I’m doing some prep for a talk next week and just discovered a disturbing fact. I’m sure all of you will have read Cisco’s annual network traffic forecast. I covered it here.
What you need to know about is that Cisco reckons that by 2017, global IP traffic will reach an annual run rate of 1.4 zettabytes, up from 523 exabytes in 2012.
Somebody asked me what a zettabyte was and it made me wonder what comes after it in the big number naming or prefix stakes. I looked it up on Wikipedia and the answer is (of course !!) yottabyte.
The real problem comes when internet traffic outgrows yottabytes, which, dear reader, I assure you it will. I couldn’t find the name for the next one up. Wikipedia stops at yotta (1024 )! The world is facing prefix-exhaustion.
This is a problem akin to IPv4 exhaustion although the fix is far simpler. We have enough warning. We simply have to run a naming competition on this blog. It just won’t be acceptable to have more and more yottabytes. I could have a special mug produced as a prize.
I’m not going to bother with it yet (unless you really want to) but as we get nearer name exhaustion date we will have to give it some thought. It will also be a good excuse for another party like the end of the IPv4 bash I held a couple of years ago.
I would think it will happen sometime in the next decade – ie the 2020’s.
You heard it first on trefor.net…
PS I’d have put an image of the Cisco traffic forecast up but their flash code didn’t work and neither did the link to generate a jpg! We won’t get to yottabytes if we have errors like that now will we Cisco, eh? Probably just a glitch.
13 replies on “Disturbing news re internet traffic growth forecast – prefix exhaustion”
PS it ain’t going to be acceptable to have kiloyotta, megayotta etc.
Infinitebytes.. problem solved, forever.
But if I was American I might propose Obamabytes, just because it sounds good, bit like a burger chain.
MarkJ
ISPReview.co.uk
It’s Zettabytes, Yottabytes, Brontobytes then Geopbytes according to http://www.whatsabyte.com/ and various other Interweb sites. That said; these are damn large numbers and ones that clearly show that this Internet thing is here to stay! Martin
A little disappointing to hear someone else got there first Martin 🙂
I’m going for a lotabyte. How much data can your datacentre store a lotabytes 🙂
http://blog.sfgate.com/techchron/2010/05/24/google-gets-behind-hella-campaign/
as in, Hella lot of data
10^27
But yeah, SI standard only goes up to 10^24 🙁
Do what Russia and others have done with currency – revalue the byte as a kilobyte
One new byte is 1000 or 1024 (let’s not go *there*) old bytes.
As a minimum we should be thinking in terms of 64 bit bytes, not 8 bit.
That said are bronto and geop formally ratified?
Bronto already exists as a range of values between 10^15 and 10^27 (so could be confusing) and actually translates as ‘thunder’ so is applicable, perhaps, to describe a brontosaurus as a ‘thunder lizard’ but not sure how it relates to data (ah hah!, I just got it ‘thunder’ in the ‘cloud’ very good – but surely that should be preceded by Astrape[lightning]-bytes?).
However, whilst yotta may appear large – just think back to your ‘O’ level Physics, Tref, (remember there’s only one true science and that’s physics. No, chemistry is just applied Physics and biology is just applied chemistry) and remember Avogadro’s Constant (six point oh two two times ten to the twenty-three). That’s the number of constituent particles in a mole of a substance, so it could make sense to adopt the mole. Then, as appropriate, continue with kmol, Mmol et cetera. That should take us up to Ymol which is >½ the number of atoms in the observable universe – so finding extra materials to put stuff on could be the limiting factor rather than what to call it. Or are we hoping for dark-matter (or other exotic material) drives – that’s one for the scientists (and that’s Physicists in case you’d forgotten).
this was actually informative – not like most of what i see online. sharing 🙂
I propose….
Yummybytes,
Tastybytes, and
Gastrobytes
to take us from 10^24 up to 10^33.
The whole prospect becomes a lot more endulgent.
And then as we truly head for the edge of mankind’s imagination and need to slow down, the last category should be called ….
Nomorebytes.
Simple!
TFG
Good ones TFG. I suspect that by the time we get to Nomorebytes the density of data will be such that it will collapse in on itself and form a black hole. I therefore propose one further prefix called Collapsabyte in which everything tends to zero.
Note also (while I’m on a roll) a collapsabyte is clearly unstable and not self supporting. It requires the equivalent of a cosmic armchair or sofa to hold it together before it blacks out in the numerical equivalent of falling asleep in front of the fire.