Video: Pre and Pixi compared side-by-side | webOS Nation
 
 

Video: Pre and Pixi compared side-by-side 41

by Robert Werlinger Sat, 21 Nov 2009 5:44 pm EST

[ youtube link ]

There are plenty of reviews for the Pre and the Pixi on the web, but few (if any) do any real comparisons on the performance differences between the two handsets. Thankfully, forums member mu7efcer has taken the time to put together a series of videos that pit the two devices against each other in a side-by-side performance showdown. The results are interesting, with the Pixi being nearly as fast (web browsing) and sometimes even faster (boot time, camera performance) than the Pre.

Here's what the videos compare:

The Pixi is pretty darn snappy, and it's interesting to see that the sluggishness some reviewers have experienced is all but absent here. With the differences in performance between the two now documented on video, I think it needs to be said: the Pixi is one heck of a phone, especially for that form factor, and doubly so when selling for as little as $25.

Category:

41 Comments

I would expect camera to be a bit faster in snapping photos, it is processing less data with 2mp vs the pre 3.2mp.

I forgot about that price drop. The phone is almost lower than a gallon of gas. Perfect for a Christmas present.

Faster boot time on the Pixi? Does that mean that it has faster flash memory? The Pre's is horribly slow (it shouldn't take an hour to write 7GB of music on modern flash memory).

So why is the Pre cpu clocked running slower than the Pixi in his test when the Pre has a better processor? Is there any way we can change the speed of the Pre's processor? Would changing the speed do anything else other than kill the battery faster?

The Pre's processor, even at 500mHz, is simply more advanced (and faster) than the Pixi's processor, clocked at 600mHz.

You can overclock the Pre CPU to 600mHz easily, though.

easily?

Look in Preware under PowerSave. There's a CPU scaling patch that dynamically scales between 250mHz and 600mHz, and a SmartReflex w/ overclocking patch that sets the CPU to a constant 600mHz.

Some people's phones might have issues with this, but most should be fine with it. I still prefer to keep mine at 500mHz though.

It takes less than 1 second to turn on these Pre and Pixi phones.

For those who do not understand these results, the boot-up time is not the time it takes to turn on your phone if you had simply turned it off. If you had simply turned off your phone, it should take less than 1 second for it to turn on. You turn your phone off by pressing on the power button once; similar to a single click on your computer

Sorry, but you have no idea of what you're talking about.

"Boot time" has been used to describe the duration that a computer goes from completely off (and consuming no power) to on and ready for user input.

When you hit the power button on the Pre, you're just putting it in some sort of screen off/sleep/standby, where it is still on, still consuming power, and by any definition, not "off."

In fact, you even use the term "boot" : "Wait for about 90 seconds while your phone boots up" but you've somehow decided that this 90 second time doesn't constitute "boot up time"

NeoteriX,

Sorry, but you are the one that is wrong! I know what I am talking about. But you my friend have read my article with very little comprehension.

If you don

FWIW, I am very interested in test suggestions. If you have some, please leave a suggestion on the thread.

Thanks.

FYI, if you have a suggestion for a test you'd like to see, please post it to the forum thread. I'm really interested in tests that can show the diffs between the 2 devices.

Thanks

It has been well documented that the Pre and most notably the iDont 3Gs are under clocked to try an preserve battery life.

Doesn't hurt heat or longevity either.

the cortex a8 is supposed to support thumb-ee which in theory supports running java bytecode at native speed. this could mean that lunasysmgr could in theory run faster.. (which also accounts for a significant part of the bootup time).

see http://electronicdesign.com/Articles/Index.cfm?AD=1&ArticleID=11555

and

http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.dui0204h/Cac...

lunasysmgr isn't java though. It's webkit.

20 cards?? My Pre just stopped at 10.

I loaded 19 cards.

Yes but as you can see, those were pretty much stock pres, with nothing loaded in them except for a few apps. plus it depends on what kind of apps you load. such as pandora and the help app. are totally different.

What?? I bought a pixi for my wife and you're telling me she has a faster phone than I do. That wasn't supposed to happen.

Uh.. I can't get past how much smaller the Pixi's screen looks.

Just saw a Pixi today in Radioshack for the first time. Awesome. I love that phone. My wife is going to get one, and I may be a little jealous.

...gripe: updated to 1.3.1 and my Pre feels very sluggish, achingly so with some apps.

I've noticed the same thing with 1.3.1 especially with facebook sync turned on.

They should make a higher end pixi.. some people like that form factor more then the pre, but want wifi :(

These tests are very interesting. It makes me wonder what's up with the Pre. I'm having a hard time comprehending why it would ever be slower.

I understand that the Pixi runs at 600 Mhz and Pre at 500 Mhz but the Pre's proc should process much more data per clock cycle. So wtf is up? Can anyone paint a clearer more technical picture for me please?

The same reason a Core i7-920 at 2.66 Ghz is much much faster than an Core 2 Quad Q6700 (or even a Q9400, or a Pentium 4) at 2.66 Ghz. The Pixi is an ARM11 architecture, and the Pre is the newer Cortex A8.

I played w/ both a Pre and a Pixi in a Sprint store yesterday and neither one was anywhere *near* as snappy as they are in that vid. That said, the Pixi was noticeably laggier. I'm not talking about browser lag either, I'm talking general UI lag. And having to swipe and/or tap multiple times because it didn't respond the first time - simply maddening! I really liked the Pixi form factor, though - it would probably be my next phone if it had WiFi and they could get the UI lag sorted.

To my mind, WebOS is still comes across as half-baked until the user experience can be equally smooth across all examples of all models of WebOS phones. And I'm talking about the phones that potential buyers pick up and try at the stores - bone stock and untweaked. The shame of it is, there is really quite a bit to like about WebOS, but the random sluggishness in the UI just kills what would otherwise be a good experience. I currently own a Centro, and it is blindingly fast in comparison (granted, not as flashy).

BTW, this is coming from someone who really *likes* Palm and has supported them for a number of years by buying multiple products. I can only imagine what someone thinks who has a neutral or even negative image of Palm going in, only to have that feeling vindicated by the horribly sluggish interface...

I'm pulling for Palm, and I really hope they get this fixed. I just hope they are doing everything they can to address this issue ASAP.

I noticed that the first time I saw a Pre in the store. It was super laggy. But there were two reasons for that. The first being that it was the original release firmware (1.1). The second was they had a demo installed that was always sitting in memory and actively waiting for a lack of input so it could start again. The Pixi I played with in the store wasn't bad so I'm not sure what you're experiencing. I'm not sure if they still have that issue with the demo tho.

As for lag spikes in general, I got them SO often with my Pre on 1.1. Like enough to get me pretty frustrated. The upgrade to 1.2 helped immensely with it, but I still get them every now and then. Resetting the phone occasionally helps when the lag spikes start comes frequently. It also helps fix the GPS signal not being caught. So yeah, webOS has a way to go, but it's slowly getting there IMO.

i used a pixi in the sprint store today as well...and i noticed the pixi can keep more cards open and still be very "snappy" when my pre sometimes gets really really slow with just 2-3 cards...

I would disregard a lot of these tests, if you notice he's tapping the Pixi with his middle finger and the Pre with his index finger. Anyone whose not blind will notice an appreciable difference in length between the two. Even with his middle finger bent in to attempt to make it a simultaneous launch, his middle finger is ahead of his index, creating these slightest faster Pixi launches.

This is not to mention the large surface impact area the middle finger supplies as opposed to the index.

Yeah, they aren't scientific. Doesn't mean we can't just watch and make observations. His two devices are damn comparable in speed it seemed, which actually kinda surprises me.

The point wasn't to measure (to the millisecond) the speed differences between the phones. The point was to see if the reviews that I'd read about the Pixi being so much slower than the Pre had any merit. My concern with all of those reviews was that they made a claim about relative performance without showing it. The showed only the experiment (the Pixi) but never showed the control (the Pre).

I wanted to see how the Pixi performed in full view of the control. I felt like I succeeded, not by determining, to the millisecond, the speed differences. Rather, I wanted to see if the Pixi really was "much slower" than the Pre. In my testing, I didn't find that.

I'd really like to know what the conditions were that caused the other reviewers to call the Pixi "much slower" than the Pre.

I'm currently in process of trying to recreate the conditions under which the engadget review showed lagginess. But I can't replicate the results. The difference may be that engadget was on 1.2.9.1 (becuase that's all that was available at the time of that review) and I'm on 1.3.1. Or it may be that he's not comparing the performance of engadget.com on the Pixi with the same thing on the Pre, and that the Pre performs exactly the same.

I found the reviews of the Pixi to be lacking. So I did some tests and made my videos of the results. If you find them lacking, I'd love suggestions on how to do better. Or you could make your own (more scientific) videos that would answer your higher precision requirements.

This is not to mention the large surface impact area the middle finger supplies as opposed to the index.

I don't find this critique to be relevant. If you watch the video's both phones respond to the application launch at (what appears to my eye) to be simultaneously. For example, the quick launch bar falls away at the same when I launch apps from there. But the applications come up at noticeably different times. I don't think the difference in app load times can be attributed to finger surface area.

lol even if he switched the position of the phones so the pre would have the same finger placement the pixi had this round the pixi wold still be slightly faster.

During my initial tests (not recorded to video) I had the relative positions of the Pre & Pixi reversed. I didn't see much difference in performance.

I calculated the differenes in the "Pre on WiFi vs. Pixi on EVDO" test for the two sites demoed in the video using the video player's time stamp:

ESPN
Pre :31 - :54 (23 sec)
Pixi :31 - 1:09 (38 sec)
diff: 15 sec (Pre is 39.4% faster)

PreCentral
Pre: 1:20 - 1:39 (19 sec)
Pixi: 1:20 - 1:46 (26 sec)
diff: 7 sec (Pre is 26.9% faster)

On average between the two site tests, the Pre is 33% faster. This is no surprise and seems significant to me even though the narrator downplayed it by saying "from this test I would not say the Pre is significantly faster" at 1 minute 54 seconds into the video.

No wonder Palm disabled WiFi on the Pixi. A key reason for purchasing the higher priced Pre would have been eliminated.

Pixi matched or beat the Pre in 6 out of 7 test posted by this youtube user. The one exception was for the EVDO vs WiFi test (although the number of cards test was not meant to test performance, there was a noticable lag on the Pixi when launching additional cards)

The only caveat is that these were stock Pres with virtually no data on them from what I could tell. My hunch is that if they had lots of real world data the Pre would be faster, maybe even significantly. Wish the youtube user could have ran the same tests on a phone with hundreds of photos, contacts, emails, messages, etc.

Thanks for the feedback. Regarding the use of "significant", I think you've shown that I probably shouldn't have said that. However, in my testing prior to recording the videos, the differences were much greater. As a result, the difference seemed quite a bit less significant than I had seen..

I'm continuing to run tests. Loading up the devices with data and then retesting is one that I will include. And if it shows any new information, I'll post another video.

mu7efer, I think you did a pretty amazing job in attempting and maintaining, for the most part, the validity of the mini-inquiry sound. I don't think anyone using their fingers to manipulate both of the devices could have done a more accurate job. It paints an accurate portrait that consistently shows the pixi slightly outperforming the pre. there is rationale behind it, but it is slightly disconcerting.

thanks for the videos.

Thanks and you're welcome.

Hey rv, it was great demo. Thanks for the experiment.

Thanks. I know how much time this stuff takes. I think I'm going to cross post on GearDiary.

I too find performance inconsistant. Sometimes downright snappy, others a lag fest. I can't tell why.

I think there are memory links, hung threads etc. This phone severely needs a windows 'task manager' equivalent to figure out whats going on.

Now I can barely load apps without getting memory errors. I think I have to blow it out and reformat for the 3rd time to get it working well again.

Frustrating.