HSDPA versus iBurst

In the mobile broadband arena whats more important to you - price or speed?

  • Price

    Votes: 37 88.1%

  • Speed

    Votes: 5 11.9%
  • Total voters
    42
Are those ping rates really true of Vodacom HSDPA. 139 ms Local ping rate.!
Wow, I thought that ibust or wi-fi was the better choice for better ping rates.
Does this mean I can play BF2 with a HSDPA connection? And if so I probably wont be able to eat every month, still a bit expensive though.
 
@RPM,

Is it possible to get the test procedure?... as I would like to do the same test in different geographical areas, different towers etc etc.
 
Hi RichardP

I don’t think it is wise to make the full testing procedure public as it will become easy to set priorities by the operators, but it is basically as follows:

1. Various local web based speed tests
2. Various international speed tests
3. Two different software based speed tests
4. Large downloads (P2P, plain downloads, download manager)
5. Basic Ping testing
6. Software based ping testing (netping, actual games etc.)
7. Web based VOIP testing
8. Actual VOIP testing

Regards,

RPM
 
rpm said:
Hi RichardP

I don’t think it is wise to make the full testing procedure public as it will become easy to set priorities by the operators, but it is basically as follows:

1. Various local web based speed tests
2. Various international speed tests
3. Two different software based speed tests
4. Large downloads (P2P, plain downloads, download manager)
5. Basic Ping testing
6. Software based ping testing (netping, actual games etc.)
7. Web based VOIP testing
8. Actual VOIP testing

Regards,

RPM
Click to expand...

Good Point :-P I am just curios as iBurst and HSDPA have been known to be crap in some geographical areas, and good in others... Being a mobile technology, I think the mobile technologies need to be tested at random locations to cancell out the 'bad tower' syndrome, also being tested during peak mobile hours (3pm on Friday, GPRS is almost dead in Sandton) and low times. Maybe run a latency test for 48Hrs on the technologies. .. Just Ideas

Best Regards
Richard
 
Last edited:
bandwidth cost

so HSDPA is more faster which is great, but the price!!!

Virginmobiles GPRS/EDGE rate is 50c/meg (out of bundle), I wonder what there 3g price will be when they finally get 3G and if they start to offer bundles aswell.

Vodacoms rate is R2.00 (out of bundle) for gprs/edge/3g/hsdpa, how does that compute, Vodacom is 4 times more expensive, was vodacom sucking the price out of their collective anal retentive asses

e.
 
Last edited:
HSDPA is really impressive when it comes to speed, at times we have been getting 1.2 Megabits in Stellenbosch through to our VPN link. For anyone wanting to do a quick connect and work its a winner, but for anyone wanting to spend hours downloading...EISH!!!
Hopefully as the subscriber base increases we will see the 2.00 per meg come down to something more reasonable.
I tested skype from Paarl to the UK a while ago and that too was fairly good, there was almost no lag in the speech.
 
This article just proves the problems that vodacom is having with HSDPA, they cannot even get it stable above the 1mb/s mark, and yet they are still advertising in every medium possible that it can get up to 1.8mb/s, I would see that as false advertising. If they said up to 1mb/s it would be ok, because it can actually hit that mark once in a while :p
 
I'm no fan of Iburst but are forced to use it atm as I've been waiting 5 months for ADSL now . I'm connected to the centurion tower and I must say , those local ping times quoted seems way to high to me , 629 ms local , come on !!! , I play games on the saix server daily and my ping times are between 70 and 150 ms , and downloads speed normally runs at 60KB/s to 90 KB/s , much higher then the 147kbps as mentioned in the article . It will be interesting to know which tower was used in this test (we all know waht a massive difference this can make).

Like I said , no fan of iburst , but this test makes iburst look like a freaking dialup solution .
 
Last edited:
How about a poll?

In the mobile broadband arena whats more important - price or speed?
 
NIGHT time testing?

Hi

I am currently using HSDPA through Vodacom in East London.

At 1am, I get great speeds (1.8Mbps) - but during the day, 236.8Kbps. I have tried over 8 times to download a new driver from HP.com, and the download fails after 5mb, every time (I only have 50gb free on my drive :P)

During the day, the service is slow, but you still pay as if you were on real broadband, which you are not.

Most people who use HSDPA are business people, and therefore use it in the day - that is perhaps when tests should be done - when contention is at it worst.

I think the article is misleading because it makes claims about HSDPA speed (which is why suckers like me went and got it) - but which area you're in makes a difference, the contention in that area makes a difference. I get different connect speeds in different parts of my house!

It is not consistent.

In fact, the only useful thing about HSDPA is that you are mobile and get an internet connection slightly faster than dialup.

ADSL is still the closest we have to REAL broadband in SA. When the mobile providers get their coverage and contention issues sorted out, HSDPA could become a serious competitor. Until then, go and rent an HSDPA card from your nearest airport (R20 a day, plus R2.5 per mb), and TRY it in your area before you commit to it.

And remember, it is UP TO 1.8Mbps, which means its at 236.8Kbps most of the time - good marketing though.

/boodah
 
Actually, coverage is more important when I want to be mobile. No coverage = no speed = no mobility.
 
I think this is very misleading. I am no Iburst fanboy however have been using Iburst in Durban pretty much since inception. I use it for skype/google/msn chat as well as ventrilo and teamspeak chat in online international games like Guildwars. Those latencies are far higher than what I experience in both online gaming and in voice chat. Ventrilo, which reports ping constantly, generally gives me about a 300-400 ping for international.

My download speed from a decent source is generally in the 100Kbs range. I don't have any difficulty filling the entire 1mbs pipe if I use multithread download programs.

I would be most interested in an article/test like the one done here, but that is done from more than one location and during more than one time of day. The nature of wireless is that there will be better and worse areas and better and worse times due to tower positioning and contention. This is like doing a poll of 10 people in Sandton city as to how much they spend on makeup and then extrapolating it over the entire country.

Unfortunately therefore, I find this article relatively worthless.
 
The problem with this sort of article is it implies that these results are what the average person can expect from these services. I have ADSL, IBurst and Hsdpa. My iburst experience is much better than the results in the article my hsdpa a bit worse.

Local latency of 600ms means either a very poor signal or a seriously contended tower.
 
As I was reading this article the thought going through my mind was "wtf, where did they get these speeds for iburst" , definately signal problems.

Those reading my reviews (which daveogg posted above), just bear in mind that those tests are under optimal conditions, I spend a lot of time tweaking signal, frame errors etc before testing.
 
noemfie said:
I'm no fan of Iburst but are forced to use it atm as I've been waiting 5 months for ADSL now . I'm connected to the centurion tower and I must say , those local ping times quoted seems way to high to me , 629 ms local , come on !!! , I play games on the saix server daily and my ping times are between 70 and 150 ms , and downloads speed normally runs at 60KB/s to 90 KB/s , much higher then the 147kbps as mentioned in the article . It will be interesting to know which tower was used in this test (we all know waht a massive difference this can make).

Like I said , no fan of iburst , but this test makes iburst look like a freaking dialup solution .
Click to expand...

Well said...I'm also no fan of iburst but i could play BF2 or wow when there was no errors on the network. my local latency is anywhere from 60ms to 200ms...avg about 90ms. and im not very close to a tower even. When I download at night I get very close to 1mbps....I can bet the test was performed on a very congested tower or there was some kind of error or problem with the signal. heres my local latencies-->

ping www.iburst.co.za -n 10
Packets: Sent = 10, Received = 10, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 53ms, Maximum = 193ms, Average = 87ms
 
AcidRaZor said:
a sad sad day when 3gb is seen as high usage..
Click to expand...

I have friends in the netherlands that use up to 10gigs a day :eek:
3gigs seems to be the SA norm when it comes to internet, simply because telkrap says so :mad:
 
This article has zero credibility. It just doesn't align with how Iburst users experience IBurst. If I had any performance even close to resembling the figures you are touting, I'd be throwing a fit, and my modem out of the window. Whatever Iburst system you used in the test is BROKEN.

My average local and international download speeds are 500-700kbps on single threads, and maxed out at >1000kbps on multi-threads. (not 187 or 147!!! Did you put the modem in a lead box???)

My average uplink speed is 200-300kpbs (not 43!!!)

My average local latecy is 50-90ms (not 679!!!!)

My average international latency is 300-450ms (not 1255!!!)

I use VOIP with Iburst regularly. Very seldom expereince any problems.

What is the point of recording "test" figures for Iburst that no Iburst user in their right mind would be prepared to live with.

Stupid, stupid, stupid. :eek:

There is no doubt that coverage for HSDPA is better, but I don't even consider Iburst as a "mobile" technology. Wireless, but not mobile. So why can't you conduct tests on a system that has been properly installed?
 
You must log in or register to reply here.
Back
Top
Sign up to the MyBroadband newsletter
X