5G Wireless Test: Faster Than Google Fiber

HardOCP News

[H] News
Joined
Dec 31, 1969
Messages
0
Tests are showing 5G speeds at four times the speed of Google Fiber? If those results are accurate, that's pretty damn impressive.

The test involves the concurrent use of 24 devices on the 5G network. According to Huawei, the average download speed was 1.34Gbps and peaked at 3.6Gbps. That’s nearly four times the speed of Google Fiber, and almost twice the speed of Comcast’s new, incredibly expensive, fiber service.
 
In other related news everyone is enjoying their new 5G devices with their 3rd eyes and ears.
 
I remember how 4G was going to be so unbelievably fast....and I barely pull 20mb on it. And how many years until this is actually available?
 
Tested in China..

Because we all know we won't get true 5g here, hell we haven't even gotten true 4g yet. Couple that with crippling data caps and you will be able to use your phones mobile data for a whopping 5 minutes a month...Progress sounds great!
 
Just what we need faster data speeds but cut off at the knees data caps unless you want to sell your organs on the black market for higher caps.
 
That is nice but how is it going to be when there are actually countless people downloading pron at the same time?
 
That's great, people will be able to hit their cap in 20 minutes!

:confused:

You think carriers will be that generous with data caps?

Unless my maths are wrong:
Sustained (1.34 Gbps) results in a download rate of 171.52 MB/s.

2 GB Cap: 11.94 s
4 GB Cap: 23.88 s

Oh I hope I'm still on Verizon's Unlimited plan should "5G" ever come out.
 
Tested in China..

Because we all know we won't get true 5g here, hell we haven't even gotten true 4g yet. Couple that with crippling data caps and you will be able to use your phones mobile data for a whopping 5 minutes a month...Progress sounds great!

I finally experienced data cap. On a multi stop vacation and the hotel we stayed first in New Mexico had terrible wifi so brilliantly I tethered my verizon phone to my laptop to watch my college football team play. They got run out of the house so I turned it off during the 3rd quarter. Used 75% of my 8gb plan in 2 hours. WTF does it matter if we can d/l that fast if we use our $200/month data plan in 3 minutes?
 
Yay 5x faster than fiber and it will be rolled out where I live maybe in 30 years oh yeah and since cell companies control it it will have data caps so you know it will be completely useless.
 
:confused:

You think carriers will be that generous with data caps?

Unless my maths are wrong:
Sustained (1.34 Gbps) results in a download rate of 171.52 MB/s.

2 GB Cap: 11.94 s
4 GB Cap: 23.88 s

Oh I hope I'm still on Verizon's Unlimited plan should "5G" ever come out.

I honestly didn't put that much thought into it. I'm kind of glad I didn't.
 
The test involves the concurrent use of 24 devices on the 5G network.
I'm not a cellular wireless expert. Does mean 24 devices connected to a single 5g antenna? Is this realistic? Don't most cellular networks have to service far more than 24 devices? Would the bandwidth devide further and further the more and more devices are connected downloading (IE 48 devices downloading simulatenously means each would get half the bandwidth of the 24 devices)
 
Like most of the clickbait articles that are posted, I'm somewhat curious about the actual test. Here is a quote from digging through all of the clicks back to the original article.

Today’s trial represents the first large-scale Multi-User MIMO (MU-MIMO) technology test, with a concurrent connectivity of 24 user devices in the macro-cell environment on the sub-6GHz frequency band, in conjunction, it is also the first time to validate the performance of Sparse Code Multiple Access (SCMA) and Filtered OFDM (F-OFDM) in the field, both of which are 5G new air interface technologies proposed by Huawei. The cell average downlink throughput of MU-MIMOs is 1.34Gbps, with 3.6Gbps on download peak throughput in a 100MHz ultra-wide band channel; these speeds are more than 10 times faster than single layer Single User MIMO (SU-MIMO) technology. The joint trial of 5G new radio access technologies is a key step toward accelerating 5G standardization and commercialization.

Things to note, it's very unlikely in the US you'd be able to get a 100mhz wide channel, most of the spectrum auctions don't even have that much spectrum available for purchase. (The majority of spectrum is more like 10mhz wide as it's in the lower bands for range, not for speed.) That said this is actually pretty close to using 802.11ac with MU-MIMO which can do 80mhz channels.

I still can't decipher their wording, but if I had to take a guess they mean they achieved an average of 1.34gbps throughput for all devices combined. This means that per device you can get around ~57mbps of real bandwidth since some of the devices are transmitting at the same time. Taking a quick peak at the specs for AC if you were using 2 antennas with 80mhz at the highest index, you still cannot get 1gbps. So it's pretty unlikely they can claim 1.34gbps per device. (Even though that's what they want you to believe)

@serpretetsky: Nope, it's most likely not talking about a single antenna. MU-MIMO allows two devices to talk at the same time on the same frequency, but afaik it has to be to separate antennas. So this might be a grid of say 4 or 8 antennas meaning that 3 - 6 clients can transmit at once. I don't know what they would be using, but given that they had exactly 24 clients I would wager they used a multiple of that, like 3, 4, or 6. Meaning that it was setup so 4 clients could use the same air time, then that was divided into 6 timeslots of 4 clients. There is simply no good way to explain that without pictures, but if you look up info about MU-MIMO it might make sense.

Cell towers can certainly have more than 24 clients at a time, but I'm not sure what they actually run per antenna array. The thing to keep in mind is that most antennas are directional, so a tower can have 3 different arrays operating at the same time. So 3 of these arrays would mean that 72 clients could be supported by the same tower and not interfere with each other. Figure 5 on Page 6 of this document has a good picture of what it might look like.
http://www.unisonsite.com/pdf/resource-center/How Towers Work.pdf

You would be able to have 72 clients operate inside of a sector without an impact to performance. If you had even more clients then it definitely could slow down speeds per device. It would probably be less than half the speed if you doubled the number of clients needing to transmit at the same time because of the overhead.
 
Time to decide that communications is as critical as the highways and roads.

National WiFi/Cell system for $9.99/mo personal unlimited, usage based commercial use.

With 100% total coverage on every habitable square inch of our 3 million sq miles of the lower 48. No more "coverage areas" or dead zones.

Will require a jobs program for the 100,000's of employees of the wireless/telecom companies who will be displaced by a far more efficient 'Single Wireless System'. :eek::rolleyes::cool:
 
easy they do the test in china where they can make everyone in fifty miles turn off every electric device to find out the max throughput. With a wired transmission everything goes down a pipe and internal collisions do not happen because it has a channel to follow. The wireless takes interference from every other emi source including light, radio waves, in the states thousands of wireless devices. So you have two choices add more power and shatter those signals or run wired lines that have to be upgraded every time you need more speed. The fiber optic signal is limited by the compression algorithm, the coax is limited by the complexity of the sine wave. The Cat 5 twisted pair is limited by heat from eletricity melting the cables if you send too much down it.
 
Tek Syndicate guys said it best, in a perfect controlled laboratory it would be. "Like 802.11AC is faster than Gigabit. Only in very very very controlled circumstances."

True. But even at 25% the speeds they've accomplished, it'd be worlds better than what we have now... which is, at best, 25% of the speed we're supposed to be getting with 4G.
 
and peaked at 3.6Gbps....
I love how they measure the peak bandwidth of an "unstable" wireless connection against a stable wired one...
 
Problem with wireless is latency, you could download at 1TB/sec if latency is 100ms whos cares... well for gamers anyway.

4G is pretty good bandwith wise anyway, depending on location of course i can get 90mbps but latency is meh
 
Back
Top