Govt network operator launches site for testing your home, or office, broadband speed. Um, why?
July 11: Using Reannz new broadband testing site at home, I found it recorded download speed (7.58Mbit/s) off-peak – a little lower than Speedtest.net, which habitually comes in around 8-9Mbit/s offpeak, and upload speed fractionally higher.
Some have argued that Speedtest.net (which has rather a lock on the popular geek imagination) is perfectly adequate.
Why should Crown company Reannz emulate it?
I think the answer is that as well as presenting your result in megabits, the Reannz site says what your speed is as a percentage of the UFB.
My speed of 7.58Mbit/s, for example, is listed as just 5.44% of what what I could achieve if Crown fibre was laid to my door.
Well, maybe.
A fibre connection can hit up to 100Mbit/s in both directions, or roughly 10 times the speed of most copper connections today.
But the first of the Big Five ISPs to announce UFB retail pricing, Orcon, has a speed limit of 30Mbit/s on its cheapest fibre plans (that is, those pitched to compete with the price of a copper/DSL connection by hitting a $75 price point. You can pay more for full speed, and more data).
And Tauranga ISP is offering just 10Mbit/s on its lowest price fibre deal – I can achieve 90% of that speed on my copper connection today (a stat that wouldn't look so flash on Reannz site).
There are a number of factors at play here. For one, fibre won't suffer the contention issues that bog down copper, meaning a DSL connection achieves only a fraction of its top potential speed at peak times when everyone piles onto the net, such as after school and immediately after prime time.
And it is not possible for Reannz Broadband Test site to factor in the messy commercial and economic reality that will see most signing up to speed-limited fibre plans. Still, that is what is going to happen.
July 10: Reannz launched the New Zealand Broadband Test website (nzbt.org.nz) at 4pm today.
This site allows people to test the speed and performance of their home or work internet connection and compare it to fellow New Zealanders and ISPs (and eventually with other similar sites around the world).
The NBR ONLINE staff already knew their TelstraClear connection at NBR Towers was not about to win any landspeed awards.
The Broadband Test website confirmed suspisions, and everyday experience.
The result was expressed both in terms of a percentage of speed that could be achieved under the UFB Crown-subsidised fibre rollout – comparison that some may see as having political overtones, or at least an agenda to stoke demand for fibre – and in the more techy terms of megabits per second up and down:
Click to enlarge.
But while the site worked well on NBR's laptop, several readers were quick to complain its Java element would not run on iPhone or Android devices (see comments below).
Reannz (Research & Education Advanced Network NZ), which adminsters the largely unsung $100 million state-owned Karen broadband network, which links universities and Crown Research Institutes.
RAW DATA: Reannz press release
Back to home
Back to top
Print
Latest
NBR SERVICES
SUBSCRIPTIONS
Sign up for NBR subscriber content or email alerts
Comments and questions21
So just a question.
Why do we have the government funding network deployments (UFB/RBI), owning network companies (REANNZ, KAREN, Koridia), competing in the Retail ISP market (Orcon).... all with our tax payer money...
Doesn't work, crashed (Mac, Chrome)
didn't work on my iPad or Samsung S3 either, no Java.
doesn't work on my windows PC running firefox.
fail
Doesn't work on Windows PC - Using Firefox & IE (Latest versions) with Java enabled = FAIL!
Sort it out - make sure it WORKS! before you put this online!
it asked too many questions, i found it very annoying.
will not use it again
Its massively inconsistant with its results as well, speedtest is more accurate than this thing. I am on a 100Mbit connection and I get results of 95/92Mbit consistantly on the speedtest sites around NZ, and randomly between 16/44Mbit and 44/86Mbit on this. I guess thats typical of something funded by our taxes, lots of money on consultants and a mediocre product at the end of it. Its like when consumer magazine launched their one and it was on a virtual server with a 10Mbit Internet connection, then they tried to judge connection speed of ISPs based off it. If you can't do it right, don't bother.....
Works fine for me on Win 7 & Chrome.
It reported around 6.2 MB/s which is close but no banana.
I get better from overseas download sites and thats using the
sea cable.
Either this is a con job, which knowing our current government it
probably is, or the test server needs a serious overhaul.
Wildly inaccurate & buggy. Appears to be in breach of the Comcom guidelines about broadband speed representations...
"Advertised download or upload speeds should be realistic and consistently achievable by consumers. They should not reflect theoretically achievable or maximum speeds as it is unlikely that customers will be able to get these speeds regularly"
so, should we ask the com com to prosecute?
Absolutely. ISPs can't sell fibre by claiming the download speed profile (i.e. 30 or 100Mbps) is the speed of the connection, so it's a bit on the nose for the govt to be comparing existing download speeds to a theoretical speed...
It is great you have set up this service. However a word of caution I was on a Fibre Optic connection with XFNET here in Northland and due to the offered speeds being capped at 10MBPS UP and Down I found it consistently slower than my Naked DSL connection. I think it might be perceived as a little misleading if you are offering comparisons of what in an ideal world they might be (Fibre Connections that is), when in reality sure the content locally will be faster but unless there is a matched pipeline out of the country nothing much will really change. Just a thought.
One of the most poorly kept secrets about the nation UFB roll out is it's big brother aspect. If you are on UFB you will have a static ip (fixed ip) and this is not up for negotiation (even with Ipv6). The reason is that these companies have signed an agreement with the government at the behest of the US that all your information is tied to your account and IP address for 7 years. So much for a national id number, they will have all your habits and message board postings tied to your ip address. Really scary stuff, the Copyright file sharing amendment bill was stage 1, the UFB with static ip stage 2 and stage 3 is anyone's guess. This is big brother stuff and being rolled out world wide.
Sigh, Orcon residential UFB plans right now are giving out dynamic IPs on a 1 week dhcp lease. Get your facts right first.
Java? Are you kidding me?
Didn't work on my work computer, not sure if it was a Java issue or blocked by network rules.
On Speednet got 94Mbps DL and 77Mbps UL, ping 5ms - BOOM!!
Great tool to see the real world performance of the Internet with all the ways we use it included in one space, warts and all.
But, mostly going to be a record of wifi test speeds as most computers are likely to be used testing while connected on wifi. Many may also be used on mobile so like Akamai, results cannot be relied on for comparisons.
Anonymous #13, can you provide a reference to support that?
This is a research network with poor connection to the rest of NZ, The crowd that do this for them "SNAP" are clearly having issues as outlined by others above
Anonymous #18 The server is hosted at VUW and they are connected to WIX, so more than likely have at least 1Gbps to most service providers. The "Reasearch" network does indeed have a poor connection to the rest of NZ internet, infact it is not connected to the NZ internet at all.
Not working for me in Internet Explorer 4 on Windows 95 either.