Commerce Commission Measuring Broadband NZ results 2019

Forums › New Zealand Broadband › Commerce Commission Measuring Broadband NZ results 2019

1 | 2

Jase2985

13416 posts

Uber Geek

ID Verified

Lifetime subscriber

#2379401 20-Dec-2019 22:19

Sideface:

Page 4 of the report reads:

"Peak hours are the times when people typically use the internet; in New Zealand this is 7pm to 11pm on Monday-Friday.
Average speeds for all technologies do not decrease substantially during the peak."

This report excludes HFC cable.

Here is my SamKnows time-of-day local downstream speed graph for December ("UltraFast HFC Max").

This shows a significant peak-hours drop - although this is not obvious to a single user.

(click to view)

would you call 20% significant? seems a little over stated

how often is your connection not doing anything during those "lower" points on the graph, maybe the box is finding it hard to find a time to do the test

Lenovo

Shop now for Lenovo laptops and other devices (affiliate link).

michaelmurfy

meow
13214 posts

Uber Geek

Moderator

ID Verified

Trusted

Lifetime subscriber

#2379423 20-Dec-2019 23:53

I removed my SamKnows probe as it just can't do Gigabit. I know my connection is nearly Gigabit across most places and yet the probe would show way different.

It is OK on less than 200Mbit IMO but anything more it just struggles.

Michael Murphy | https://murfy.nz
Referral Links: Quic Broadband (use R122101E7CV7Q for free setup)

Are you happy with what you get from Geekzone? Please consider supporting us by subscribing.
Opinions are my own and not the views of my employer.

hio77

'That VDSL Cat'
12999 posts

Uber Geek

ID Verified

Trusted

Lizard Networks

Subscriber

#2379424 20-Dec-2019 23:56

michaelmurfy:

I removed my SamKnows probe as it just can't do Gigabit. I know my connection is nearly Gigabit across most places and yet the probe would show way different.

It is OK on less than 200Mbit IMO but anything more it just struggles.

this is exactly what i commented on earlier, networks are capable. but for some reason the samknows box just cant do it unless it has perfect conditions.

hio77:

the max results i suspect are CPE related, I've seen the results on these go from pretty variable to rock solid, with just a change of router.

Both routers were well within spec of gbit and the end user was more than capable of full gig on tests, the one swapped out probably one that's rated near the best around here... for a RSP device.

#include <std_disclaimer>

Any comments made are personal opinion and do not reflect directly on the position my current or past employers may have.

Sideface

9273 posts

Uber Geek

Trusted

Lifetime subscriber

#2379498 21-Dec-2019 08:37

Hammerer: Why are there a couple of hours missing from that graph? It starts at 1am and ends at 10pm instead of 12am to 12pm.

All of my SamKnows graphs are like that.
The graphs are generated in the UK - possibly time zones have something to do with it?

Jase2985: would you call 20% significant? seems a little over stated

how often is your connection not doing anything during those "lower" points on the graph, maybe the box is finding it hard to find a time to do the test

20% is not significant in terms of user experience, as mentioned in my post.
However it is much greater than the peak-hour drop listed in the report for wireless, xDSL and fibre technologies.

The SamKnows box will only test if the connection is not in use at the time.

Recently, when I was away for two weeks, with zero internet usage, the graph was the same.

Daily manual testing with the Ookla speedtest desktop app confirms the trend, but shows slightly higher speeds.

I get the same results using a Huawei HG659 or an EdgeRouter 4 router.

Sideface

k1w1k1d

1485 posts

Uber Geek

#2379514 21-Dec-2019 09:52

OK, looks like I won't bother re-volunteering my gigabit fibre connection.

Outside my field of knowledge, but I assume that most/all(?) fibre connections are capable of gigabit speeds so would handle 100/20Mbps without any issues? Wouldn't this make testing 100/20 rather meaningless???

(Corrected spelling mistake)

1 | 2