Evo 4G Wifi speed test

So we've already taken a look at the strength of the Wifi signal on the Evo 4G. And that's important. You have to have a good connection to get anything done. But signal strength isn't everything. And so we re-ran our tests, this time measuring data speed. How did the Evo 4G compare? If the picture above is any indication, our results might surprise you.

Let's discuss, after the break.

Evo 4G Wifi speed test - Phil's results

Same setup as last time. The particulars:

My router's in one corner of my house. I started in the corner of the house farthest away, working my way back toward the base, testing speed at the same five locations (and using the Atlanta server) as I tested the signal. I held the each phone in the same manner, pointed in the same direction, and ran the FCC test. Rooms are listed from farthest from the router to closest. Fasted speed is noted in bold.

  Evo 4G Nexus One Droid
Laundry room -- 2.27/1.88/57 4.81/1.37/56
Dining room 4.63/1.86/64 4.72/1.85/59 6.85/1.45/52
Living room 7.24/1.83/59 7.45/1.92/53 7.58/1.72/55
Bedroom 8.08/1.86/64 7.77/1.91/56 7.26/1.64/54
Office (router location) 8.39/1.89/58 8.29/1.90/55 7.79/1.80/57

Results are downstream/upstream/latency. Speeds are Mbps, latency is in milliseconds.

So here's what I found: I ran each test five times in each location, for a total of 75 results. Took the average of each location, and that's how I came to the numbers above. The Evo did the best in the two locations closest to the router, and the Droid did the best farthest from it. The Evo refused to register data in the laundry room -- farthest from the router -- which is why it has no results.

Evo 4G Wifi test - Jerry's results

Same deal as above. Five tests in each location (on the Washington, D.C., server), with the averages below. The particulars:

  • Router: Netgear WNR3500L 802.11 b/g/n at 2.4GHz
  • Phones: Evo 4G (with 1.39.00.05.31 radio), Nexus One, Sprint Hero, Google Ion (dev phone)
  • Test app: FCC Mobile Broadband Test
  Evo 4G Nexus One Sprint Hero Google Ion
Back porch -- 0.58/0.85-117 -- --
Living room 5.10/2.1/39 5.5/2.7/47 4.30/0.9/116 3.90/1.10/88
Kitchen 3.30/1.1/88 3.77/1.3/28 1.60/0.97/61 1.40/1.00/58
Office (router location) 11.2/3.1/22 9.7/3.0/41 5.73/1.37/88 4.90/0.90/104

So what did we learn?

Whoa, there! Jerry and I were both a little surprised by our findings. The Evo 4G had the fastest average closest to the router for both of us. And when it doesn't have a signal -- which pretty much was the case in my laundry room and on Jerry's back porch -- it doesn't get good speed. Natch.

But in between, the Evo 4G was right up there with the other phones. It beat out the Sprint Hero and Google Ion in Jerry's test, and the difference was mostly negligible in my tests. Repeat: The Evo 4G was not not noticeably slower than the other phones. (Except for when the Motorola Droid kicked everybody's butt, beating the Evo and the Nexus One in a few tests, as well as signal strength.)

So what do we take from this? Slow Wifi speeds may be in our heads, at least a little bit. While our tests were far from scientific, we both saw the same trends and had the same "Ah ha!" moment -- the Evo 4G did much better than we thought it would and is mostly on par with the other phones we tested.

Will you get different results? Maybe. There are a lot of variables to take into effect. (And we still need to test against a Droid Incredible.) Will we see signal strength and/or speeds improve with a software update? Maybe. Maybe not. But you won't hear any more complaints out of Jerry or myself.

 

Reader comments

Evo 4G Wifi speed test -- you might be surprised by the results

37 Comments

Just tested my Incredible with the same app, just in the room where my router is. Linksys WRT54GS
My results are:
11.8 Mbps download
2.23 Mbps upload
40 ms latency

and second test after I killed all running apps

12 Mbps download
2.4 Mbps upload
43 ms latency

How do we do that?

Because I really am getting... Well, angry a bit... With the lack of Incredible coverage.

If the EVO is #1, the Incredible is a very strong #2.

Just sayin'. =)

Well I am pretty sure the fine folks at AC will graciously take your Incredible off of your hands so they can include it in their tests. How about you send it to them?

I just got 10.26 mbps...0.53 mbps upload...65ms latency out of chicago, il server im in cincinnati area...actually hamilton ohio

Signal Seems poor in my room... little to no bars but still connects... yet i get quicker speeds then my friend on his Hero or MyTouch...

I think the Icon is screwy personally

Yeah WiFi final strength isn't that great but the speeds have been top notch. It hasn't slowed down on me in my house.

The idea of whether this is potentially a driver issue or not reminds me a lot of the wifi experience on my i5 17" Mackbook Pro. For example, when i'm booted into OSX, i get outstanding wifi reception and signal (as reported by the os). However, when i'm booted into Windows 7 via boot camp, its a totally different story. The signal is much weaker from the same location and the speeds seem to reflect that. I'm by no means an expert, but that would lead me to believe the driver does in fact have a large impact on things like reception and speed...considering the hardware is exactly the same in both environments.

I think this test would be MUCH more interesting if you did a test from within your own network rather than the internet-based test you did.

802.11g goes up to 54 Mbit/s, so why not run a web or FTP server on your LAN and measure that way? You're interested in WiFi performance here, so eliminate the variables that don't matter (your ISP, the speed test site over the internet, etc). Hopefully you would see much higher speeds this way and you would start to see the limitations of the phones more.

The thing I would take away from this article is that weak WiFi isn't too important when you're ultimately bottlenecked by your internet speed anyway (e.g., if your weak 802.11g is only giving you 11 mbit/s, who cares? You probably don't have >11 mbit/s internet access anyway).

Yep. But it's hard to have a standardized test that anyone can perform that way. I did happen to test this situation when I was investigating 802.11n on the nexus one:
Slackware 10 server hardwired to wireless router
Running standard LAMP server
test preformed using Andftp client from the market, sitting about 2 feet from the router.

N1 and Hero were tested, both were able to hit the 54 mb/s (802.11g) mark as advertised, but sustained xfer rates were about 6 for the Hero and about 12 for the Nexus 1. Not bad considering that the wifi radio is tweaked to draw very little power.

Nexus 1 sustained xfer rate did not increase using 802.11n mode either BTW. Chipset/RAM/processor are the real bottlenecks IMO.

The test Phil devised show a real-world example and clearly tell us that the Evo performs as expected, even though the signal strength reported doesn't seem kosher.

I agree wit Nate.
Unless you can control ALL traffic on your internet connection for the duration of the test this this is essentially useless.

Mail is always flowing into my servers, idling desktops and phones are are checking mail, or social media.

Testing 6 times from opposite corners of my house and 4 feet from the router gave counter intuitive results. I've seen upload speeds exceed download speeds.

This app dramatically differs with Speedtest.net's app run mere seconds apart.

All of these are flaky at best.

I agree with the first 2 comments. How about using the Incredible for a benchmark since it is, you know, Incredible. It does still exist does it not?

Are you sure? iPod Touch gets much faster speed in my apartment than the EVO. As an aside, should I think of ridding myself of the iPod Touch?

I get far better signal with my Incredible than my buddy does with his new Evo , page after page loads much faster in my home than his evo , turning off wifi my Incredible also smokes his evo with 3g as well , now of course 4g will be even better but for real world use everyday on wifi or 3g the incredible smokes the Evo , at least where I live . I didnt think I would care for the Evo front facing camera but I must say it was cool when he called and had a video chat with his wife , this I cant do but I will when Motorola comes out with there Evo killing devices , I must say as well Verizons signal and Sprints signal in my area is equal 4-5 bars , strong signal on both

Well my EVO's 3G data was about twice as fast as my Pre's 3G data in the same location and a week earlier my Pre's data was about 1 and a half times faster then my friend's Incredible so that would mean my EVO's 3G data is two and a half times faster then the Incredible's 3G data. Hmmmm... I'll be keeping my EVO over the Incredible. :)

The issue isn't the EVO's wifi speed, it's connectivity. The speed is fine if I have a signal, but if I'm farther than about 20' from the router, it's no go.
HTC apparently has a bad batch of radios or something, because the complaints about poor wifi are rampant.
This had better be fixed in a few days or it's bye-bye Sprint/HTC/EVO, and hello Apple..

Someone certainly hasn't been keeping himself updated lately. Verizon among other carriers have said on numerous occasions that capped data is a GREAT idea. Now that AT&T has taken the first steps, I'd expect the remaining to be on that wagon before Christmas, possibly before end of summer.

Verizon has always been capped. Their "unlimited" plan is capped at 5Gb but I don't think they enforce it much. AT&T's new capped plan is the same price but at 2Gb I believe.

Please stop perpetuating false information, the only VZW data plans that are capped are the for data cards OR if you add tethering to your smartphone. Otherwise it is still truly unlimited (until they decide to copy AT&T that is).

I think what people are failing to remember is the speeds of the phones. More memory + faster processors if everything else is equal will mean more speed. Once processing and storing are solved then transmission is the issue.

Not quite sure where any of this matters though considering your phone isn't fast enough to render a page at much faster than probably 300-500kbps. So until the mobile rendering engine and processor in a phone gets fast enough you won't see it in anything other than synthetic benchmarks.

I think it is a driver issue. My brothers Droid beat the Evo in some WiFi test the the EVO stomped the Droid in others, at the same distance from the router. I think there will be an update to address it or 2.2 will.

I'm really disappointed with the lack of love the incredible is getting. I bought the incredible because ac told me that it was the best Android phone to get on Verizon and on the market at the time. but now it seems like I was told wrong. it seems like they really prefer the Droid and nexus over the incredible and the evo seems to be on every headline. incredible hasnt been out long but it seems ac has already forgot about it.

Maybe Phil should just put a disclaimer at the top of every article saying: "This article in no way involves the Incredible because I don't physically HAVE one to test."

Anyone know if the Droid having a metal body would make a difference? Does it use the body at all to pick up signals?

Phil, how did you get/test the data transfer speeds?..[EDIT: I just noticed you did use a broadband internet speed test!!!]......are they internet speeds, like using "speedtest", or file transfer speeds....and, too bad you didn't collect/report the connection speeds, in Mbps.

My complaints, and others, have not been about "internet" speeds, but connection and file transfer speeds. For example, if you want to transfer big video files to your EVO to your computer. Theoretically, all of your tests would only require a connection speed of 12Mbps.

But, in general your tests support our concerns w/ the signal being week over all, and dropping off w/ distance much sooner than most other gear.

***MY MAIN POINT** Had your transfer speed testing been done w/ "files" you might have found out that, for example, the Evo connecting @ 11Mbps would not transfer "files" as fast as a Droid, in the same location, that might be connected @ 54Mbps.

Kudos for doing more test, but I hope you see this additional point. This has been a source of confusion re: comments on the internet since day one.