iPhone 4 Wi-Fi performance benchmarks
After seeing Anandtech’s detailed (and impressive) review of the iPhone 4, we were surprised to see that Anandtech only managed to get Wi-Fi performance of 7.2 Mbps downstream and 7.96 Mbps upstream using the SpeedTest.net app for iPhone OS. This wasn’t a broadband performance issue because AnandTech’s testers used a broadband connection capable of 30 Mbps downstream and 8 Mbps upstream when using the desktop connection.
AnandTech theorized that this may have been a CPU bottleneck, but it could be a limitation with the Wi-Fi network they tested or it could be a limitation of the SpeedTest.net app on the iPhone. We decided to run our own Wi-Fi test on the iPhone 4 using Nick Brown’s home network and iPhone 4.
We wanted to eliminate the Broadband connection from the test loop so we set up IIS 7 web server on Nick’s desktop computer running Windows 7 Ultimate. Then we had to find a way to get the iPhone’s web browser to pull huge amounts of data from the local IIS 7 web server. We initially tried an 11 megabyte PNG image file but that test run was too short to draw accurate conclusions, so we tried to stream an Apple QuickTime file from the web server but that didn’t work. Finally we tried downloading and playing an 80 MB MP3 file hosted on the home web server and that did the trick.
Using the built-in Windows PERFMON tool on the Windows 7 IIS web server, we captured the following screen shots showing bytes/second send throughput while the iPhone 4 downloaded and played the MP3 file.
Figure 1 – iPhone 4 802.11g throughput
Figure 2 – iPhone 4 802.11n throughput
Note that we have to multiply bytes/second by 8 to get bits/second. Then we divide by 1,000,000 to get megabits/sec (Mbps).
The sustained peak Wi-Fi performance on the iPhone 4 hovered between 17 Mbps and 20 Mbps. This applies to both 802.11g and 802.11n and the peaks were roughly 10 seconds in duration. The dips were caused by the iPhone taking a break when it had enough of the MP3 buffered so that it can cut radio power and conserve battery life. We had to skip around on the MP3 to force the iPhone to keep using the Wi-Fi radio.
Conclusion
It appears that Anandtech’s Wi-Fi performance for iPhone 4 is running into some bottlenecks that are not caused by any limitations of the phone’s hardware or built-in software. That means either the iPhone speedtest app was capped in performance or the Wi-Fi network they tested on limited the Wi-Fi performance of the phone. Hopefully, Anandtech will update their Wi-Fi test using a similar technique to the one we used here.
It also appears that the iPhone 4 doesn’t get any performance benefit out of its 802.11n capability as it runs slower than 802.11g or 802.11n mode on a typical laptop or netbook. My netbook for example can get 24 Mbps in 802.11g mode. However, these speeds on any smartphone are still respectable but the next generation of competing smartphones may do better. This isn’t a big issue for most users because they will be limited by the performance of their broadband connection since most broadband users get less than 15 Mbps. Most users download from their desktop computers although it is a shame that users can’t sync with iTunes over Wi-Fi and an even bigger shame that Apple blocked a 3rd party Wi-Fi sync application.
We don’t know if the limited Wi-Fi performance of the iPhone 4 is due to the Wi-Fi chipset or if it is a limitation of the iPhone’s CPU or if it is a limitation of the software. We can conclude that the promise of 802.11n for now rings hollow on the iPhone 4 since it can’t even get full 802.11g speeds.
Update 8:51PM PST – Brian Klug of Anandtech has retested using a methodology similar to ours. They used PDFs instead of MP3s which might actually be better because iPhone’s web browser seems to be more aggressive at downloading. They’ve now managed to get their iPhone 4 to pull 24 Mbps downstream in 802.11g mode and 29.6 Mbps in 802.11n mode. That is as fast as one can expect from an 802.11g network but the iPhone 4 falls short of the 70-80 Mbps you can get on an 802.11n enabled notebook or netbook. However, the iPhone 4 seems to be much faster than the Google Nexus One in this round of testing. The Nexus One only managed to pull an average of 14.4 Mbps in 802.11n mode and peak at 20 Mbps.
[...] iPhone 4 Wi-Fi performance benchmarks [...]
I don’t know where the bottleneck lies, but n at 20Mbps should be lower power than g at 20Mbps because the duty cycle is lower. So even if the new iPhone’s wi-fi isn’t faster, it may contribute to longer battery life.
That is a great point Wes. It is a theory worthy of testing.
Volunteering Wes for that one. :D
[...] Posted by nick r brown on July 1, 2010 · Leave a Comment Print This Post jQuery(document).ready(function($) { window.setTimeout('loadGBuzz_272()',5000);window.setTimeout('loadFBShareMe_272()',5000);window.setTimeout('loadFBLike_272()',5000); }); function loadGBuzz_272(){ jQuery(document).ready(function($) { $('.dd-gbuzz-272').remove();$.getScript('http://www.google.com/buzz/api/button.js'); }); } function loadFBShareMe_272(){ jQuery(document).ready(function($) { $('.dd-fbshareme-272').remove();$('.DD_FBSHAREME_AJAX_272').attr('width','53');$('.DD_FBSHAREME_AJAX_272').attr('height','69');$('.DD_FBSHAREME_AJAX_272').attr('src','http://widgets.fbshare.me/files/fbshare.php?url=http://www.techlobbyist.net/archives/272&size=large'); }); } function loadFBLike_272(){ jQuery(document).ready(function($) { $('.dd-fblike-272').remove();$('.DD_FBLIKE_AJAX_272').attr('width','450');$('.DD_FBLIKE_AJAX_272').attr('height','25');$('.DD_FBLIKE_AJAX_272').attr('src','http://www.facebook.com/plugins/like.php?href=http://www.techlobbyist.net/archives/272&show_faces=false'); }); }Check out the tests George Ou and I ran over at Digital Society here. [...]
No problem Nick; my job is power management. Just send me an iPhone 4… :-)
Like other phones and embedded devices, the iPhone uses single-stream n so the raw data rate is limited to 65 Mbps in a 20 MHz channel, which would give real throughput of up to 40 Mbps. Laptops are probably using two or three streams to get 70-80 Mbps.
Wes, laptops that use 20 MHz can do 70-80. They can do around 150 on 40 MHz.
That’s what I said. Laptops use multiple MIMO streams in 20 MHz to get 70-80 Mbps, but a phone using a single stream cannot possibly reach such speeds and thus we shouldn’t expect it to.
@Wes,
OK, I thought you were saying multiple channels, but I obviously misread you.
Laptops I think use 2 to 3 streams using 2 or 3 antennas I believe. I think it takes at least 3 antennas to do 70-80. A smartphone might only be using 2 antennas so that would make sense.
Leave your response!
Twitter Feed
About Us
Digital Society is a digital think tank that believes culture and commerce are inseparable, that the digital economy flourishes when people are free and rights are secure, and that free markets free people.
Digital Society is an independent 501(c)3 non-profit organization, funded by donations from Jon Henke and from Arts+Labs. We advocate for a pro-culture, pro-commerce digital society through research, analysis and debate on emerging technology issues.
Reply Comments
Transparency and interactivity are trademarks of the Internet era, and we aim to foster them here at Digital Society. It is inevitable that some people will disagree with the technology policy positions we take. We want to have that constructive debate.
The Reply Comments feature gives our critics a chance to respond to our viewpoints and the Digital Society audience convenient access to competing arguments. Any time we directly challenge the views of an individual or a group on this site, the party in question may substantively respond in a guest post.
Please contact executive director Jon Henke by e-mail.
Subscribe
Daily Digest Email
Recent Posts
Recent Posts
Most Commented
Most Viewed