impact on bandwidth, depending on things like packet size, retransmission, and acknowledgement). Wireless will be worse on all these things than wired (unless you've got loose connections or sharp bends in your wires), but whether it's hardly noticeable or a real problem will depend on, well, pretty much everything. Router quality, device quality, wireless band, interference from your neighbours, interference from your walls, etc. etc.

It'll also depend on exactly what you're doing; someone playing hair-trigger games will worry a lot more about latency than someone streaming videos. Someone uploading and downloading files over a VPN will be affected by all of the factors in ways that I wouldn't care to try to predict.

As usual with benchmarking, the only benchmark that's worth anything is the one that matches what you're trying to do. Do what you're trying to do a bunch of times with a wire, and then do it again a bunch of times wireless, and time all your times, and see if one is any faster or slower than the other, or if one has more variation in speed than the other.
posted by clawsoon at 10:32 AM on July 1, 2018 [1 favorite]


It's hard to say if you're getting maximum speed, as this depends on how far the wireless signal has to traverse and the actual model of the router. I think the best way to go about this is to run an carefully controlled test to find your maximum possible connection speed (see below).

As the other poster said, I wouldn't even use 60 MB as a starting point as it's all too common for ISPs to make wild claims about their speed but offer something different. ISPs often oversubscribe their infrastructure, and last-mile line condition and length may affect the quality (especially true of DSL). I would toss 60 MB figure out the window and strictly go by your experiment results.

To do this, you need to (1) connect through Ethernet, (2) make sure no other devices are connected via Ethernet, and (3) disable wireless. If modem and router are separate, bypass the router and connect directly to the modem. Then run speed tests, preferably using several sites. I would suggest speedtest.net and fast.com, and take notes. Then plug everything back in the way you originally had it, and run the same sorts of tests, and take more notes. If you get similar numbers, you're golden, if not, you can turn devices off and on and change the setup to see what happens.

I've run tests like this in the past, and a couple of times with DSL setups I actually carried the modems outside to the demarc box and plugged them in with extension cords, and connected them to the box with the shortest, highest quality cable to get the best possible benchmark value. Attention to detail like this gives you the most accurate measurement of your maximum Internet throughput.

It might also be best to do this late at night instead of at 7 pm when your neighbors are usually hammering their servers, which will distort your results. Good luck!
posted by rolypolyman at 10:35 AM on July 1, 2018 [1 favorite]


For what you're doing, which seems to be TV streaming, games and computer use, 60Mbps is enough for lots of people to be on at once. An HD TV stream uses about 4Mbps, so quite a few people can be watching different shows before the connection starts to show any strain. The only problem you'll notice is if you're playing the type of game that's sensitive to latency - doing that and doing a big download at the same time could lead to increased ping times because of bufferbloat (higher ping is worse!).

Wi-fi can be more of a bottleneck than it appears if you're in a congested area or a house with thick walls. Even though a lot of routers are advertised as having quite high speed wi-fi connections nowadays, if next door are using the same channel it'll knock quite a lot off your speed. Same goes for running from one end of a house to the other.

To be honest, if you aren't noticing any problems doing what you need to use your connection for, it's not really worth worrying about the speed too much. If you are noticing buffering while watching video, or slowness while working or gaming, then you can start to look into it. You don't mention in your question whether you're facing any connection issues or if you just want to check whether you're getting what your ISP is advertising.

Perhaps I'm one of the lucky ones in that my ISP advertises my service as 100Mbps and the various speed-tester sites usually show it around 115Mbps, even at the highest of peak times.
posted by winterhill at 12:12 PM on July 1, 2018 [1 favorite]


« Older Overcake: how long will 238 pieces of cake last?   |   Pro-Choice Support In Under Represented Areas Newer »
This thread is closed to new comments.