Slashdot Mirror
Speed Report Finds T-Mobile and Xfinity Providing Fastest Mobile and Broadband Internet in US (geekwire.com)
Mobile and fixed broadband internet speeds in the U.S. are improving, but not all carriers and providers are created equal and not all areas of the country are benefitting equally from fast speeds. From a report: Those are the findings according to a new market report from Speedtest by Ookla out of Seattle, which relied on data it captured from user-initiated tests during the first half of 2017. And for customers using T-Mobile for mobile internet and Comcast Xfinity for broadband, the results are especially good. Speedtest credits infrastructure investments and upgrades as well as increased affordability of higher tiered packages for the fact that fast broadband keeps getting faster. The average download speed in the U.S. over fixed broadband during Q1 to Q2 was 64.17 Mbps (ranking 15th in the world) and average upload speed was 22.79 Mbps (24th in the world). Xfinity is the top provider when it comes to Speed Score -- which incorporates low-end, median and top-end performance for both download and upload speed -- with a score of 69.58. Speedtest says that Comcast has been aggressively seeding the market over the past year with advanced modems capable of delivering a more consistent experience for customers. The cable provider has also been increasing the amount of DOCSIS 3.1 channels in order to deliver faster speeds, according to the report.
Sorry, gotta ask, who is reviewing these articles before it gets posted. This feels a a lot like ads/spam on the site by a marketing/PR specialist. Whatever the original source, doesn't change an article that is in fact a promotion/ad.
"Imagination is more important than knowledge" - Einstein
When it comes to service, Comcast is at the bottom of the heap. I had them years ago, and I'd never go back.
The mean can be elevated by vastly improving speeds to a select few customers. e.g. If 10 people have service with an average (mean) speed of 10 Mbps, and you upgrade just one of them to 1 Gbps fiber, the mean speed of all 10 people will now be 109 Mbps, even though the mean speed of the slower 9 people is still 10 Mbps.
The median can only be elevated by improving the service speed for more than half your customers.
Microprocessor manufacturers find ways to game the system when it comes to benchmark tests on their products, and it's entirely likely Comcast and T-Mobile are doing the same.
I wonder how selective this is, or how a few good connections can raise a low of mediocre ones. I have Comcast that claims 25Mbps down but I routinely get 8-12, and use a Arris surfboard 6190 3.0 with 32 downsream channel bonding. Precisely 8 down are actually offered and they do not support 3.1. Not only that but I had problems with my cable to my house and it took 5 years to get it fixed properly because they had to run a line under the street (during which time I got as low as 0.32 Mbps. I seriously lost track of the calls (over 50) and every time it was a 15 minute exercise to show it wasn't me being a dumbass. Only positive thing was of the 9 techs they sent out, each one kept saying "What the hell is going on here, this is so screwed up. I'll fix everything and do it right. Believe me." Every FKN time.
I can understand regular media using the term "fast" to classify broadband characteristics, but as a tech site can't we do better? The term "speed" doesn't really make sense, especially when applied to bandwidth ( as is the case here ).
Using the same definition, a minivan full of thumb drives is "fast" in that it can deliver gigabytes of bandwidth, but I don't know that anyone would actually want to use such a connection.
An internet connection can be measured by 3 main metrics; latency, bandwidth and reliability ( well, and media if you want to get technical ). Perhaps we could start using that here?
Mod me down with all of your hatred and your journey towards the dark side will be complete!
I can hit a Speedtest.net server even when my general internet is dead with Comcast. They've obviously got an OOB route for Speedtest.net that allows that traffic to flow regardless of the actual status of the network. It also somehow always manages to get my full allotment of bandwidth even when I have several streaming video services running with Steam downloads which should be taking a chunk out of what can be used to hit Speedtest.net.