Download Accelerators: Do They Actually Work???

I recently found a download accelerator that I liked and it seemed to work. I did wonder though – do they actually work?

I will test the normal Windows download program against Download Accelerator Plus Premium 8, all at the latest version possible. The tests do not take into account latency or traffic loads; however, all downloads will be from the same source AND will have no gap in between each method. The internet speed used is 768 Kb/s down, 512 Kb/s up. All files tested will, for the Accelerator’s sake, have RESUME support.

There will be 3 tests for each program:

  • Test A: A small file ~500KB (Large Document / PDF)
  • Test B: A medium file ~50MB (iPod updater)
  • Test C: A large file ~750MB (ubuntu CD ISO)

ALL tests will be timed in 2 ways:

TEST I: Full download process (clicking, target, etc)
TEST II: Actual download process (just the actual download)


Test A

TEST I

TEST II

Windows Standard Method

9s 46ms

5s 34ms

Download Accelerator Plus

12s 56ms

6s 03ms

The winner for the actual downloading time is, by a close 29ms, Windows. The winner for full download process is also Windows.

Winner: Windows Standard Method


Test B

TEST I

TEST II

Windows Standard Method

34m 21s

34m 12s

Download Accelerator Plus

38m 12s

37m 42s

The winner was Windows by a long way – 3m 52s, in one case. The winner is also Windows for actual downloading time.

Winner: Windows Standard Method


Test C: ~750MB (CD ISO) Chosen file: ubuntu 5.10 Standard CD ISO
Test C

TEST I

TEST II

Windows Standard Method

3h 23m 12s

3h 23m 3s

Download Accelerator Plus

3h 24m 45s

3h 24m 12s

So, the winner here was also Windows, but only by a few minutes. As usual, but closer, Windows was also quicker for the actual download.

Winner: Windows Standard

So, in conclusion, we have learned, Download Accelerator Plus and probably, IMHO, all download accelerators not only don’t work but actually SLOW DOWN the downloading process. However, it is important to note that the download accelerator companies do say the most improvemant is seen in dial-up, but broadband does still speed up!

The next time I need to download a file, I know which way I’ll do it, especially with DAP costing $44.90!!!

Contrary opinions/test results welcome! (pls send these to Joe)

Readers respond on page 2…

Tayo Moore

Many thanks to all who responded, pointing out that download accelerators shine when multiple paths are available. There is a wealth of free software which can markedly improve your download times – take the time to read selected reader’s comments and give some of this software a try.


Dan

The test was flawed and didn’t take the biggest advantage of Download Accellerator (or others, like Free Download Manager) into account; that feature is multi-source downloading. I have 8 MB cable and with Free Download Manager I can add multiple sources and download concurrently from many sites, thus increasing my available incomming bandwidth quite a bit.

Jicin

The benefit of download programs lies in finding mirrors and downloading from more than one source and thus maxing out the line’s speed. I get 6600 KB downstream, so single servers that max out my line are rather rare; by connecting to multiple serrves I often get my speed maximised. Also, download managers/accelerators were more common earlier (for example GetRight) when Windows wouldn’t support resuming properly (or not at all) and were rarely used for speeding up downloads.

I’ve got FlashGet installed for now but actually I don’t even bother to use it for files less than 100 MB. For bigger files, I sometimes get much (sometimes 5x) more download speed if mirrors are available. Also, I can query files so that a whole bunch will be downloaded one by one without me having to start every single one when the previous has finished. That’s what download accelerators are for, not downloading one file from one server.

Sebastian

There are situations where download accelerators really do help:
If the download is somehow limited by latency or geographic distance to server, instead of lack of bandwidth,
you can benefit a lot. Example:

I download a file from a server in the U.S. (I’m from Sweden) and the download caps out at
80kb/s. I know for a fact that both my connection (24 MB/s) and the server’s connection is faster,
but the geographic distance and sheer number of routers is making the download slow.

Download accelerators will help here because they will start several threads, each one downloading
at 80 KB/s until either mine or the servers connection get capped.
Yesterday I downloaded a file from the U.S. at 400 KB/s using my download accelerator versus
90 KB/s using the built in download manager in Firefox. Another plus in using a download accelerator is all the extra features such as, resuming, queing, and more.

And what did I pay for my download accelerator? Nothing. I’m using the Firefox browser with a free extension
called DownThemAll.

x586

I don’t use them, but I have tested a few and it’s really quite simple
how they work and what they will and won’t speed up.

They compress the file/s server side and this reduces the size, seemingly
making the download faster. They do not make it faster, they make what
your downloading smaller thus it takes less time.

Images are re compressed at a reduced quality – any sort of text can be
drastically compressed. What all this means quite simply is that web
pages load faster, especially on dial up connections. What it does not do
is help when downloading anything besides text and pictures, and since
it reduces quality of the pictures, people may not want to enable that
feature anyway.

At best a web page will load about 5x faster on dial up,
but usually you’re only going to see 1.5 – 2 times faster, and that’s only for
dial up. On 128 KB/s or higher connection, you won’t see any measurable
improvement unless for an exceptionally large page and you’re using the
image compression.

In general they are not very useful, even for people on dial-up.

Ray

In my experience, my Windows (or rather, Internet Explorer downloader) is pure crap. It doesn’t download files over 11 MB completely. I started using the free version of Star Downloader and it has worked well.

True – not all downloads will be faster, but some do; I can manage 4 connections with 200-300 KB/s download speeds from well funded sites/servers, although some servers just won’t let you automatically download from 4 connections and it does take longer to connect and actually start downloading than IE downloader. But for me it’s a matter of having a complete download vs not. And you can always turn off Star Downloader, so it doesn’t get in the way when downloading smaller files.

Peter

How about testing Google’s new Web Accelerator as well since it is free?

Thomas

There are many conditions where an accelerator can make a big difference, especially if the connections between your PC and the server are not all that perfect, this is often the case for people living outside of the USA, like myself.

Due to routing conditions and who knows what else affects long distance TCP/IP connections, downloading a file by IE’s utility can often only reach a maximum speed that is a fraction of the actual bandwidth your ISP supplies. If a download accelerator uses multiple threads, each one may be slow, but they add up and give you a much speedier download overall. Not always, but often.

For example, this morning (this was before reading the article, so I did not make any detailed measurements), I had to download the same file on two PCs on the same network. The file was a Logitech driver from their US server, weighing in at about 25 MB. On one PC, a fresh install, I used the IE downloader and downloaded the file at speeds between 15 KB/s to 30 KB/s. Only my own PC I used FlashGet to download the file – it was split in 5 parts and the overall download speed was between 70-75 KB/s. This is on a 1 Mbps downstream ADSL line, so there’s plenty of extra bandwidth to go around!

But if, like in the conditions of the article, the IE download can max out the connection, then there is an advantage of a single thread download. That is, less overhead compared to the download accelerator’s individual threads – each have their own overhead which add up and thus allow less bandwidth to the actual data being downloaded.

So, in conclusion, download accelerators do actually work in certain conditions, and you do not need to use one every single time you download a file, but they are still an invaluable tool to the savvy Internet user.

Mic Terpstra

I use Flashget – it lets me download a file using 10 streams simultaneously, which is still not enough to totally saturate my own broadband, but it lets me download huge files fast enough to be happy.

Bit Torrent uses the same concept – it opens a large number of streams to various sources to download different pieces of the same file/s. Unfortunately, the Bit Torrent networks are limited by the UPLOAD speed of its users (which is typically less than the download speed), and is also tuned to reward uploaders with more download speed.

Bit Torrent really has the potential to totally saturate your download bandwidth, delivering files at the fastest possible speed, but only if you have a large number of incoming streams (and since you can only upload to just so many other Bit Torrent clients at once, you probably won’t ever get a huge amount of fast incoming streams.) Morpheus was also excellent at saturating my bandwidth.

Bottom line – My testing shows I can download a 100 meg file in a half an hour or more, or in one minute or less, depending on how many streams I can open up to the server that hosts that file. A good download accelerator is what makes that difference.

Note: Here’s another site to check out: Internet Speed Booster Software

Email Joe

Be the first to comment

Leave a Reply