4K HD on a mobile phone seems to me to be just as pointless as living in a big city and having a car that can do 400 kph.
3 Likes
Like 400kph cars which few of us have the capacity to drive let alone utilise in an urban environment. For high res fast moving video:
Finally, a human being (optical resolution of the eye being 1 arc second) can probably not process much more information than about 4 HD channels would pack (sound is trivial in information density compared to video, so is touch, smell), so there appears to be not much need for an individual to be connected at download speeds that substantially exceed the limit of about 4 HD channels incoming and 4 HD channels outgoing. Multiply this by the size of a family.
FFTH optical bandwidth likely already exceeds our biological capacity to process, only the price of the electronic side might remain a practical limit for a few years.
The techie bit can be found here:
1 Like
Maybe not on a 6" screen, but as some telcos are moving towards 5G dongles/hotspots to compete with the fixed line NBN. It is likely that the 5G will be used for IoT in the house including any media devices. Many IoT connect by wifi and not an ethernet (wired) LAN connection, so a 5G dongle or hotspotting device may satisfy the needs of almost most households. It also provides the flexibility of mobility where the high speed connection is not limited to the house, but anywhere where the 5G network coverage exists.
While fixed line NBN will suit some, 5G is likely to suit a lot more if the pricing is comparable to the NBN.
1 Like
Yes each mode on Fibre is capable of roughly 1 petabit per second (1,000,000,000,000,000 bits). If noted in terabytes, a terabyte is 0.008 parts of a petabit. A 10 terabyte sized driveās capacity is therefore 0.08 of a petabit and at the data rate of fibre would be filled in 0.08 ths of a second (if a drive was capable of storing data as fast as it arrived). 5G just (and just barely) tickles the edge of that speed.
From a 2016 article to put it into the realms of what a human brain could store:
āResearchers discovered that, unlike a classical computer that codes information as 0s and 1s, a brain cell uses 26 different ways to code its ābits.ā They calculated that the brain could store 1 petabyte (or a quadrillion bytes) of informationā
So in about 8 seconds a single mode fibre could transmit what is approximately a human brainās total capacity.
A 100 Mbps (0.0000001 of a Pbps) nbnā¢ fibre plan connection compared to all of what a fibre connection is capable of is just a very sad joke.
3 Likes
I meant a ethernet connected LAN. I have amended the above post to make it clear.
1 Like
āA lot moreā yes, but I venture to believe if one does not have reliable coverage and consistent response times from mobile (eg 5G) data whether stationary or however, even at near $0 pcm it is not going to sell as a substitute (let alone a replacement) for the NBN.
That which does not work is priceless, since no price will be paid 
5 Likes
Well I am sitting back watching the wired NBN vs wireless 5G battle with interest to see what eventuates. Which one I choose, donāt know yet.
Bit like watching the high def DVD vs Blueray battle, or way back the Betamax vs VHS battle.
Useability, availability, reliability, and price will determine my choice, not theoretical speed.
It would be if one had a choice.
No need to ponder 5G in itās higher speed incarnations.
In order of reliability here.
ADSL2+ over copper, 4G Mobile data from the next town or NBN Fixed Wireless from the other side of the pine plantation.
If I jump to Telstra I get to keep the copper fixed line and phone for free, but cannot port the number. If I stay with iiNet/Westnet I have to pay extra for the copper service to remain connected but get to keep the number.
I fully expect to get 5G when 4G is given itās use by date. But only on the lower speed spectrum of the current 700-900Mhz bands used for 3G and 4G rural services.
Singtel (our Optus owners) to Singapore users (Singapore Dollars are roughly the same as AU Dollars)
1Gbps at Singapore$40.00 per month (thatās the basic speed package)
10 Gbps plans at S$169 per month
2 Gbps at S$63 a month
Japan:
USD$55 a month for 10 Gbps
https://hothardware.com/news/japanese-10-gbps-internet
USA
A list of 19 ISPs who provide 1 Gbps services prices vary but AT&T offer a package at about US$80 per month
Just one of the selection āLike other top services like Google Fiber, Sonic gigabit internet is symmetric, so you get both 1000 Mbps downloads, and 1000 Mbps uploads. In most service areas, service is an affordable $70 per month.ā (my highlighting)
India:
NZ
They have āentry levelā 100 Mbps fibre plans then 200 Mbps then 1 Gbps then 10 Gbps starting prices from the 100 Mbps at NZ$60 to around NZ$105 a month for 200 Mbps. ATM 1 Gbps plans are starting at NZ$85 a month to around NZ$155 a month. No 10 Gbps plans are yet showing a cost but the rollout has commenced in some areas.
1 Like
According to an industry source, development of 5G has focused on economics. Physics, it seems, is now beginning to bite.
" Signals are hitting walls" 
āThe cost of the energy needed to power 5G has proved to be one of the biggest headaches for Chinese telecommunication companies.ā
4 Likes
Very interesting article.
It does not auger well for 5G in regional Australia.
1 Like
It is an odd headline, I donāt recall seeing any hype that promised 5G would have long range or not require many extra transmitters. As for deployment in outback Oz - tellāim heās dreaming.
1 Like
That pesky physics does get in the way of the hype.
In a similar way, microprocessor makers tried to make them run faster, but saw the power requirements go up exponentially and big cooling problems arise. Anyone remember the Pentium 2 CPUs? So much lower clock rates now, but multiple cores to get the same amount of processing done.
I reckon 5G will end up in general use as an incremental improvement on 4G; maybe 5 times more bandwidth per connection, and 10 times more connections per base station in the cell network, and similar range. Mmwave high speed 5G will only be for specific applications and areas due to that pesky range and signal blocking physics problem.
2 Likes
Maybe. The power demand is quite a shocker. Industry is already talking 6G. 5G might turn out to be a fizzer.
As an analogy, technology has changed with different fabrication processes, improved micro architecture, larger quality dies and higher density.
I donāt think the Intel CPU clocks or FSB are any slower now than 1998. I understand they are much faster.
What that says for 5G which is radio frequency spectrum limited is not clear. The physical elements available to manufacture CPUās has evolved enabling more powerful processors. For 5G the physical environment it operates with is very much a fixed quantity. The cost of manufacturing the radio and supporting equipment can improve. The RF spectrum is finite and so are the transmission losses over distance.
The whole article leaves it open as to whether it is really relevant to Aust. It does not dig down into any core statistics that enable meaningful comparisons. Telstra and Optus now have working 5G. It would be more useful to have an assessment of the power consumption needs of their solutions, and the corresponding customer service delivered. The last might offer a few clues as to just how many customers need to be in a footprint to make it pay.
P.S.
Isnāt 6G just 5G on steroids?
1 Like
In microprocessors, the equation is Power = Capacitive load X voltage squared X frequency. Designers hit the power wall many years ago. What did they do?
- Reduced the capacitive load, that is transitor to transitor connections called fanout by some fairly serious pipelining. Clock may be fast between pipeline stages, but the instruction may have to go through 20 stages.
- Reduced voltage. The early micros needed 12 volts. Later on 5 volts was common. Now, 1.5 volts is all that is needed. By the equation, voltage is squared.
- Reduced the frequency, or clock speed. And designers went to RISC designs, and used the extra space available to put lots of memory cache on the chips as well as multiple CPUs.
1 Like
A good friend who is a professor in specialises in smart technologies (has presented to senate committees on smart cities and technologies) is talking about 7G in about a decade. 7G is supposed to get over some of the barrier issues with 5Gā¦from past discussions 6G is an incremental step to 5G (like 3G and 4G), where 7G is a substantial leap.
You people are way behind. They are already talking about 10G. Google it.
Now seriously, the international standards have not even been set for 5G, let alone even looking at anything beyond. You want 5G now, prepare to accept a non-standardised protocol that could change. I think 5G will go off into a multitude of variations with increased capabilities. What the boffins are talking about with 6G or 7G will just be incorporated into 5G for at least a decade. We have seen this with standards like 802.11. Just add a suffix like a, or b, or g, to add capabilities.
But Japan doesnāt want to be left behind on 6G. Theyāve given up on the race for 5G.
Apparently, 6G will be a googolplex faster than NBN fibre. 
1 Like
