So Much for That Exaflood, Huh? Latest Cisco Report Again Shows Internet Growth Slowing Thursday May 30 2013 10:39 EDT Cisco's latest Visual Networking Index is chock full of interesting data nuggets on Internet growth, like the fact that annual global IP traffic will pass the zettabyte mark (1.4 zettabytes) by the end of 2017. According to Cisco, by 2017, global IP traffic should approach around 1.4 zettabytes per year, or 120.6 exabytes per month. Global IP traffic will reach 1.0 zettabytes per year or 83.8 exabytes per month in 2015. With a vested interest in selling companies the latest gear, Cisco has every incentive to make traffic patterns look as unwieldy as possible. As such, most of their highlighted bullet points from the report look very, very impressive: •Metro traffic will surpass long-haul traffic in 2014, and will account for 58 percent of total IP traffic by 2017 (thanks largely to CDNs, or content delivery networks). •CDNs will carry over half of Internet traffic in 2017. 51 percent of all Internet traffic will cross content delivery networks in 2017 globally, up from 34 percent in 2012. •Nearly half of all IP traffic will originate with non-PC devices by 2017. In 2012, only 26 percent of consumer IP traffic originated with non-PC devices, a number that will soar to 49% by 2017. •Traffic from wireless and mobile devices will exceed traffic from wired devices by 2016. By 2017, wired devices will account for 45 percent of IP traffic, while Wi-Fi and mobile devices will account for 55 percent of IP traffic. •The number of devices connected to IP networks will be nearly three times as high as the global population in 2017. There will be nearly three networked devices per capita in 2017, up from nearly two networked devices per capita in 2012. Once you get past all the splashy eye-grabbing bullet points, you'll note that Internet traffic growth is slowing down substantially across the board, especially on fixed-line networks. By and large, most of the growth rates I see in the report are dramatically lower than they were just a few years back in earlier Cisco reports. In the United States, Cisco predicts a compound annual growth rate of 26% for Internet traffic. That's massively less than what Cisco predicted previously (92% at one point), and for obvious reasons the company doesn't bullet point these adjustments. Cisco's latest numbers are an ever further cry from what telecom sector lobbyists and think tankers were predicting in 2010 and before, when they were using a looming "exaflood" to scare regulators and the press and public into buying into bad telecom policy. Companies like Nemertes Research and The Discovery Institute (the latter a PR firm paid directly by carriers, the former long accused of having a rather cozy relationship with AT&T) insisted we'd be seeing Internet "brown outs" by this point courtesy of unsustainable growth rates of up to 100% or more. The scary predictions were effective. Said lobbyists, think tankers, astroturfers and "fauxcademics" convinced many people that if the telecom industry wasn't given "X" (X being anything from fewer consumer protections and more subsidies to the right to bill by the byte or avoid network neutrality rules), that the Internet would collapse. That obviously never happened and intelligent engineers and networks adjusted, but few of the people who massaged data for their own financial ends over the last five to eight years were ever really held accountable. But hey, at least your tubes aren't clogged. |
FFH5 Premium Member join:2002-03-03 Tavistock NJ |
FFH5
Premium Member
2013-May-30 10:59 am
Growth slowed by switch from landline to cell networksAs IP traffic migrates from wired landlines to mobile cell networks the combined growth rate has slowed. And in no small part because of caps on wireless traffic. | |
| |
1 recommendation |
zod5000
Member
2013-May-30 11:52 am
Re: Growth slowed by switch from landline to cell networkssaid by FFH5:As IP traffic migrates from wired landlines to mobile cell networks the combined growth rate has slowed. And in no small part because of caps on wireless traffic. That's what i don't get. My cell phone is ok in a pinch but I never used it to do the heavy lifting. Even if I want to run updates on it, I do it over Wi-Fi on my home connection. It's still my home broadband connection that does all the heavy lifting. Even though I would agree that there will be portable devices using the internet than PC's, the different in bandwidth usage between the two should still be pretty large. Mostly because there are severe caps on cell phones coverage, and much more generous caps on home broadband. Maybe by lack of PC's they are referring to people using smart tv's, set top media boxes or game consoles to access netflix or other streaming video? I still can't see cell phone traffic exceeding home broadband any time in the near future? | |
| | | |
Re: Growth slowed by switch from landline to cell networksAgreed. Even though I'm using my cell for more tasks, most of the time it's connected to a wi-fi network, which is connected to a wired network. I only use data from my wireless carrier when I have no other option.
I think that a big reason for the slowdown in growth is because people and businesses have settled into somewhat stable patterns of Internet use. Also, most people and companies that can connect to the Internet and who have a desire to connect have done so. Granted, the speed they're getting may not be all they want, but it's probably the best they can do at the moment. The two groups left unserved are those who can't connect and those who simply don't want to connect.
Now, this isn't to say that there won't be an increase in growth in the future, but there needs to be a catalyst for it. For instance, if a company with multiple locations can get a very cheap and very fast connection to those locations, then you could see them use a VPN to merge their LANs into a single, larger network, and that could really increase data traffic. OTT video streaming could also drive up data, provided additional content becomes available. | |
|
| Sammer join:2005-12-22 Canonsburg, PA |
to FFH5
said by FFH5:As IP traffic migrates from wired landlines to mobile cell networks the combined growth rate has slowed. And in no small part because of caps on wireless traffic. Yes caps and the lack of network neutrality for wireless along with the the fact that our country isn't moving toward fiber to the premises is killing internet innovation in the USA. Welcome to third world broadband. | |
|
IowaCowboyLost in the Supermarket Premium Member join:2010-10-16 Springfield, MA |
Data is more efficientIf I look up a website on my iPad, it steers to a tablet optimized website on some sites.
Sometimes I prefer the full desktop version.
They are making websites more data efficient, especially in an age of data caps and slow speeds in rural areas.
Kind of reminds me of the '90s when websites had a text only version (such as someone on a 28k dial up connection) and a full graphics version (for those with the latest 56k modem).
Maybe web developers can have a full version for faster connections with generous caps (like cable or fiber) and a lesser version for those with slower connections or with limited caps (like DSL, wireless or satellite). | |
| | |
Corehhi
Member
2013-May-30 12:50 pm
Re: Data is more efficientsaid by IowaCowboy:Kind of reminds me of the '90s when websites had a text only version (such as someone on a 28k dial up connection) and a full graphics version (for those with the latest 56k modem).
Maybe web developers can have a full version for faster connections with generous caps (like cable or fiber) and a lesser version for those with slower connections or with limited caps (like DSL, wireless or satellite). I kind of forgot about all that and I don't know if the developers take into account what kind of connection the person has. Lite and full would be a good choice or if you could use some automatic flag that would be nice for some people. For those who don't remember all that it was a pain in the neck, I can remember I couldn't go to the comedy central site because the opening page had to many things going on like video and sound. It simply took to long to load up the first page, think that was when I had a 28K connection. | |
| | | IowaCowboyLost in the Supermarket Premium Member join:2010-10-16 Springfield, MA 1 edit |
Re: Data is more efficientMaybe software could be developed at the server level to detect the IP address and associated ISP that a request is coming from so they could direct to the lite or full version.
An example would be to direct a request to the lite version of the site if its coming from VZW BroadbandAccess (with low caps) or the full version if from Comcast.net (with high or no caps).
A lite version of a site could look more like a site from my high school and community college days (late '90s and early 2000's) and a full site could look like today's sites.
Same with streaming video, they could downgrade video quality on slow or capped connections. As for streaming video, I find it more suitable for short videos (like YouTube) but I don't find it to be a replacement for OTA or pay TV. In my opinion streaming video consumes network resources that could be used for other applications (node saturation) when video already has other resources allocated for its delivery (OTA/Cable/Sattellite).
I can tell because I have problems with dropped connections, especially in the evening. The problem is the node is overloaded with people streaming their Netflix.
Think of it as the information superhighway at rush hour: If you have too many tractor trailers with double trailers on the MassPike (Netflix) there will be no room for the cars (e-mail, web surfing, legal music downloads). Say you have a dedicated lane (cable, OTA, Satellite) on that highway for the double tractor trailers (1080p HD Video), then there will be plenty of room for the cars (e-mail, web surfing, legal music downloads).
Edit: Another advantage of cable/OTA/satellite is a provider can send a single stream of video to all of the receivers vs 600 or so separate streams over the same network of the same show.
It would be kind of nice if the cable ISPs could work with Netflix to deliver their streams over dedicated video channels to specialized CPE (such as a Netflix receiver connected to the cable line separate from the modem and fed into the customers home network via an Ethernet cable and they could allocate a QAM channel to Netflix streams). Or they could use a Netflix modem which would be a gateway that receives the standard DOCSIS channels plus has a special tuner to receive the dedicated Netflix QAM channel. | |
| | | | |
Re: Data is more efficientsaid by IowaCowboy:It would be kind of nice if the cable ISPs could work with Netflix to deliver their streams over dedicated video channels to specialized CPE (such as a Netflix receiver connected to the cable line separate from the modem and fed into the customers home network via an Ethernet cable and they could allocate a QAM channel to Netflix streams). Or they could use a Netflix modem which would be a gateway that receives the standard DOCSIS channels plus has a special tuner to receive the dedicated Netflix QAM channel. Uh, no thanks. Completely defeats the purpose of Netflix. You don't need a separate receiver or set top box/CPE.. you have your computer. Your service will get better when television providers start to realize that their old-fashioned television delivery models are on the way out, and start dedicating more resources to internet traffic. | |
| | | | rradina join:2000-08-08 Chesterfield, MO |
to IowaCowboy
Wouldn't it be a helluva a lot easier and more reliable to have the client add an HTTP header to the request? The device should know its connection speed or it can be configured by the customer to report a desired connection speed. For instance, someone's device might be connected over LTE but they still might want the low bandwidth option to prevent problems with their cap. With the client providing this information, the customer is also free to choose high bandwidth for one site and low bandwidth for another. | |
|
|
What about peak hours?While overall traffic may have increased by only 23%, how much of that increase is concentrated within peak hours? The more interesting data points should be peak-hours since that's what drives network build costs.
In past years, peak-hour traffic growth was in the 50-120% range depending on market and application. I'm surprised they omitted those figures from their summary this year. | |
| | |
Mojo 77
Anon
2013-May-30 2:41 pm
Re: What about peak hours?Globally, peak Internet traffic will grow 3.5-fold from 2012 to 2017, a compound annual growth rate of 29%. | |
|
sivranVive Vivaldi Premium Member join:2003-09-15 Irving, TX |
sivran
Premium Member
2013-May-30 10:17 pm
Discovery Institutequote: Companies like Nemertes Research and The Discovery Institute (the latter a PR firm paid directly by carriers, the former long accused of having a rather cozy relationship with AT&T)
So the Discovery Institute's branching out from their anti-science, anti-education efforts and getting into PR now? Well, at least their expertise in bullshit will serve them well. | |
| fg8578 join:2009-04-26 San Antonio, TX |
fg8578
Member
2013-Jun-12 1:55 pm
Moore's Law for data?"Because of smartphones, tablets, social media sites, e-mail and other forms of digital communications, the world creates 2.5 quintillion bytes of new data daily, according to I.B.M. The company estimates that 90 percent of the data that now exists in the world has been created in just the last two years. From now until 2020, the digital universe is expected to double every two years, according to a study by the International Data Corporation." » www.nytimes.com/2013/06/ ··· ach.html | |
|
| |
|
|