Happy with, or perhaps dreaming about having, gigabit internet service at home? And that’s only 1,000 megabits per second (Mbps). How about 44 terabits per second (Tbps) service? That’s 44 million Mbps. Can you say, “Download a thousand movies a second”?
Really, one thousand movies? A second? Well, a typical movie might have a file size on the order of 5 gigabytes, or 5 × 109 bytes. So at that speed, in one second one could download 44 × 1012 bits / 8 bits per byte / 5 × 109 bytes per movie, which comes out to, yes, around 1,100 movies.
“Don’t bamboozle me with numbers,” I can hear you muttering as you read this column. And of course you have a point. Newspaper editors around the world couldn’t resist that 1,000 movies-a-second metric. After all, this is a moment when people everywhere are struggling to meet the demands of work-from-home technology – among them, internet bandwidth. And often for multiple members in a family, perhaps two parents and children, all vying for a slice of the home internet service pie.
Scientists in Australia recently pushed the boundaries to achieve that speed of 44 Tbps. This record-setting test was conducted in downtown Melbourne using what is described as “existing communications infrastructure.” Essentially a device being called a micro-comb, a highly specialized soliton crystal chip about the size of a fingernail, was used to send hundreds of separate laser or optical signals along existing optical fibre lines.
Each of these laser signals is independently controllable for several properties. The original research paper mentions that signals were sent over 75 kilometres of existing fibre cables. This is actually the key aspect of the research. It suggests that capacity on existing fibre in the ground can be massively increased. Were it instead some specialized, new form of fibre that was needed, this story would likely not have surfaced.
Such separate signals can either be combined for a massive bandwidth, as the news story featured, or, more practically, to give higher bandwidths to more users on a single cable. It is the latter that is really at the heart of the development. Let’s not expect to see 44 Tbps service to our homes any time soon, although I am reticent to say “ever.”
Even commercializing the micro-comb, as used in this test, is a few years off, but developments in these COVID-19 times have shown the earth needs all the bandwidth it can get. Besides, even without the extra demands of these past two months, world internet bandwidth consumption was growing at around 25 per cent annually.
In the wake of COVID-driven lockdowns, and the growth of the work-from-home movement, some major internet services such as Netflix and YouTube were either required to reduce the maximum or default quality of their data streams or did so voluntarily, in the process greatly diminishing their share of the overall worldwide bandwidth.
Demands placed by the widespread adoption of services such as Zoom may actually give us a good indication of what needs to be done to adapt to the demands of a 5G world, where far more devices than we have currently will be connected to the internet. In other words, don’t expect overall internet bandwidth usage to decline once a coronavirus vaccine is found and fewer people are working from home.
Most articles in the wake of the publication of the actual research conclusions tended to conflate two different things, backbone speed between devices in switching centres, where cost isn’t much of an issue, and end user speed. Backbone links, at least in the exchanges, are often 100 gigabits per second (Gbps) already.
Still, from 100 Gbps to 44 Tbps represents an increase by a factor of hundreds. It is an impressive accomplishment.
Here in North America we average about 50 Mbps for home download speeds. Certain countries such as Singapore and South Korea have average speeds approaching, say, 100 Mbps. Around 60 per cent of the world population has access to the internet in some form, but in some places the average speed is well below 0.1 Mbps. Downloading a movie at that rate might take half a day.
I should note that speeds on mobile networks are often faster than the terrestrial feeds to our homes and workplaces.
Existing average home internet download speeds are on the order of a millionth of the Australian test result. Even Gigabit download speeds, which are available from some providers in the Vancouver area, are around 40,000 times slower than the Australian test figure.
Results of the Australian experiment were published in Nature Communications under the title “Ultra-dense optical data transmission over standard fibre with a single chip source.”
According to one of the primary researchers working on the micro-comb project, “What our research demonstrates is the ability for fibers that we already have in the ground … to be the backbone of communications networks now and in the future.”
Meanwhile, I shall continue writing “straight into the cloud” with my pedestrian 600 Mbps service, although a quick speed test here on the porch, using fast.com, shows my older Acer Chromebook is actually getting 150 Mbps.
Follow me on Facebook (facebook.com/PeterVogelCA), on Twitter (@PeterVogel), or on Instagram (@plvogel)
