2018 Speedtest U.S. Fixed Broadband Performance Report by Ookla
With gigabit expanding across the nation, fixed broadband speeds in the United States are rapidly increasing. Speedtest® data reveals a 35.8% increase in mean download speed during the last year and a 22.0% increase in upload speed. As a result, the U.S. ranks 7th in the world for download speed, between Hungary and Switzerland. The U.S. ranks 27th for upload, between Bulgaria and Canada, during Q2-Q3 2018. Though 5G looms on the mobile horizon, fixed broadband speeds in the U.S. continue to outpace those on mobile showing both faster speeds and greater increases in speed.
During Q2-Q3 2018, the average download speed over fixed broadband in the U.S. was 95.25 Mbps. Average upload speed was 32.88 Mbps. [...] According to Speedtest Intelligence data for Q2-Q3 2018, Comcast was the fastest provider in the U.S. with their XFINITY Internet receiving a Speed Score™ of 104.67.
Mean download speeds varied widely across the U.S. during Q2-Q3 2018 with the fastest state (New Jersey) coming in 139.8% faster than the slowest (Maine). The East Coast fared well with 5 states (New Jersey, Massachusetts, Maryland, Delaware and Rhode Island) and the District of Columbia ranking in the top ten. [...] Kansas City, Missouri topped the fixed chart as the fastest city in the U.S. during Q2-Q3 2018 with an average download speed of 159.19 Mbps. [...] Memphis, Tennessee came in last with an average download of 44.86 Mbps and Laredo, Texas was second to last at 55.37.
On average, U.S. consumers should have few complaints about recent increases in internet speeds. [...] As ISPs continue to build out their fiber networks and gigabit-level speeds expand we only expect to see internet speeds increase across the U.S. We'll check back in next year to see if all parts of the nation are experiencing the same improvements or if some states fall even farther behind.
Should consumers look forward to hitting their data caps now faster than ever, or are data caps being rolled back on gigabit connections?
Related Stories
Submitted via IRC for SoyCow1984
Ajit Pai's rosy broadband deployment claim may be based on gigantic error
Ajit Pai's latest claim that his deregulatory policies have increased broadband deployment may be based in part on a gigantic error. Pai's claim was questionable from the beginning, as we detailed last month. The Federal Communications Commission data cited by Chairman Pai merely showed that deployment continued at about the same rate seen during the Obama administration. Despite that, Pai claimed that new broadband deployed in 2017 was made possible by the FCC "removing barriers to infrastructure investment."
But even the modest gains cited by Pai rely partly on the implausible claims of one ISP that apparently submitted false broadband coverage data to the FCC, advocacy group Free Press told the FCC in a filing this week. Further Reading Ajit Pai says broadband access is soaring—and that he's the one to thank
The FCC data is based on Form 477 filings made by ISPs from around the country. A new Form 477 filer called Barrier Communications Corporation, doing business as BarrierFree, suddenly "claimed deployment of fiber-to-the-home and fixed wireless services (each at downstream/upstream speeds of 940mbps/880mbps) to census blocks containing nearly 62 million persons," Free Press Research Director Derek Turner wrote.
"This claimed level of deployment stood out to us for numerous reasons, including the impossibility of a new entrant going from serving zero census blocks as of June 30, 2017, to serving nearly 1.5 million blocks containing nearly 20 percent of the US population in just six months time," Turner wrote. "We further examined the underlying Form 477 data and discovered that BarrierFree appears to have simply submitted as its coverage area a list of every single census block in each of eight states in which it claimed service: CT, DC, MD, NJ, NY, PA, RI, and VA."
In reality, BarrierFree's website doesn't market any fiber-to-the-home service, and it advertises wireless home Internet speeds of up to just 25mbps, Free Press noted.
Related: Just How Rigged is America's Broadband World? A Deep Dive Into One US City Reveals All
FCC Chairman Ajit Pai Proposes Raising Rural Broadband Speeds
Speedtest.net Report Concludes That Broadband Speeds in U.S. Are Improving
It's Now Clear None of the Supposed Benefits of Killing Net Neutrality Are Real
FCC Struggles to Convince Judge That Broadband Isn't "Telecommunications"
Democrats To Push To Reinstate Repealed 'Net Neutrality' Rules
(Score: 2) by MichaelDavidCrawford on Thursday December 13 2018, @06:28PM (1 child)
"Don't buy version 1.0 of anything", as a support caller once advised me:
Ima gonna camp me out by the Apple Store next thanksgiving so I can be the very first to lay claim to my collectible.
But seriously: while the 4g works well in my iOS 10.0.1 iPhone 7, the throughput of its personal hotspot is poor, and is insufficient for HD videos.
Perhaps that's just a driver problem, but I'm avoiding firmware updates as I puzzle over whether to jailbreak
Yes I Have No Bananas. [gofundme.com]
(Score: 2) by takyon on Thursday December 13 2018, @06:32PM
5G iPhone will come a year later than other 5G phones, so you'll have to wait until 2020.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 2) by Runaway1956 on Thursday December 13 2018, @06:50PM (5 children)
Average speed, nationwide, is abysmal. So, a few super-fast lines are installed, thereby raising the overall average speed? Well, that's wonderful, if you happen to live in those three towns with super-fast, but all the rest of us still live with 2MB, or 10MB, or whatever.
Maybe I'm being a little bit unfair. Maybe. There is fiber being installed in some of the backwoods towns around me. But, so long as I suffer with 2MB, it's all meaningless to me.
(Score: 4, Interesting) by The Mighty Buzzard on Thursday December 13 2018, @07:01PM
My backwoods TN town's provider just started offering gigabit for $50/month ($15/month less than I'm paying for 100Mbps); no caps either hard or soft. Same ISP that has doubled my speeds twice in three years without a price increase and dropped the price on the year it didn't. That, my friends, is how you build customer loyalty.
My rights don't end where your fear begins.
(Score: 0) by Anonymous Coward on Thursday December 13 2018, @07:21PM
This result is biased in two ways. First is that it uses the mean, as opposed the median, or even a geometric or harmonic mean. It is a much bigger difference to move from 1Mbps to 2Mbps than it is from 20Mbps to 21Mbps, but they both have the same weight here.
Secondly, the kind of people who test their speeds are those who are more likely to have fast internet. Most people aren't going to test it unless they are getting slower than they feel they should or to confirm they are going fast. Once you confirm you are going fast, there isn't a need to check it again and if you are too slow, you'd call your ISP to fix it, who will then run a speed test once fixed. Therefore, the data will be biased to the last speed test, which is more likely to be fast.
(Score: 3, Interesting) by JoeMerchant on Friday December 14 2018, @02:02AM
In Florida they laid a high capacity fiber backbone down US-17, straight through many "backwoods" towns with populations around 10,000 each. It was YEARS before they saw any access through this fiber (though they also laid a gas pipeline along the same route at about the same time which did benefit the locals by ending the streak of acid-rain making sulfur coming out of the coal fired powerplant, when it was taken offline in favor of a cleaner gas burner.)
🌻🌻 [google.com]
(Score: 0) by Anonymous Coward on Friday December 14 2018, @04:30AM (1 child)
I would love a cheap low bandwidth option. Something like 2-10Mb $5 - $10 per month. I had ~10Mb down plan through Charter (quite a bit more than $10/mo), but they refused to renew it, stating it did not meet their new minimum speed standards... oh, and the new minimum plan costs 50% more. Guy on phone said, "You don't have a choice." I said, "Yes I do. Please cancel my service."
Been without internet service, at home, for a few years now (also affected some of the neighbors in the public housing across the street that used to use the open wifi I put up). Tethering works pretty well for low-bandwidth sites like Soylent.
But, it would suck to be in Runaway's situation too-- being willing to spend more, but having nothing better offered.
(Score: 2) by Runaway1956 on Friday December 14 2018, @03:35PM
I haven't seen $10 internet since dialup, 56k modems. My 2MB costs more than people in Philadelphia are paying for fiber optic. If I were only paying ten, fifteen, or twenty dollars per month, I wouldn't be so terribly pissed.
(Score: 3, Informative) by TheFool on Thursday December 13 2018, @06:53PM
I don't know, but they should certainly look forward to websites and content providers squandering the the extra bandwidth. Personally, I am looking forward to videos only being available in 4K and all websites including at least 50MB of random crap for "a better user experience".
(Score: 0) by Anonymous Coward on Thursday December 13 2018, @08:12PM (4 children)
There is no way average US home internet speed is 95 Mbps. Who here gets that at home? Mine is nominally 60 but usually slower.
We tested a new faster connection from Starry, but they did a sh*t install, and so we went back to Xfinity.
(Score: 0) by Anonymous Coward on Thursday December 13 2018, @10:49PM
Sure it is, all you need is 26 people getting 60 Mbps for every outlier getting Gigabit at home and you have an average of just over 95 Mbps.
(Score: 2) by JoeMerchant on Friday December 14 2018, @02:08AM (2 children)
Oookla just rated my Comcast at 13ms ping, DOWNLOAD Mbps: 70.68, UPLOAD Mbps: 5.89 and we're on a cheap-ish plan.
95 average does seem optimistic, but if you cook the numbers a certain way I suppose you might get there.
Instead of average or even median, I think the better measure of "access" would be the speed available to the bottom quartile, or even bottom 5% of people who are seeking internet access.
Now, what's more meaningful: landline speed available at home, landline speed available at work, wireless speed available during the workweek, or wireless speed available on weekends out in the boonies?
🌻🌻 [google.com]
(Score: 2) by canopic jug on Friday December 14 2018, @05:44AM (1 child)
A while back I read that most of the speed test sites are known and are simply given priority and left unthrottled. If you set up your own speed test, you'll usually find much lower speeds than what the well-known sites indicate you have.
Money is not free speech. Elections should not be auctions.
(Score: 2) by JoeMerchant on Friday December 14 2018, @05:35PM
A not unsurprising supposition. I have noticed that my speeds, on good days, tend to match what the bill says I'm paying for.
Of course, when you attempt connection to any site, your net transfer speed is never better than the worst of your speed, their speed, and the speed of every hop in between.
Our transfer speeds at work are laughably bad, with particularly ping times. Something that takes me less than 5 minutes to download at home takes a minimum of 30 minutes at work, and sometimes as much as 6 hours if there's a bad (read: Windows based) transfer protocol involved. One of the many prices of a central corporate security structure.
🌻🌻 [google.com]
(Score: 2) by cmdrklarg on Thursday December 13 2018, @08:38PM (3 children)
Last month I upped my 20/1Mb DSL to 40/5 for a few bucks more (I could get up to 80/10 if I paid more). Speedtest shows actual speeds of 39/4.5, so I'm happy. No caps either. Been with this company for almost 20 years since I signed up for a 512k connection to replace the 48K dialup.
I could go with Charter and get faster, but then I'd have to deal with Charter.
The world is full of kings and queens who blind your eyes and steal your dreams.
(Score: 2) by MichaelDavidCrawford on Thursday December 13 2018, @08:57PM
T-Mobile's business phone service, seventy Ducats per month.
I don't make voice calls a whole lot, but I really do use a lot of bandwidth, mostly by listing to YouTube videos.
Note I said "listening", after I start my first track I switch the screen off then set my phone on the shelf underneath the table just to one side of my easy chair, then I Troll The Tubes.
Yes I Have No Bananas. [gofundme.com]
(Score: 3, Interesting) by captain normal on Thursday December 13 2018, @09:26PM (1 child)
Try running the bandwidth meter at ping.eu. I'll bet it shows a lot less download speed. I think speedtest.net is run by Verizon (or at least tied to the big tel-cos).
Everyone is entitled to his own opinion, but not to his own facts"- --Daniel Patrick Moynihan--
(Score: 2) by cmdrklarg on Friday December 14 2018, @02:11PM
Well sure, if I'm testing from a European server.
No, I'm getting nearly 40Mb; tested it on some game downloads via Steam.
The world is full of kings and queens who blind your eyes and steal your dreams.
(Score: 2) by legont on Friday December 14 2018, @12:58AM
My little server in unfriendly jurisdiction gets the same speed as here but for 20% the price.
"Wealth is the relentless enemy of understanding" - John Kenneth Galbraith.
(Score: 0) by Anonymous Coward on Friday December 14 2018, @02:03PM
An interesting survey, I wonder what might be missing from any conclusions drawn.
Things that come to mind.
Cost of service and service caps
Any service provider special treatment for Ookla packets
What subset of customers does Ookla
This is good stuff, probably better than anything from the Fcc.
It might be nice to see stats from the services actually using the b/w. (G, FB, YT, NF, etc)