Average Broadband Speed in US Rises Above 50 Mbps For First Time (techcrunch.com)

Internet speeds are getting faster in the United States, especially in cities such as Kansas City, Austin, Seattle, San Francisco, and Phoenix, according to a new Speedtest Market Report. The report, by Ookla's popular service, found that fixed broadband customers saw the biggest jump in performance this year with download speeds achieving an average of over 50Mbps for the first time ever. The result marks a 40 percent increase since July 2015. From a TechCrunch report: That average, 54.97 megabits per second is 42 percent higher than the same period last year, and upload jumped even more -- 18.88 is 51 percent higher year over year. This is all based on the 8 million or so daily tests conducted on Speedtest's website and apps, by the way, so the data is pretty sound. Comcast Xfinity took the honors for fastest speed on average, but its 125 megabits wasn't that much higher than the competition: Cox with 118 and Spectrum with 114. [...] On mobile, Verizon and T-Mobile are tied for first place with 21 megabits and change download speed on average, though the latter beats the competition by a long shot with upload speeds averaging 11.59 megabits. Poor Sprint, though.

Comments

This

[johnsmithperson123]

Is probably the result of people using mobile phone data instead of DSL .

Re:This

[gfxguy]

Why are people rating this funny? I was getting 25Mbs with my Comcast wired connection. While waiting in a parking lot one day, I decided to speed test my T-Mobile connection and got 90Mbs. I've since upgraded my modem with Comcast, but the point remains that I was getting faster data through my cell service than my wired connection (and for a lot less money, although the high speed data limit is 1/100th of what Comcast's is).

Re:This

[ITRambo]

Data caps are the real issue that is mostly ignored in the US. What is the point of having 50-Mbps speeds if you are charged for going over your (X-GB.month) data cap? It's far to easy to do so. We need a class action suit against the land based ISP's for this gouging tactic.

Re:This

[gfxguy]

I agree - especially when what most people signed on for was "unlimited," and especially if someone signed a long term (year or more) contract that the provider unilaterally changed after the fact. But really, instead of going the route of more government regulation, they should be attacking providers for dividing up territory to keep monopolies. If we had more competition, we'd have better prices and service. I would dearly love to dump Comcast, but now I'm getting 90Mbs, and nobody else in my area comes close.

I call fould

[mrlinux11]

I and my neighbors are nowhere near 50mb, I have the fastest and it is just 3mb's

Re:I call fould

[110010001000]

Me too. I live in Seattle and we only have 300 baud modems. We share them with 25 other people too.

Re: I call fould

[Anonymous Coward]

300 baud modem? My family of 50 has to share a connection, which is just cousin Cletus shouting "one" and "zero" from the top of a telegraph pole and listening for a response from cousin Billy-Joe standing on Comcast's roof.

Time Warner Maxx

[butchersong]

I'm not surprised. Time Warner, the largest provider in the states has been rolling out decent upgrades recently. I'm in Kentucky (not typically the best availability) and am at 300/20 mbps as of this year. Not sure if this is a trend with other ISPs

Thanks, Google

[pak9rabid]

I can directly attribute this bump in speed in my town (Austin) to Google Fiber. Before Google announced they were coming to Austin, the absolute fastest consumer-grade connection one could get was 50 Mbits, through TWC. As soon as Google mentioned their intentions to enter Austin with their Fiber service, TWC immediately started offering 100, 200, and even 300 Mbit plans, with plans for a 500 Mbit service level on the horizon. AT&T did something similar with their U-Verse service as well. Hell, I can even get these speeds in the next town over (Buda), where Google hasn't even announced they're going to go into. A little competition goes a long way.

Re: Thanks, Google

[ZeroWaiteState]

Yes, it absolutely has to do with Google. Otherwise they would just amortize the older DOCSIS infrastructure over more years and only upgrade to avoid equipment failure. Just because new tech is available doesn't mean there is any reason to deploy it if you have no competitors. In the telcos case they did fiber projects in order to get grant money, and Verizon sold off their fiber business to Frontier after the grant money ran low. I've seen other smaller fiber operators setting up shop because the incumbents aren't interested in upgrading their networks. This usually works out with the incumbent purchasing the small operator and driving the price back up, but Google is the first one I've seen that AT&T etc can't afford to buy out.

Lots of data does not mean representative

[Kjella]

This is all based on the 8 million or so daily tests conducted on Speedtest's website and apps, by the way, so the data is pretty sound.

So how many people on the same old DSL line run a speed test to check that there speed is the same as it was 10 years ago? People use speed tests when they got a new line, they've upgraded it or they're troubleshooting. They don't do it at random. Our national statistics here in Norway is based on collection of subscription statistics, which seems far more reliable as users would probably complain if they didn't get what they paid for. Last figures are 1,914,431 broadband connections, average of 40.2 Mbps with a median of 25.6 Mbps.

How is this measured?

[oneiros27]

They say this is "broadband" speeds, but broadband was redefined last year to require 25Mbps downloads.

So, someone could be sneaky and say 'oh, those 10 Mbps connections aren't broadband anymore', and you just drop out the lowest numbers, and miraculously the average goes up.

Schools were using this trick by keeping the poorly performing students from taking standardized testing to raise their test averages.

Re:Pretty Sound?

[Calydor]

*looks out window*

So this is what Seattle looks like? Weird, can't see the Space Needle from here.

...and mirrors

[fyngyrz]

My family has to stack wood so we can emit a burst of one's and zero's with our signal fire. The baud rate is terrible, and we keep getting parity errors when the blankets burn through. The cost of enough cords of wood to keep the connection up is horrific.

To be fair, they did try to put a telegraph line in, but the Smith's down the road a ways cut and burned the poles trying to watch porn, and so that never came about.

Re:...and mirrors

[Yvan256]

To be fair, they did try to put a telegraph line in, but the Smith's down the road a ways cut and burned the poles trying to watch porn, and so that never came about.

Well at least he did get wood.

For the first time?

[Eponymous Coward]

So do you think one day there will be a headline:

Average Broadband Speed in US Rises Above 50 Mbps Again

Keeping it on the down low.

[lazyron]

Well I'm doing my best to keep it down with my 13mb CenturyLink DSL line... and patiently (grrr...) waiting for Google Fiber.

Akamai says 15 Mbps

[TheSync]

The Akamai State of the Internet Q1 2016 has a US average Internet bandwidth of 15 Mbps, which is far more believable.

I agree that there are plenty of people in the US with 50 Mbps+ (I have that myself), but there are still a lot of people on the end of long DSL loops who will never get higher than 5 Mbps.

Why poor Sprint?

[Solandri]

I've been on an ancient Sprint unlimited data plan (true unlimited data - I've used up to 112 GB in a month with no complaints and no throttling) for $50/mo for over a decade now. They had a rough spot with the WiMax bungle, but I stuck with them gambling that they'd pull through. And they have. I almost always have LTE service now in Southern California. TFA lists Sprint's (and AT&T's) LTE coverage as 93% vs Verizon's 98% (then uses a graph whose scale is apparently 90%-100% to exaggerate the difference). It's the LTE coverage which is key. It was really painful when Sprint was down near 50%, but it's actually pretty rare for me to see my phone in 3G mode nowadays.

As for LTE speed, the average of the other carriers is 21.2 Mbps down, 9.3 Mbps up. Sprint's is 15.8 Mbps down, 4.9 Mbps up, or 75% down and 53% up vs the other carriers. Unless you're streaming 4k video to your cell phone, or regularly shoot a lot of videos and insist that they be uploaded to cloud backup immediately, these differences simply don't matter. They're all "fast enough" - they correspond to a few seconds or even a split second difference in most use cases.

The speeds are to the point where consistency (better coverage, fewer dead spots) is a more important factor. And by that metric there's now only a 5% difference between the best and worst mobile carrier in the U.S. Hardly worth the 2x price Verizon wants for service. That's why I gambled and decided not to give up my unlimited plan on Sprint by switching carriers. Once your coverage reaches about 90% (which was where Verizon was at when Sprint was around 50%), you're pretty close to maxed out. There's simply not much more improvement you can make. Whereas Sprint at 50% had a lot of room for possible improvement. (Your experience will vary with location. I hear Sprint still sucks in the Bay Area.)

(And if you're curious, no I'm not a bandwidth hog. My monthly data use is usually down around 1-3 GB. Just every now and then I go on a business trip or vacation, and use my phone as a hotspot so I and my family/friends can get Internet on our laptops and tablets. I'd have to pay $15/GB for overage if I switched to Verizon. The month I used 112 GB would've cost me over $1500. No thanks.)

Biased data source.

[jdharm]

Ookla/speedtest.net is used by people on fast connections to see how fast they can push data, not by people on slow connections to see how bad they are, and not by the general public to accumulate representative data. This report would be like going to a drag strip and then claiming that the data shows that the average American car does a 1/4 mile in 8 seconds.