Does anyone have AIS Fiber?

[ethaniel]

Hello everyone, I'm doing an ultimate test of all major fiber optic providers in the area. I can reliably monitor TOT, CAT, 3BB 24/7, but I don't have access to AIS.

If you are connected to AIS Fiber and have a few minutes to spare, it would be much of a help.

 

I will provide the results (with graphs) in about a week.

[balo]

In Pattaya yes. 

[sammieuk1]

AIS 100 mbps down 20/30 mbps up on ethenet 60/70 mbps on wifi  perfect for the last two years constant speed 990 baht pm. 

[ESCAPIS]

My experience with AIS:

 

I have a 50/10 connection with AIS for a year now. I pay ฿950.16 monthly.

 

It came with a Playbox that I could hardly understand or view as it is all in Thai,  and disconnected it altogether from my TV.

 

AIS will supply you with a very outmoded Fibre modem/router, to which they attach another router that operates on both 2.4 and 5 hertz denomination. You will effectively end up with 3 networks. Surely there must be a modem/router capable of such feat, but AIS knows best I suppose.

 

They will keep bombarding you with ads to upgrade to the platinum package for a modest sum for the first few months and charge the full sum subsequently, but I never fell for that bait, as my watching is modest and confined to BBC, CNN and the likes, which I get through a browser.

 

The connection itself is wonderful according to Ookla. But when you choose more unbiased sources the real connection is more like 30/7 on http protocols, other protocols fair way better.

 

There are certain times of the day where the connection is completely absent or it is so slow that the beach ball spins for a considerable time. I suppose it is the area one lives in and the prevailing infra structure that would make it tolerable or otherwise.

 

If you are after support, the English speakers almost non existent. I could imagine them looking at the English line and waiting for each other to pick the call, just like they do when you walk in any Thai commercial establishment where they suddenly get absorbed in whatever they are not doing, while looking sideways praying that someone else would have the guts to greet you...

 

In the event of getting really lucky talking to a CSR, and after ascertaining you are not a drunk farang by suggesting many tests you could or would perform on your own, a promise of a technician who will contact you within 24H or... months...

 

Recently, I was informed by a Thai friend that I can dislodge the Playbox from my subscription and save about 400 baht per month, which I did immediately.

 

 

[StevieDee]

30Mbps download, 15 Mbps upload, had it for just about a year in BKK. Never missed a beat, never been anything less than the 30 & 15 paid for. Very happy with their service.

[HeavyHands]

Had 50/20 in CM for few weeks. No prob so far always around the 45/50mbs promised. @ 590b/month.

 

[Wake Up]

I have AIS Fibre 50/20 at a condo near On Nut BTS.  I love it and it works great and wi fi fast and also love the tv channels.  I have HBO and I pay 799 baht plus tax which equals 924 baht a month. After the first three months price increases by 100 baht a month. Paid 650 baht to have the playbox installed and it took the young man about 20 minutes and he showed up on time. Play box has an English language option and is easy to use and your installer will help you understand the playbox and set the language you choose. HBO go is my favorite amenity and worth every penny. 

Edited November 12, 2017 by Wake Up

[Chaiwallah]

I have AIS Fiber at my condominium on Rama 3 Rd. I've been using it for about two years I think after changing from True. I am very satisfied with it, better service, speed and price than in the US. I pay 630 baht a month for a 30/10 internet access. On the few occasions I've had a problem (no internet), the company has been very responsive and remedied the situation quickly. I highly recommend them.

[ethaniel]

 

Results for Fiber connections so far (in megabits - NOT MEGABYTES - per second)

1st place: 3BB (2 connections on top). Stable at 2-3 Mb/sec during the day and night.

2nd place: CAT and AIS (but their speeds are unreliable and constantly changing during the day). Jumping from 0.7 Mb/sec to 1.5-2 Mb/sec.

3rd place: TOT (3 connections on the bottom). Absolute garbage at 0.5-0.7 Mb/sec.

 

I am doing a long-range international speedtest every 30-60 seconds on each connection.

The remote server is 250-300 milliseconds away (in Moscow) which gives a pretty accurate idea of what the internet connection is actually like.

 

I need someone with a 100 Megabit AIS to join the test to confirm the results.

 

Edited November 12, 2017 by ethaniel

[ethaniel]

If owners of TRUE Fiber want to join, I can add them to the graph. However, I don't think that we have TRUE Fiber on the island yet.

[muratremix]

On 11/12/2017 at 6:49 PM, ethaniel said:

If owners of TRUE Fiber want to join, I can add them to the graph. However, I don't think that we have TRUE Fiber on the island yet.

Koh samui has True docsis (which is carried via fiber) on island and they can provide 200 or 300 mbps packages.

[suzannegoh]

 

Results for Fiber connections so far (in megabits - NOT MEGABYTES - per second)

1st place: 3BB (2 connections on top). Stable at 2-3 Mb/sec during the day and night.

2nd place: CAT and AIS (but their speeds are unreliable and constantly changing during the day). Jumping from 0.7 Mb/sec to 1.5-2 Mb/sec.

3rd place: TOT (3 connections on the bottom). Absolute garbage at 0.5-0.7 Mb/sec.

 

I am doing a long-range international speedtest every 30-60 seconds on each connection.

The remote server is 250-300 milliseconds away (in Moscow) which gives a pretty accurate idea of what the internet connection is actually like.

 

I need someone with a 100 Megabit AIS to join the test to confirm the results.

 

 

Can you explain these results and your methodology a little more? What is confusing me is that the numbers seem to be about a tenth of the speeds that I get with AIS Fibre. Is it because you measuring in megabytes per second rather than megabits per second?

 

[ethaniel]

6 hours ago, suzannegoh said:

 

Can you explain these results and your methodology a little more? What is confusing me is that the numbers seem to be about a tenth of the speeds that I get with AIS Fibre. Is it because you measuring in megabytes per second rather than megabits per second?

 

It's megabits per second. 

 

What I'm measuring is a speed of a single connection to a remote international server. The server in Moscow, Russia is chosen because it's far enough to test all the international routing policies that a ISP has. It's a "worst case" scenario - the lowest speed that you can possibly get on a chosen ISP.

 

Each test takes 10 seconds which is enough for a download to "speed up". Normally the speeds are super-high for the first 1-2 seconds, then they drop, and then they gradually increase (weird ISP policies). 10 seconds is enough for "averaging".

 

The reason I'm doing a single connection test is to show how fast a single image/movie/iptv stream can load over a given connection if it's downloaded from outside of Thailand.

 

It's worth noting that Facebook, Youtube, other big companies have servers in Bangkok, so their speed is much faster.

Also, if you're downloading through torrents, then multiple connections are established - so you might not notice the difference between a fast ISP and a slow one.

 

I decided to do this test because sometimes my internet connection feels very slow. The pages took 3-5 seconds to load completely (which is annoying). And some of the channels on my iptv couldn't play without hiccups. Also, if I'm watching a rare youtube video (which isn't cached in Bangkok), the HD stream wouldn't download fast enough to play without hiccups.

 

 

 

 

 

 

Edited November 30, 2017 by ethaniel

[suzannegoh]

It's megabits per second. 
 
What I'm measuring is a speed of a single connection to a remote international server. This server (in Moscow, Russia) is chosen because it's far enough to test all the international routing policies that a ISP has. It's a "worst case" scenario - the lowest speed that you can possibly get on a chosen ISP.
 
Each test takes 10 seconds which is enough for a download to "speed up". Normally the speeds are super-high for the first 1-2 seconds, then they drop, and then they gradually increase (weird ISP policies). 10 seconds is enough for "averaging".
 
The reason I'm doing a single connection test is to show how fast a single image/movie/iptv stream can load over a given connection if it's downloaded from outside of Thailand.
 
It's worth noting that Facebook, Youtube, other big companies have servers in Bangkok, so their speed is much faster.
Also, if you're downloading through torrents, then multiple connections are established - so you might not notice the difference.
 
 
 
 
 
 

But is a worst case scenario like that predictive of a users' experience? I think not - there might be some small village in Somalia to which I'd only be able to get a 300 baud connection but that doesn't affect me. I watch HD video streamed from the US and EU all the time, and if your data was typical then I would not be able to.

[ethaniel]

1 hour ago, suzannegoh said:

 

But is a worst case scenario like that predictive of a users' experience? I think not - there might be some small village in Somalia to which I'd only be able to get a 300 baud connection but that doesn't affect me. I watch HD video streamed from the US and EU all the time, and if your data was typical then I would not be able to.

 

I would love to add a speedtest to Los Angeles if someone could run iperf3 on some server there.

Can you show where you are streaming HD video from?

[suzannegoh]

19 hours ago, ethaniel said:

I would love to add a speedtest to Los Angeles if someone could run iperf3 on some server there.

Can you show where you are streaming HD video from?

There’s another thread ongoing thread on ThaiVisa where people have been providing benchmarking results and discussing the theory & practice behind it all.  One guy there is reporting download speeds in excess of 100 Mb/sec  from San Francisco to Bangkok using AIS 200/50 Fibre.  That thread is here:
https://www.thaivisa.com/forum/topic/949508-3bb-vs-ais-fiber-vs-true-online-fttx/

In some cases the HD Video that I’m streaming might actually be coming from servers in Asia.  I don’t actually know how to trace the origin of the feed.  The ones that I’ve streamed from most often are MLBTV (baseball), NBA League Pass (Basketball), and Player Klub (general TV watching).  The first two stream at 720p/60fps and need about 5 Mbps to support that resolution; Players Klub generally needs less bandwidth because they are using pretty heavy compression. 

 

[ethaniel]

7 hours ago, suzannegoh said:

One guy there is reporting download speeds in excess of 100 Mb/sec  from San Francisco to Bangkok using AIS 200/50 Fibre.  That thread is here:
https://www.thaivisa.com/forum/topic/949508-3bb-vs-ais-fiber-vs-true-online-fttx/

Speedtests are normally performed in several parallel connections (5+). They are also allowed to run longer. I've noticed that local ISPs tend to speed up connections that take longer than 10 seconds to run. This is why people are reporting high results on different speedtests (although regular internet browsing may feel sluggish).

 

In my tests I'm drilling down do the basics - how fast a single connection can be within a 10 second time limit. 

This allows me to show the differences between different providers in terms of how fast a page load experience would "feel".

 

 

Edited December 1, 2017 by ethaniel

[suzannegoh]

Speedtests are normally performed in several parallel connections (5+). They are also allowed to run longer. I've noticed that local ISPs tend to speed up connections that take longer than 10 seconds to run. This is why people are reporting high results on different speedtests (although regular internet browsing may feel sluggish).
 
In my tests I'm drilling down do the basics - how fast a single connection can be within a 10 second time limit. 
This allows me to show the differences between different providers in terms of how fast a page load experience would "feel".
 
 

That's discussed extensively in that other thread too. Some speed testers such as speedtest.net are multi-threaded while others let you select between multi-threaded or single-threaded. Some of the testers also measure how much the speed varies over time. So while I'm not sure exactly what you're trying to do, I'd be a little surprised if there was a need to invent a new test.

[ethaniel]

1 hour ago, suzannegoh said:

That's discussed extensively in that other thread too. Some speed testers such as speedtest.net are multi-threaded while others let you select between multi-threaded or single-threaded. Some of the testers also measure how much the speed varies over time. So while I'm not sure exactly what you're trying to do, I'd be a little surprised if there was a need to invent a new test.

One of the reasons I'm also doing my test is to show how speed changes during time of the day. My tests run non-stop. You would not see that fluctuation on a regular speedtest.

 

So, for example, if you did a regular speedtest on CAT, you might be lucky and hit the time when they give you maximum speed (10 megabits per thread). But most of the time the speed is slow (around 700 kilobits per thread).

Edited December 2, 2017 by ethaniel