Comments Locked

159 Comments

Back to Article

  • MadMan007 - Saturday, April 3, 2010 - link

    How were these networked to the internet? I assume the same way because otherwise you wouldn't be comparing just the devices themselves but the network as well but you never say specifically how they were connected.

    Also, 4th pargraph "...it's operating frequency" should be "its."
  • dnd728 - Saturday, April 3, 2010 - link

    There's also no mention of Flash. Was it not installed?
    I guess he was too much in a hurry to publish.
  • cassiohui - Saturday, April 3, 2010 - link

    flash? what flash?
  • Souka - Monday, April 5, 2010 - link

    There isn't flash on the iPad....
  • Quake - Saturday, April 3, 2010 - link

    Flash?! Apple doesn't permit flash in it's iPhone/iPad products!
  • MadMan007 - Saturday, April 3, 2010 - link

    Flash? FLASH? THIS IS AAAAPPLLLE!!
  • Anand Lal Shimpi - Saturday, April 3, 2010 - link

    Flash isn't officially supported on any of the devices I tested here (yet, Android is getting official 10.1 sometime soon). I tested all devices in their native shipping configuration, in other words - no flash.

    All three devices used WiFi.

    Take care,
    Anand
  • rhadife - Saturday, April 3, 2010 - link

    All devices where tested using wifi... what about cache servers on the network???? The first device to access the webpage would be the slowest... then the other devices will benefit from the cash servers.
  • ViRGE - Saturday, April 3, 2010 - link

    That's why you run multiple tests, which he did.
  • Anand Lal Shimpi - Saturday, April 3, 2010 - link

    Correct. I threw out any abnormally high results, ran each test a minimum of 5 times and noted if there was ever anything that looked strange (e.g. the Engadget results).

    If one number looked particularly fast I also re-ran the other devices at the same time to make sure all three were indeed hitting the same content. If you want network independent results, look at the SunSpider test I just added in.

    Take care,
    Anand
  • furui - Sunday, April 4, 2010 - link

    Anand, wouldn't this not prove anything processor speed wise? Apple and Google both use the WebKit browser, however Google's WebKit browser is branched off as Chromium. They aren't actually exactly the same code base and therefore aren't you testing the weakness/strength of SnapDragon + Chromium _software_ vs the weakness/strength of A4 + Apple's WebKit _software_? SnapDragon + Chromium vs A4 + WebKit is not the same thing as SnapDragon vs A4. To test comparative processor speed, you would need the same code base wouldn't you?
  • ash9 - Sunday, April 4, 2010 - link

    comparing apples and oranges is normal here-couldnt resist
  • Sharpie - Monday, April 5, 2010 - link

    Interesting to see from a user experience point of view; but not a very meaningful benchmark when comparing actual processors.
  • T2k - Monday, April 5, 2010 - link

    Anand,

    while Nexus One runs a newer Android version I was very impressed by SE's Xperia X10 when I recently tested - it runs a heavily modded Android v1.6 - on TMO's 3G with the latest Opera Mini 5 beta... it was far faster than N1's stock Android browser and about 2-3x faster than Safari.
  • Zoomer - Thursday, April 8, 2010 - link

    ---> latest Opera Mini 5 beta

    Probably why!
  • Mr Jimlad - Sunday, April 4, 2010 - link

    Hi,

    " I had been preparing for it's arrival not by downloading apps but by running comparative benchmarks"

    Far be it from me to question the test enviroment but.... these tests were not run in the same time period, network connections are not a constant.

    Also surly the review title is misleading, loading webpages isn't a cpu benchmark.

    cheers
    Jim
  • Anand Lal Shimpi - Sunday, April 4, 2010 - link

    The numbers that were published were all run within close proximity of one another as I mentioned in the article:

    "The rests were also run at around the same time to ensure that content on the sites was as similar as possible (and thus shouldn't be compared to this morning's Nexus One results). "

    I'll edit the intro to make it more clear though :)

    Take care,
    Anand
  • Alexstarfire - Saturday, April 3, 2010 - link

    What a useless article. I'd have been surprised if it was slower, but not that it's "faster." Considering the differences in hardware you can't even compare the two head-to-head.
  • EarthwormJim - Saturday, April 3, 2010 - link

    What do you mean differences in hardware. They're both phone hardware. It's perfectly valid to compare their speeds.

    What' surprising is how the A4 (and thus the iPad) is barely any faster than Snapdragon (Nexus One).

    Hardware wise it may not be any faster at all, and it all could be software optimizations (as Anand mentioned).
  • Alexstarfire - Saturday, April 3, 2010 - link

    Name one thing that's the same between them. The CPU is arguably equal, but the whole point of this article is to show how they aren't equal. The 3G and Wifi are probably the same, though the signal could be a bit better since it could have a larger antenna. Nothing is said about signal quality though. Actually, not much is really said in general about the differences/similarities. From what I know though there isn't anything similar.
  • EarthwormJim - Saturday, April 3, 2010 - link

    The iPad is essentially a phone SOC with a big battery and big screen. It's even running a phone OS. It's like comparing the iPhone to the Nexus One. They're all ARM variants too. They're also both priced similarly. It's perfectly valid to compare their web page rendering speeds. It also can be a nice preview comparison of what the next iPhone will be like speed wise.

    I wouldn't mind seeing a comparison between a similarly priced ultra-portable or netbook in this iPad review of similar tasks (like web page loading).
  • Anand Lal Shimpi - Sunday, April 4, 2010 - link

    Don't spoil the surprise :-x

    :)

    Take care,
    Anand
  • Fleeb - Sunday, April 4, 2010 - link

    Atom?
  • LuxZg - Sunday, April 4, 2010 - link

    Than we'll wait for a full review, as I believe that size/price comparison puts it closer to (at least) netbooks than it does to phones..
  • Alexstarfire - Sunday, April 4, 2010 - link

    "It's like comparing the iPhone to the Nexus One."

    And that is my point exactly. This is comparing one platform to another, not CPU vs CPU. I've got no qualms with them comparing an iPad to a phone, although I think it's stupid since it should obviously be at least a bit faster. My issue is that they are trying to compare CPU vs CPU in which case they simply can't. There are far too many variables to find anything even remotely definitive.
  • EarthwormJim - Sunday, April 4, 2010 - link

    I understand what you're saying. You're right, CPU to CPU is basically impossible to do.
  • tim851 - Monday, April 5, 2010 - link

    Whenever you are comparing an AMD CPU to and Intel CPU you are also comparing platforms. Doesn't usually bother anyone.

    People also compared PowerPC Macs to x86-PCs, where you had completely different architectures with completely different OS'es running different compiles of Photoshop and they were interpreting the results.

    All of this might be methodically wrong, but it's still kind of interesting. Although pointless in the end. Because a purchase decision for the iPad won't likely be made upon performance considerations.
  • zdzichu - Saturday, April 3, 2010 - link

    Dissappointing. CPU speed is compared based on benchmarks with diferent versions of software.
  • Griswold - Saturday, April 3, 2010 - link

    I have to agree and he even says it himself. Why compare a device with a 25Wh akku to a smartphone that has to work with much, much less capacity and thus more strict power envelope? Its like going to a gunfight with a knife...

    And then the whole software difference. Pointless from start to end.

    Bottom line is, I dont expect my smartphone to perform on par with a 1.5 pound device.
  • WaltFrench - Sunday, April 4, 2010 - link

    Pointless?

    What *would* have had a point absent a tear-down or architectural description?

    If nothing else, the article shows that, if you're sitting at home with your iPad next to your Nexus phone, you can get pages on the iPad a couple of seconds faster. Despite a couple times the pixels that get driven. Despite claims of virtually identical clock speed, which is all that many people understand about CPUs, and almost all that we officially know about the A4. Despite the work that Google has done on Webkit.

    If none of that is relevant, what drew you to go past the headline?
  • Anand Lal Shimpi - Saturday, April 3, 2010 - link

    Both platforms use smartphone class SoCs. There's no increase in footprint or thermal requirements (based on the teardown report) for implementing the A4 SoC vs. the Snapdragon (the A4 does require an external modem, however the same is true for the current SoC in the iPhone).

    The two platforms are comparable in the tests we've run. Where they aren't comparable is in things like screen size, battery life, etc... and those aren't the comparisons we're drawing. The question is whether or not the A4 is or has the potential to be faster than the Snapdragon. This matters because it's likely that the A4 or some similar piece of silicon will find its way into the next iPhone.

    Take care,
    Anand
  • WaltFrench - Sunday, April 4, 2010 - link

    EE Times is out with a report of a 64-bit data path; that would help explain higher performance at similar clock speed. There are many parts to this puzzle.
  • fabarati - Saturday, April 3, 2010 - link

    In your review, Anand, can you check some ergonomics issues that have been bothering me about all these larger tablets:
    How is it to actually use it when on a couch? Laptops are made so that you don't have to bend your neck all that much when using them in your lap, but a larger tablet, from what I can tell, are only going to work in three modes:
    1) you hold it up in one hand at eye level; this will be annoying long term (just try to do it woth a heavy book)
    2) lying down in your lap; will be cause neck pains
    3) you put up your knees and lean the tablet against them; may not always be possible.

    Thanks.
  • dagamer34 - Sunday, April 4, 2010 - link

    I'd recommend buying the Apple case which solves the issue of having to hold up your iPad all the time. It works as a stand for both typing and viewing content. A must have.
  • tim851 - Monday, April 5, 2010 - link

    How do you read a book or magazine? I'd imagine that whatever way you accomplish this will be similar to how you work with the iPad.
  • LordSojar - Saturday, April 3, 2010 - link

    Saying their hardware is different and that you can't compare them is like claiming you can't compare performance between an AMD processor and Intel processor... I mean.... really? That's just plain silly to suggest. The iPad is basically an enlarged phone with a larger battery, larger screen, and a "bit" more connectivity (and that is debatable at best). Why on Earth can it not be compared to the Nexus One or iPhone 3GS?
  • VashHT - Tuesday, April 6, 2010 - link

    That's not really the same, Intel and AMD processors are able to benchmark using the same OS and versions of software, that isn't true with these devices. This isn't really a CPU benchmark, it's more of a platform benchmark.
  • melgross - Saturday, April 3, 2010 - link

    When this is described as being "blazingly fast" as most writers have been describing it, they are thinking that in comparison to a "fast" netbook, which in my experience is sloggingly slow. Now that I've tried the iPad, it seems to work "just right". The speed is good enough so that everything works without delays or jerkiness. I suppose that when compared to products that don't do that, it's blazingly fast. It's actually fster than I expected it would be.

    But then, I'm not evaluating it with a technical eye, just for the pleasure, and that way, it's plenty fast.
  • jameskatt - Saturday, April 3, 2010 - link

    The fact that the A4 is 20% faster than the Snapdragon at the same speed (when you sum all the clearly repeatable site load times) indicates substantial optimization improvements compared to the Snapdragon.

    That was a crock to remove the smallest and largest differences in load times - particularly when the load times were repeatable - to come up with a 10% improvement.

    In real world use, it will be 20% faster. Period. And this is without the use of the GPU.

    With the use of the GPU in apps, given OS X's capabilities for using the GPU in parallel processing, it will be interesting to see how different real apps perform, other than the web browsers.
  • ganeshts - Saturday, April 3, 2010 - link

    The form factors of the devices compared are completely different, as you have noted.

    The Sanpdragon also has some image and video processing silicon to control (probably switched off during web browsing) which may contribute leakage power, and this has to be considered in the power envelop. Apple has no image and video processing in A4 to talk home about (HW accelerated video encode / decode).

    Also, Snapdragon may claim their core to be capable of 1 GHz, but most of the time it is going to be underclocked to save on battery life. Can you repeat the tests with both the Snapdragon products connected to a power outlet? (Hopefully Qualcomm was prudent enough to let the chipset run at full speed when in operation while being charged).

    IMHO, Apple A4 is very suite and optimized for doing one thing and doing that thing well (and that thing is not a well rounded smartphone experience). Credit must be given to Apple marketing rather than the tech group if A4 goes down as a success (note, not talking about the iPad, but as the A4)

    Let Apple try to put the A4 as-is on an iPhone, and the backlash would be tremendous! (Literally hot iPhones in the pocket, anyone?!)
  • metafor - Sunday, April 4, 2010 - link

    "Also, Snapdragon may claim their core to be capable of 1 GHz, but most of the time it is going to be underclocked to save on battery life. Can you repeat the tests with both the Snapdragon products connected to a power outlet? (Hopefully Qualcomm was prudent enough to let the chipset run at full speed when in operation while being charged)."

    I don't believe cell phones work like that. They don't have different power profile modes like laptops do as they're almost never expected to be used while charging. And for the most part, the actual applications processor has no knowledge of "battery vs hooked up" on a hardware throttling level.

    Android may throttle the processor more aggressively than the iPhoneOS. But we'll never know.
  • leatherseat - Saturday, April 3, 2010 - link

    You are comparing apples and oranges. The browsers are different. The support for flash is different. I would hazzard a guess that m.cnn.com is a more accurate overall speed comparison, because of the lack of flash on that particular website.

    You should also show the connection, since I'd take it the ipad is based on wifi, while the phones could be both.

    It pains me to see when anandtech is reviewing intel/AMD processors they go to the highest extent to keep most of the factors the same, while revewing anand's beloved apple products, it jumps into conclusion so fast, without proper experimental design.

    Your results do not show that A4 is better or worse than the snap dragon. It only shows that ipad browsing is faster than the nexus one browsing, under your testing conditions.
  • Anand Lal Shimpi - Saturday, April 3, 2010 - link

    My apologies for not making this clear earlier - flash wasn't enabled on any platform and all three used WiFi, connected to the same access point from the same physical distance (all 4.5 feet away with line of sight, no obstructions).

    Take care,
    Anand
  • Sapient2007 - Sunday, April 4, 2010 - link

    hi Anand,

    the difference between snapdrAgon and cortex A8 is not just clock speed. To me it SeeMs that the architectural differences Qualcomm employed to get snapdragon to get to 1GHz may be partly responsible. do you have access to an over clocked droid running at 1 GHz or an OMAP 3630 based device? So that we may check other 1GHz cortex A8 processors.
  • Crowl - Sunday, April 4, 2010 - link

    What speed was the wifi connection used, were you limiting the ipad to the same wireless g supported by the other two or was it connecting as a wireless n device?
  • ltcommanderdata - Saturday, April 3, 2010 - link

    http://www.ifixit.com/Teardown/iPad-Teardown/2183/...

    iFixit's teardown show the iPad is using 2Gb RAM chips just like the third-gen iPhone and iPod Touch. However, they conclude that the iPad is using 2 of these chips for 512MB total. This should be ideal for the multitasking support hopefully coming in iPhone OS 4.

    The K4X2G643GE RAM part number also seems to indicate that the iPad has a 64-bit memory bus compared to the 32-bit memory bus I believe you said was used in the iPhone and other smartphones so that could definitely be contributing to the speed increase.
  • Anand Lal Shimpi - Saturday, April 3, 2010 - link

    Those two chips are actually MLC NAND devices, not RAM. The DRAM is actually located on the A4 package itself. ifixit has apparently x-rayed the chip and revealed that it's got 512MB of on-package DRAM, identical to Qualcomm's QSD8250.

    Take care,
    Anand
  • ltcommanderdata - Saturday, April 3, 2010 - link

    The K4X2G643GE part number that iFixit and I are referring to is printed on the A4 chip itself presumably because it is integrated into the package as you say and refers to the DRAM. The two separate NAND flash chips have a K9LCG08U1M part number.
  • Anand Lal Shimpi - Sunday, April 4, 2010 - link

    My mistake! I thought you were referring to the NAND devices that ifixit originally referred to as DRAM. Two devices doesn't necessarily mean it's a 64-bit wide interface. I believe such a wide memory interface is still unheard of in a smartphone class SoC.

    Take care,
    Anand
  • ltcommanderdata - Sunday, April 4, 2010 - link

    In the end it looks like I was mistaken afterall in quoting iFixit on 512MB of RAM since they've since changed their interpretation of 2Gb to refer to both RAM chips together for 256MB of RAM total. This reportedly matches software developer feedback finding available free memory on the iPad being similar to the iPhone 3GS.

    I did post over at iFixit to see what they make of the 64-bit memory bus question since I believe in Samsung's nomenclature the 6th and 7th character in the K4X2G643GE part name refers to interface width.

    http://www.samsung.com/global/business/semiconduct...

    Samsung's Mobile RAM brochure only mentions up to 32-bit organization but I guess the 2 RAM chips could be arranged in a dual channel configuration for 64-bit width. Doubled memory bandwidth would be some consolation for the lack of doubled memory amount and would probably be most helpful to the GPU.

    And thank you for the time you spend in responding to readers comments. It's definitely appreciated and adds to the quality of the articles and website.
  • imgod2u - Saturday, April 3, 2010 - link

    At the end of the day, end-user performance is what matters. And in that respect, the iPad is just faster. I don't know if you can quite extrapolate that as a difference in the CPU cores themselves though; especially on something such as loading a webpage, which would be a task I'd think is far less CPU-dependent as network/software/memory dependent.

    Are there any apps out there that would be close to CPU-bound benchmarks? Even something as simple as, say, prime number calculations or image resizing (using a codec that isn't dedicated hardware accelerated) would probably give you a much better indication of the actual CPU inside the A4.
  • Devilsmurf - Saturday, April 3, 2010 - link

    Wow, talk about a bunch of ingrates !!!!!

    Look at the posted times of the last two articles from Anand. 14 hours ago for the comprehensive Nexus One review, and two hours ago for the A4 Snapdragon article.

    It's not meant to be a true comparison, but a 'preview' of the 4th generation Iphone and the possible performance based on not dis-similar hardware. It's certainly a lot more imperical than anything else that has been possible so far.

    Some points that should be addressed :

    For those quoting different hardware/OS/and software, how do you think comparisons should be done on closed OS's ? You can't just install benchmarking software on the Ipad, and browser usage is a very quick yet effective way for comparing load speeds of a site.This is especially true when the objective is to show a user comparative experiences with the devices, and the benchmarks used are a mile ahead of sensationalist media claims that the Ipad is 'blazling fast' or even as one person commented here ' just right' .

    Next, we all read this site due to the integerity and throughness that Anand has displayed over the years. Now while this article isn't chock full of benchmarks and in depth analysis, what you can rely on Anand to have done is use the same method of connecting to the internet (WiFi in this case) to make sure his abbreviated tests were as correct as possible. I don't view the fact that he left out his testing regime as a fault, especially when you consider the time he had to get this article out. Instead I see people who think he would be so lax as to use different methods of network connectivity for testing browsing speeds. The fact that this was even brought up is at best laughable, at worst insulting.

    Lastly, for those people who want to yell unfair that the Ipad is a tablet and the other's are smartphones, again I bring up the fact that this is the closest you are going to get to an Android vs next gen Iphone until the new Iphone releases. Instead of crying about what Anand has done, how about we instead just thank him for his review, and wait patiently for his Ipad review which he will be no doubt working on right now.

    I'm all for constructive criticism, and justified criticism. However I've only read two pages of comments of Wahh so far, and I took the time to register and post to redress this imbalance.

    Thanks Anand for taking the time to get this article out in a very short period of time. Looking forward to your Ipad review.
  • Makaveli - Saturday, April 3, 2010 - link

    Cry babies 0

    Adult post 1
  • KaarlisK - Saturday, April 3, 2010 - link

    Seconded
  • mickmcd - Tuesday, April 6, 2010 - link

    Good comment!
  • has407 - Saturday, April 3, 2010 - link

    Apple's A4 SoC: Faster than Snapdragon?

    Given the number of variables, this might be the kind of question/comparison where synthetic benchmarks and artificial/controlled conditions are appropriate. E.g., use javascript-based benchmarks; run tests over wifi to a local server; run plugged in and not on batteries. No substitute for real-world tests, but it might help answer the question and better highlight specific areas where there are significant performance differences.
  • Anand Lal Shimpi - Saturday, April 3, 2010 - link

    Done and done :) I just updated the article with SunSpider Javascript results, completely network independent :)

    Take care,
    Anand
  • Anand Lal Shimpi - Saturday, April 3, 2010 - link

    Also, I ran benchmarks both plugged in vs. unplugged, no performance difference.

    Take care,
    Anand
  • has407 - Saturday, April 3, 2010 - link

    Thanks! Wow... these little guys they pack a pretty impressive punch by any measure; 2.2GHz Core-2 Duo (T7500) laptop Vista IE8 clocked in at 5604, and Firefox 3.0.19 at 3165.
  • Granseth - Sunday, April 4, 2010 - link

    well my P7350 2GHz C2D did give me a result of 582 with Sunspider and Opera 10.51. So these result say much say much more about the browser than the CPU, but if you used a webkit based browser it could at least give you some indication
  • has407 - Sunday, April 4, 2010 - link

    Agree. On identical hardware/OS there's typically a 10:1 range in SunSpider performance depending on browser, but the front-runners are closing up (see e.g., http://ie.microsoft.com/testdrive/benchmarks/SunSp... However, not all WebKit browsers use the same javascript engine (e.g., Safari, which uses a proprietary Apple engine).

    As you suggest, it's a pale reflection of raw CPU performance, but it helps illuminate an area of platform performance of significance and of interest--which for these devices/appliances is an important measure--in particular for javascript/AJAX-intensive apps. Probably a good time to rethink what and how to benchmark for these types of devices. E.g., for synthetics javascript, rendering and network performance tests come to mind.
  • calvin80 - Sunday, April 4, 2010 - link

    How about some basic ones too to compare the CPU/SoC platform -- memory read write tests, CPU intensive tests, file system read/writes.
  • Chloiber - Saturday, April 3, 2010 - link

    It's a great preview, thanks for that.

    Of course it's completely valid to compare the iPad to an iPhone or a N1 or any other smartphone as they use nearly identical hardware.
    But as long as you have the difference in battery power, it's still impossible to come to a conclusion like "A4 > Snapdragon". As long as you don't have the same surroundings (or at least comparable) it's impossible IMHO.

    But I still like the short review (and hopefully more thorough review later). Because in the end, what you wanna know is how much faster the iPad is compared to your smartphone and you dont care about "the surroundings".
    Still, you can't extrapolate to a "real" mobile version of the A4 imho. It should say "iPad vs. N1" or "A4 in iPad vs. Snapdragon in N1".
  • sxr7171 - Saturday, April 3, 2010 - link

    Did anyone really think the iPad was anything but a hyped up iPod touch with a bigger screen? If anything I'm actually surprised it beats the Snapdragon ever so slightly.
  • sxr7171 - Saturday, April 3, 2010 - link

    I mean they are completely different software platforms. This is a Nexus One vs. iPad comparison.

    I bet the Apple software is highly optimized and as a platform it is faster, but we still have no way of know how SNAPDRAGON VS. A4 compares. So really this article needs a new name. Unless you can put Android on the iPad or iPhone OS on the Nexus you can't call this a CPU or SOC review.
  • BreakingStrata - Saturday, April 3, 2010 - link

    Hey Anand, lets say the A4 is just a faster clocked A8. What if the performance scales with clock speed? Then the 3GS should be roughly 60% as fast as the iPad. Well if you divide the test results...

    SunSpider Javascript test:
    iPad: 10475
    3GS: 17360
    10475/17360=0.6033982

    To me it looks like the A4 really is nothing than a faster clocked A8, which would explain the secrecy. Trying to make something magical out of nothing.

    Your thoughts?
  • kalster - Sunday, April 4, 2010 - link

    what about the processor in the newly announced samsung galaxy s. that thing is supposed to be (on paper) faster than snapdragon, will be interesting to see how it stacks up against the A4
  • TemplarGR - Sunday, April 4, 2010 - link

    I agree with all those that said this isn't an apples to apples comparison. And this isn't:

    1) It should be obvious that a much bigger device would be faster. They are not in the same class. Even if they use the same class SoC, there is room for other improvements for the bigger device. For example larger/faster RAM. Some co-processing chips. Faster interconnection etc.

    2) Nothing is said about the die area of both chips. I imagine the ipad's cpu is bigger than Nexus's. I could be wrong of course on this one, but i don't have much information.

    3) The iPad has much more room for battery and thermals, and this gives it a nice advantage.

    4) The iPad OS is much more refined than Android at the moment.

    5) Similarly the browsers are different.

    6) The smartphones never stop to act as phones, they aren't simple PDAs. There is some overhead to that, especially considering small RAM sizes.

    So in the end, this comparison is invalid. I want to believe Anand isn't biased and trying to boost iPads image, but he has made a big mistake here. Please compare the iPad to netbooks and CULVs, not Smartphones. Besides, even if the iPad's Soc finds its way into iPhone, it will be a crippled version due to smaller factor...
  • Anand Lal Shimpi - Sunday, April 4, 2010 - link

    It's not directly an apples to apples comparison, but it is the best we can do at this point unfortunately:

    1) Remember these are smartphone SoCs - memory, interconnects, coprocessors are all on the same package.

    2) Die area is unknown (and rarely quoted), but package size appears to be comparable based on the leaked FCC data and ifixit teardown.

    3) The larger battery doesn't give it an advantage in this performance test. It's possible the thermals do, but the device isn't actively cooled and doesn't have a heatsink. At 1GHz it would almost have to built on a 45nm process. A 1GHz 45nm Cortex A8 has already been shown off by both TI and Samsung for future smartphones. I don't think there's any reason to believe that the A4 couldn't make it into an iPhone sized device if it is comparable to the 45nm Samsung/TI SoCs.

    4) This is potentially very true (I didn't write either OS and refined is pretty broad, each platform appears to have its own maturity points). It's because of this that I made the following statement after the Sunspider results:

    "It could be as simple as the the iPad OS being better optimized than Android, a definite possibility given how much longer Apple has been working on it compared to Google. "

    5) Also agreed. Keep in mind that the software advantage can't be *that* great because the Nexus One still manages to step all over the iPhone 3GS.

    6) That's very possible, this goes back to the optimization point earlier. I don't believe any overhead present would make up the gap we saw between the Nexus One and the A4 however.

    I'll be comparing the iPad to both netbooks and smartphones in the review. The fact of the matter is that it bumps into the netbook category in terms of functionality, but has the performance characteristics of a high end smartphone - making a comparison to both very important.

    Take care,
    Anand
  • bsoft16384 - Sunday, April 4, 2010 - link

    Anand,

    Remember that the iPad is running a newer version of the iPhone OS than the iPhone 3GS (3.2 vs 3.1). Considering the recent JS performance wars, it's entirely possible that the JS engine shipped with the iPad is faster than the JS engine shipped with the iPhone 3GS (since Apple/Google/etc. are constantly working on their JS engines).

    I would be hesitant to compare JS results between different software platforms and versions.

    As I pointed out, my 2.53GHz Core 2 Duo ThinkPad T400 is 29 times faster than the iPad in the SunSpider benchmark, and that's with one core active. That works out to 11x faster per clock, which is absolutely stupid. Some of the difference is certainly due to Core 2's vastly superior 2 IPC (much bigger caches, out of order execution, higher memory bandwidth, better branch prediction, 4-issue vs. 2-issue, etc.), but I'm willing to bet that at least part of the difference is due to the fact that the Chrome build I'm using (latest beta) has a better JS engine than the iPad.

    The bottom line is that JS benchmarks aren't really useful comparing CPUs when you have different JS engines (or at least different JS engine versions). Chrome, for example, is twice as fast as Firefox 3.6 on my T400.

    I think you're probably right that the A4 beats the Snapdragon, but it's hard to tell. People keep saying that the iPad is "super fast", and, compared with the typical smartphone, they're right. But the iPad isn't a smartphone and it doesn't compete with or replace a smartphone. What it does compete with are netbooks and CULV notebooks, and it's nowhere close to the former - let alone the latter.

    I'm going to go run some benchmarks on my Mom's 1.4GHz Core 2 Solo Acer 1410. I'm betting that Engadget is going to load in way less than 14 seconds.
  • bsoft16384 - Sunday, April 4, 2010 - link

    The Acer 1410 (1.4GHz Core 2 Solo) took 4.6 seconds to load Engadget (Chrome Stable 4.1.249, Flash disabled). My guess is that an Atom netbook would be around the 6-8 second mark.

    SunSpider on the Acer 1410 was 729.8ms (Chrome Stable 4.1.249), or about 14 times faster than the iPad (and about half as fast as the ThinkPad).

    Looks like ARM Cortex A8 (even in the form of the A4) still can't hold a candle to Core 2 in terms of performance.
  • Anand Lal Shimpi - Sunday, April 4, 2010 - link

    I agree that the software differences do contribute to the performance difference - the question is: how much? Unfortunately we'll probably never get an identical software stack running on A4 vs. Snapdragon, so this may end up being the closest we get.

    I've updated the title and added some text in the article itself to hopefully make the comparison and its potential influencers more clear.

    Take care,
    Anand
  • misaki - Sunday, April 4, 2010 - link

    I wouldn't be so sure about #5.
    While the base rendering engines are all based on webkit, the javascript engine in android/chrome is completely google's own code (V8), and NOT from webkit. Chrome's release of the V8 engine really lit a fire under all the browsers to work on performance tuning and the comparison to even a year ago is astounding.
  • bsoft16384 - Sunday, April 4, 2010 - link

    I'm not so sure V8 is faster. Safari 4.0 on my Acer 1410 did the SunSpider benchmark in 820ms, which isn't far behind the Chrome result.

    All of this is academic since you can't compare CPU performance by looking at a JS benchmark running on different engines on different CPUs.

    Unless Safari on Windows is magic though (or the iPad JS engine is terrible), it does seem that a Core 2 Solo at 1.4GHz basically destroys the A4 in the iPad. Which kind of makes the fact that the OS and browser are decently responsive that much more impressive.
  • TemplarGR - Sunday, April 4, 2010 - link

    1) To clarify, it may be a SoC and all major systems on package, but it may use for example more advanced parts like WiFi controllers etc. As i said i do not claim to know what exactly is in there but it may affect performance.

    3) The larger battery allows more power for the same autonomy, but i believe most will be used for the bigger screen anyway. My comment was more about thermals, but more power available could have some impact in the form of fewer cuts needed.

    My overall point is that we cannot compare CPUs directly in those devices, and iPAD is obviously faster than Nexus. If it was't faster, it would be a disaster considering their difference in size and role.
  • Anand Lal Shimpi - Sunday, April 4, 2010 - link

    1) All devices have WiFi support. The implementation of WiFi may differ across them but the SunSpider test is network independent so that should rule out WiFi as a cause of the gap. All in all it boils down to the SoCs (which are comparable) and the software stack (which aren't).

    3) The verdict isn't out yet on thermals. Like I said, there's nothing that blatantly screams the A4 is running too hot for a smartphone and other manufacturers will be shipping 1GHz Cortex A8s with PowerVR SGXs into smartphones later this year. Expecting this level of performance out of the next iPhone I don't believe is unreasonable.

    I agree that we can't make a perfect CPU comparison between the A4 and Snapdragon, however I don't believe that means we shouldn't try :)

    Take care,
    Anand
  • Cali3350 - Sunday, April 4, 2010 - link

    1. They are absolutely in the same class. The iPad uses the same hardware as phones. Its larger in terms of screen real-estate and dimension, however NOT in terms of whats actually powering the thing. There are numerous places to see this
    http://www.ifixit.com/Teardown/iPad-Teardown/2183/...

    2. Package size is identical when comparing actual dimensions, which makes complete sense given number 1.

    To say something along the lines of "I want to believe Anand isnt biased..." when you have so little information with which you base your claims is simply foolish, and thats exactly how your coming off here. The Nexus One, the iPhone 3GS, and any other smartphone, are directly comparable to the iPad from a pure hardware standpoint which is the reason this article exists and why Anand is so excited to test this processor.

    Now, your comments about the OS are true, however out of Anand's hands. There is no "Super PI" port that runs identical code via a native app for all systems, so he needs to use javascript and the like as a baseline.
  • Rob94hawk - Sunday, April 4, 2010 - link

    The iphone is one thing but people lining up for this overrated piece of crap is borders on irresponsible spending. But you know the saying, "there's a sucker born every minute."
  • TemplarGR - Sunday, April 4, 2010 - link

    From what i can understand, this is Apple's strategy overall. To target suckers. I may be wrong, but apart from design i see nothing special in all Apple's devices and OSes.

    I haven't done a detailed benchmark on my Acer Aspire One netbook, which runs ArchLinux with KDE, but it certainly loads pages much faster than iPad. To me this means a lot, as my netbook is 8,9", comparable to the iPad. Of course it is not as thin, but it sports a big hard drive, a web cam, two card slots and a true keyboard.

    It should not be difficult for netbook makers to clone the iPAD, and they will trash its performance easily, while having much more software AND Flash support... I believe they will also be cheaper. So when this happens, anyone still buying the iPAD could be considered a sucker...
  • metafor - Sunday, April 4, 2010 - link

    I'm curious whether it's hardware or software. From the numbers of the SunSpider numbers, it would appear the A4 is exactly what a Cortex-A8 would be at 1 GHz (60% higher than the 600MHz A8 inside the iPhone).

    To isolate software vs hardware, just take a Cortex-A8 Android phone using a similar version of Android (1.6 I believe?). The Droid is the closest thing I can think of OTOH. But I'm sure you know more than I do.

    I believe the Droid and N1 also use the same browser? If it isn't too much trouble, put up the Droid performance numbers and we can probably extrapolate how much of the difference is software/browser based.
  • metafor - Sunday, April 4, 2010 - link

    NVM, found them:

    http://arstechnica.com/gadgets/reviews/2010/01/nex...

    Droid's 550MHz A8 on Android 2.0 was significantly slower than the iPhone 3GS's 600MHz A8 running 3.1.2 of the iPhoneOS. What's interesting is that, since the N1's release, the browser has apparently gotten faster compared to the 3GS.

    Also, this doesn't take into account the rest of the SoC hardware.
  • Alexstarfire - Sunday, April 4, 2010 - link

    I took a look at that site and one of the charts is labeled incorrectly. If you take a look at the part where they show the average load times for the iPhone you see that with the except of 2-3 websites that all the phones are pretty equal. For 2 of the websites the iPhone is significantly slower, yet when you look right below that chart where they just look at the average load time over all the web pages you see that the iphone somehow comes out below the Droid. I did the calculations myself and based on the previous chart the iPhone should be the one loading in 9.3 seconds, not the droid.

    I don't which one they got wrong, but because of that you can't really use any of the results that are on that site.
  • metafor - Sunday, April 4, 2010 - link

    Arstechnica is a pretty reputable site. In either case, with data as sparse as it is, it's the closest source I can find. Added Droid data would be awesome using the same test setup used here.
  • Alexstarfire - Sunday, April 4, 2010 - link

    They might be, but that doesn't change the fact that something is simply WRONG. Without knowing which is which, in other words without them actually reviewing the data to make sure it's accurate, it's all useless.
  • Brian Klug - Sunday, April 4, 2010 - link

    You find the 60% increase not surprising? Since I generated those numbers, I have to disagree. Remember that the increase from 614 MHz to 1 GHz represents a roughly 30% increase in clock speed, but the JavaScript performance is double that at 60% better.

    I find that surprising. It's clear to me at least that there is much more at play here that we haven't uncovered yet. I pulled the user agents from both browsers and found that while they're running different versions of webkit (unsurprising), theyre not that different. It'll be interesting to see if the performance is architectural or software. Likely its some combination of the two.

    Cheers,
    Brian Klug
  • misaki - Sunday, April 4, 2010 - link

    You can get results like this just from comparing different versions of webkit and google chrome on the same pc... They are constantly leapfrogging each other with every version bump.

    All this says is iPad is faster than the Nexus One, not how well the CPUs perform.
  • bsoft16384 - Sunday, April 4, 2010 - link

    My experience today (about 30 minutes with the iPad) was also that it was "acceptable" but not "fast".

    I'm not sure why everyone keeps calling the iPad "super fast". Yeah, it's faster than the iPhone 3GS, and it's a bit faster (looks like about 20%) than the Nexus One.

    Here's the problem: the Nexus One isn't exactly fast, and the 3GS, while fast for a phone, is positively anemic for a PC.

    When I used the iPad today, small sites (like Google) loaded plenty fast, but more complex sites like Engadget were pathetically slow. Engadget took around 25 seconds for me (possibly because it was loaded with huge images this afternoon), which is stupid. And, no, it wasn't the Internet connection, since the MacBooks at the same Apple store were very fast today (the Apple store in Boulder, CO has at least a 45Mbps dedicated connection from AT&T; I've benchmarked it before).

    In the days of dial-up, 256kbps DSL, and Pentium III CPUs, 15+ seconds to load a site might have been acceptable. But today even a moderate broadband connection (like my 12Mbps Comcast) and a moderate CPU (like the 2.53GHz Core 2 Duo in my ThinkPad) can load even big sites like Engadget in a couple of seconds.

    Oh, and by the way - my 2.53GHz Core 2 Duo T400 does the SunSpider benchmark in 351ms (latest Chrome beta). That's 29 times faster, or 11 times faster per clock. And that's with one core (I locked Chrome's affinity to CPU0 only; the JavaScript engine doesn't benefit much from the second core).
  • metafor - Sunday, April 4, 2010 - link

    True but that Core2 of yours is eating up ~20W. These SoCs, including the graphics engine, are taking up less than 1W.

    Also, the Pentium III at 1GHz would still be faster than these Cortex/Scorpion based CPU's. It was 3-issue (more if you worked the 1-1-3 rule) out-of-order processor.
  • 96redformula - Sunday, April 4, 2010 - link

    How are you going to compare Intel Atom Netbook vs iPad without having backround processes in windows(manufacturer installed software) affecting the Netbook. Are you going to do a clean install of Windows 7 on the netbook?

    I am asking this knowing well that many people are seeing poor performance on many of the netbooks due to bloated manufacturer software bogging down the Atom CPU. I.E. Lenovo S10-3t is a good example of one with lots of crud pre-installed bogging down the processor, people seem to be seeing big differences reinstalling the OS as there is so much preinstalled garbage running.
  • TemplarGR - Sunday, April 4, 2010 - link

    You are wrong, 600mhz to 1ghz is not 30% increase in speed, it is ~65%.
  • Brian Klug - Sunday, April 4, 2010 - link

    Wow, I have no idea what i was thinking yesterday, (1000-614)/614*100=62% ;)

    I failed at my arithmetic last night. Sorry guys ;)

    -Brian
  • duda - Sunday, April 4, 2010 - link

    This is the biggest fail on an article I have read in a long time. In fact I registered just so I could point out why this is sooo bad. The writer is probably some Apple douche noob.

    You are trying to compare PROCESSORS, yet you end up testing the web-page loading speed of two completely different devices. Geez - absolutely inconclusive!! There are so many things wrong with this article I don't know where to start.

    Ok, Analysis 101:

    If you are going to compare two things, and two things only (in your case the Apple A4 Processor, and the Snapdragon Processor). Then you have to make sure that any other variable factors are either eliminated or minimised. If you don't your results and measurements may be influenced by something you are not testing (in your case, the whole rest of the hardware and software of the devices!)

    You are comparing a Google Nexus One to an Apple iPad, the two devices are completely different. You can't say you are comparing the processor of the devices when there are sooo many other variable factors involved in taking your measurements, such as the networking and graphics hardware. You really don't think the Processor does everything do you??

    Secondly, you are not even measuring output and functionality in the processors, you are measuring how fast web-pages load.

    What this article should really be saying is, "How does the Apple iPad match up with the Google Nexus One on Webpage loading times?"

    There you go dumb a..

    Have a good Easter :)
  • Screammit - Sunday, April 4, 2010 - link

    Holy not-getting-the-point, batman!

    I'm going to save some time and energy and direct you to DevilSmurf's post, on or around page 4 of the comments, that person stated it better than I ever could.
  • tushaar - Sunday, April 4, 2010 - link

    As you said, the better performance might be because of the larger battery and this might mean that the A4 in the new iPhone might NOT surpass the Nexus One.
  • Chloiber - Sunday, April 4, 2010 - link

    This would be the same if you said:

    500Mhz to 1000Mhz is an increase of 50% in clock speed. Obviously that's wrong - it's double or 100%. 1000Mhz -> 500Mhz != 500Mhz -> 1000Mhz
  • vigg0 - Sunday, April 4, 2010 - link

    The soon to be released Samsung Galaxy S has a 1 Ghz Cortex A8 combined with a PowerVR SGX540. Given the fact that Samsung is producing the A4 SoC for Apple, could it be so that iPad/Galaxy use the same hardware?
  • LyCannon - Sunday, April 4, 2010 - link

    Now, from what I have read, the iPad is supposed be be a replacement for a netbook. If that's the case, why the heck are you comparing it against a smartphone? Shouldn't you compare it against other touch-based netbooks in the same price range?
  • misterPaul - Sunday, April 4, 2010 - link

    A lot of other sites are linking to this article as a review of the iPad's CPU. Reading the defensive comments here it's obvious that this is more a comparison of the CPU expected to be used in the 4 gen iPhone.

    The fact is there's a lot of hate for this tablet and Apple in general at the moment. Don't ask me why, I guess it's what they represent - tightening up control of hardware. Anyway, the fact is this isn't a review of the iPad, as the iPad isn't a smartphone and doesn't really share many characteristics with them (except the poor hardware and limited user input methods).
  • yacoub - Sunday, April 4, 2010 - link

    Your numbers are calculated the way a marketer would do them (% faster), instead of the real world / conservative method that most folks would appreciate (% improvement, or the difference in the amount of time from start to finish between the two devices).

    For example, the first benchmark you state 6.2 vs 8.8 is "41% faster", yet in reality the improvement seen over the other device is 29.54%, because the iPad at 6.2s takes roughly 70% of the time the Snapdragon takes at 8.8s. The correct formula is ((6.2 / 8.8) * 100) -100
    It's taking ~70% of the time that the Snapdragon takes to load Anandtech.com.

    I know, I know, the way you're doing it is also "right" in a strictly mathematical sense, but it's a math trick used by marketers to be able to post higher numbers, but it rubs the wrong way for a technical review site to be doing that.

    Here are the correct numbers for each site load, and as we see, the bigger the difference between the two devices' scores, the more pronounced the exaggeration in the marketing numbers, which is why they like to calculate that way (% faster) instead of what matters more (% improvement):

    A4 percentage improvement over Snapdragon:
    29.54%
    7.8%
    8.14%
    29.09%
    20.93%
    11.9% (you actually got pretty close on this one (TechReport). Using the incorrect/marketing calculations you used with other sites, i think this one should have been 13.5%?)
    37.27%
    15.57%
    -15.38% (m.cnn.com is still slower on the A4, as noted)
  • Anand Lal Shimpi - Sunday, April 4, 2010 - link

    I did it for consistency with previous performance testing in smartphone reviews (I started out doing it the way you're describing).

    The % reduction in time and % faster are both valid ways of expressing the data here. Folks usually like to know the latter in my experience, but if the overall demand is for the former I've got no problems switching :)

    Take care,
    Anand
  • yacoub - Sunday, April 4, 2010 - link

    Thanks for the reply, Anand. You're right that they're both valid, however, I think when most people ask "how much faster is it?", what they really want to know is "how much faster am I actually going to be loading the page/site?". After all, that's the real-world usage you're examining with that test. And that is what the % improvement numbers tell. The other method, "% faster", is traditionally just a way for marketing to make the differences seem greater.
  • Mumrik - Sunday, April 4, 2010 - link

    "Jon Stokes recently stated that Apple's secrecy surrounding the chip is because it's anything special, just a Cortex A8."

    I guess that should be "it isn't anything special" ?
  • Anand Lal Shimpi - Sunday, April 4, 2010 - link

    Fixed :)
  • Mumrik - Sunday, April 4, 2010 - link

    Hopefully you'll be comparing the iPad to netbooks/notebooks in the same price and screen size range in the review and not just phones?
  • vigg0 - Sunday, April 4, 2010 - link

    I agree, the iPad should be compared to equally priced laptops/netbooks, not phones.
  • Flunk - Sunday, April 4, 2010 - link

    I'm interested in how Futuremark's Peacekeeper test would rate these mobile devices. As far as synthetic web rendering tests go it's pretty comprehensive.
  • doctorpasty - Sunday, April 4, 2010 - link

    Newbie question, but why is it that an iPad is still way slower at loading a webpage than a regular Mac/PC? For example, the Ars Technica website loads in about 2 secs on my Mac. Is it software, hardware, or both that causes this difference? I've always wanted to know!
  • Anand Lal Shimpi - Sunday, April 4, 2010 - link

    It's software being limited (either at runtime or during development) by much slower hardware. For starters, the CPUs in these SoCs are all dual issue, in order cores compared to the much wider, out of order cores we have in netbooks, notebooks and desktops today. If you want to understand what that means and how it impacts performance, I'd suggest starting here in my iPhone 3GS review:

    http://anandtech.com/show/2798/2

    Take care,
    Anand
  • doctorpasty - Sunday, April 4, 2010 - link

    Great primer, thanks for taking the time to let me know about it!
  • Anand Lal Shimpi - Sunday, April 4, 2010 - link

    I just wanted to chime in and thank everyone for the discussion. I've taken the comments to heart and clarified the comparison in the title as well as further clarified it in the review itself just so no one is mislead by what we're doing here.

    Thanks again :)

    Take care,
    Anand
  • Chloiber - Sunday, April 4, 2010 - link

    Is it somehow possible to change the way comments are displayed? Right now I always have to scroll through every comment, because even on the first page there could be a reply to a really old comment. That's somehow really annoying :-(

    @Anand

    I really appreciate your work and how you respond to comments - rarely seen on the web todays :)
  • Goffers - Sunday, April 4, 2010 - link

    With this test aren't you mainly comparing GPU's?
  • Anand Lal Shimpi - Sunday, April 4, 2010 - link

    The GPU isn't involved with web page rendering, it simply displays the final image but the scanning, parsing and interpreting of the HTML code and running anything on the pages is all done on the CPU.

    Take care,
    Anand
  • Zink - Sunday, April 4, 2010 - link

    Anand, if you have some time run some atom netbook benches.
  • PhilipHa - Sunday, April 4, 2010 - link

    Over at pocket gamer:

    http://www.pocketgamer.co.uk/r/iPad/Apple+iPad/new...

    In terms of performance, benchmarking the iPad's OpenGL ES performance suggests it has a capability of 1.83 million triangles per second.

    That compares to 2.15 million/sec for the MacBook Air (which runs on a 1.6 GHz Intel Core 2 Duo CPU with 2 GB of RAM) and 0.423 million/second for the iPhone 3G (ARM1176 clocked at 412 MHz with a PowerVR MBX Lite GPU).
  • skyward03 - Sunday, April 4, 2010 - link

    It is a nice start in review SOC for smartphone. I for one would like to see how it scale when it is overclock. I also would like if there were test done on the pre as well.
  • ecophreak - Sunday, April 4, 2010 - link

    I think that it's a reasonable comparison to make between the smartphones and the ipad, simply because many users will be using internet browsing on either in similar situations i.e. at a cafe or elsewhere with wifi access, personally I like the idea of the ipad, but not at apple's prices.
  • windywoo - Sunday, April 4, 2010 - link

    As others have said, the iPad is competing with netbooks so it would be nice to see some comparable benchmarks for one of those devices. Yes I know the hardware is not the same, and will vary greatly from netbook to netbook, and will be affected by the OS and other things. I would still like to see it because I am sick of Apple's products selling on the hype and their flashy marketing.
  • Anand Lal Shimpi - Sunday, April 4, 2010 - link

    Oh it's coming... :)

    Take care,
    Anand
  • joe_dude - Sunday, April 4, 2010 - link

    Netbooks have already used the PowerVR SGX 535. Intel rebranded it as the GMA 500 for the Atom Z500.

    So probably the key difference in speed for the iPad vs. a netbook is the fact the iPad runs a stripped down OS with no multitasking.

    And the fact apps have to run in 200 MB of RAM (or whatever it turns out to be) user space means it's gonna be snappier compared to full-blown apps on a Mac or Windows.

    But of course this mean "HD" games on the iPad are really nothing more than enhanced iPhone or portable games.
  • johnrb85 - Sunday, April 4, 2010 - link

    The iPad support 802.11n, the iPhone 3GS only supports 802.11b/g. Just wanted clarification that you were not on a 802.11n network. Otherwise all you proved is that 802.11n is faster than 802.11g.
  • Anand Lal Shimpi - Sunday, April 4, 2010 - link

    Correct, 802.11n was disabled on the router - only 802.11b/g were used.

    Take care,
    Anand
  • ltcommanderdata - Sunday, April 4, 2010 - link

    http://furbo.org/2010/04/03/benchmarking-in-your-l...

    Craig Hockenberry has done some benchmarks to compare the native performance of Cocoa Touch apps between the iPad and iPhone 3GS. He is seeing about 2 times CPU performance improvement between the iPad and iPhone 3GS. Granted the iPhone 3GS results are older results using the original iPhone OS 3.0, so presumably it'd be a little tighter if iPhone OS 3.1 is more optimized, but I think we're still seeing a bit better than pure clock speed scaling between the iPad and iPhone 3GS. Interestingly, the iPad is more than 8000 times faster than the original iPhone 2G in sin(x) calls probably due to the stronger FPU and NEON instruction set.

    http://twitter.com/zodttd/status/11592511853

    As well, the PowerVR SGX535 has been confirmed as the GPU in the iPad. This is the same one as in the iPhone 3GS, just presumably clocked higher. I guess this isn't unexpected since the SGX535 has 2 TMUs and presumably ROPs just like the faster SGX540 and SGX545. The SGX540 and SGX545 have more ALUs for more shading power, but don't really have more fillrate than the SGX535 which is probably what was important for the iPad's larger screen in comparison to the iPhone. Since they aren't leaving fillrate on the table Apple probably stuck with the familiar SGX535 and kept the power savings compared to the higher shader power of the SGX540 and SGX545.
  • shiboos - Sunday, April 4, 2010 - link

    Why are you comparing a smartphone with iPad? Isn't iPad vs Snapdragon-powered smartbook a better comparison?
  • MGS88 - Sunday, April 4, 2010 - link

    Opera is my 2nd browser (even if its in beta) after Dolphin witch both faster than original one.
    My test of Opera 5.0.18302 on N1:
    www.engadget.com: 8.4 & 7.7 & 9.2 & 7.1 & 8.4
    www.anandtech.com: 6.9 & 6.2 & 7.9 & 8.6 & 12.3
    www.tomshardware.com: 4.8 & 4.9 & 6.9 & 5.4 & 5.7
    www.gizmodo.com: >>> have some problem with this site. some times loads fine other takes for ever
  • windywoo - Sunday, April 4, 2010 - link

    gizmodo isn't even worth bothering about these days since they are all iPad maniacs there. I know that it gets a lot more traffic from iPad stories, but the tone is always one of fawning awe.
  • rdeva - Sunday, April 4, 2010 - link


    Its not just processor metrics, but (android vs iphone-os), (wlan stack/chip performance) parameters also impact greatly..

    its completely misleading to use browser loading time to processor metrics!!!
  • joe_dude - Sunday, April 4, 2010 - link

    With limited RAM, CPU and GPU power, this thing's only good to play 5 to 10 year old games. No wonder Scrabble and Tetris were launch titles.
  • nitussi - Sunday, April 4, 2010 - link

    Fanboy .. there is no way you telling me that running a browser is a good benchmark comparison of similar type cpu's with different operating systems.. very misleading.
  • TopDNBass - Sunday, April 4, 2010 - link

    The Pad dimensions:
    Height:
    9.56 inches (242.8 mm)
    Width:
    7.47 inches (189.7 mm)
    Depth:
    0.5 inch (13.4 mm)
    Weight:
    1.5 pounds (0.68 kg) Wi-Fi model;
    1.6 pounds (0.73 kg) Wi-Fi + 3G model

    The Nexus dimensions.
    Height
    119 mm
    Width
    59.8 mm
    Depth
    11.5 mm
    Weight
    130 g (with battery)
    100 g (without battery)

    The nexus is a fuckin' PHONE, with an overclockable 1gHz processor, 512MB RAM, GPS, 3G, Multitouch, Flash coming soon, Amazing everything.

    Compared to a damn Pad, <256MB RAM, OPTIONAL 3g, will NEVER have real multitasking, and probably not FLASH, since their trying to make HTML5 the new standard...

    If you haven't noticed, the Pad is 2-3x bigger than the nexus in every way.
    Yet still has a disgustingly low amount of RAM, and you have to pay extra for 3g.
    If cr@pple couldn't get it right with that much space, and the best CPU they could churn out just barely compares to a phone's imagine what google and htc could do!

    Pad, pick on someone your own size! You don't see me comparing phones to netbooks or laptops!
    Not to mention this is a flawed benchmark in every way..
  • TopDNBass - Sunday, April 4, 2010 - link

    The Pad dimensions:
    Height:
    9.56 inches (242.8 mm)
    Width:
    7.47 inches (189.7 mm)
    Depth:
    0.5 inch (13.4 mm)
    Weight:
    1.5 pounds (0.68 kg) Wi-Fi model;
    1.6 pounds (0.73 kg) Wi-Fi + 3G model

    The Nexus dimensions.
    Height
    119 mm
    Width
    59.8 mm
    Depth
    11.5 mm
    Weight
    130 g (with battery)
    100 g (without battery)

    The nexus is a fuckin' PHONE, with an overclockable 1gHz processor, 512MB RAM, GPS, 3G, Multitouch, Flash coming soon, Amazing everything.

    Compared to a damn Pad, <256MB RAM, OPTIONAL 3g, will NEVER have real multitasking, and probably not FLASH, since their trying to make HTML5 the new standard...

    If you haven't noticed, the Pad is 2-3x bigger than the nexus in every way.
    Yet still has a disgustingly low amount of RAM, and you have to pay extra for 3g.
    If cr@pple couldn't get it right with that much space, and the best CPU they could churn out just barely compares to a phone's imagine what google and htc could do!

    Pad, pick on someone your own size! You don't see me comparing phones to netbooks or laptops!
    Not to mention this is an inherently flawed benchmark in every way..
  • TopDNBass - Sunday, April 4, 2010 - link

    Sorry about the dbl post, I can't edit?

    I also wanted to add, the nexus has a camera with flash, it's a PHONE so it makes CALLS.Has live wallpapers, revolutionary Voice-To-Text, amazing apps, not a bunch of b.s. like in the app store.

    I've lost respect for you Anand, you put this under smartphones? Is apple really so far behind that they need to inflate the size of an iphone by more than 3x to compare to an HTC cell phone?
  • colddarkdrink - Sunday, April 4, 2010 - link

    Palm Pre Plus Web OS 1.4.0

    sunspider benchies
    I got 17360@600
    20734@500
  • vol7ron - Sunday, April 4, 2010 - link

    Despite the problems mentioned w/ comparing "Apples" to oranges - pun intended - I am thankful for including the IPhone 3GS in there. I think this is the best that can be done, since the 3GS and the iPad are almost the same product. You didn't design the product, so there really is nothing in the same class as the iPad. It's less than a tablet, or a netbook, but a little more than an iPod Touch.

    Thanks Anand.
    I'm looking forward to future reviews of the iPhone 4g and some of the other smartphones when they're out.

    vol7ron
  • nitussi - Sunday, April 4, 2010 - link

    The ipad may bring up a web page over wifi faster than my quad-core laptop.. but there is no-way in hell it has a faster processor. Hell, my laptop brings up a webpage faster with chrome than it does with explorer, so what does that mean?
  • Kidster3001 - Sunday, April 4, 2010 - link

    Thanks for the article.

    The iPad is not a netbook replacement, nor is it a smartphone replacement. It doesn't replace either and isn't intended too. Anyone who wants to use one to replace one of the other platforms will not like it. Anyone who expects it to compete against either of those platforms is crazy.

    It's an appliance, a toy, whatever you want to call it, to allow for nearly instant access to small things quickly. When the wife and I are watching a movie and can't remember the actors name we can get it in 10 seconds with the iPad. I can check forum posts or blogs in a few seconds in the morning over coffee These things can be done with a net book or smartphone, but the tablet makes it so much more convenient.

    Posted from my iPad in the Lazy-boy
  • Alexstarfire - Sunday, April 4, 2010 - link

    No offense, but more convenient? Really? You do know that there have been other, better tablets released already, right? Not to mention IDK how a tablet is either to read than a laptop/netbook. I suppose it could be on par if you want to spend extra money to get the stand, but that's about it.

    It's not much use trying to use logic to argue with those who defend most Apple products though. If you guys want your stuff for looks, ease of setup (assuming iMac anyway), and essentially not having to worry about viruses/malware (if only for the moment) then that's fine. I just find it laughable when people try to defend them in any other way. This especially includes the iPad.
  • T2k - Monday, April 5, 2010 - link

    "When the wife and I are watching a movie and can't remember the actors name we can get it in 10 seconds with the iPad. I can check forum posts or blogs in a few seconds in the morning over coffee These things can be done with a net book or smartphone, but the tablet makes it so much more convenient."

    You iTards are hilarious - which one of these features are more convenient on a coffee table-sized iPod than on, say, any 8-9" tablet or laptop, let alone a decent 4" smartphone?

    You guys are so hilarious with your twisted, idiotic (ill)ogic, bending backwards to justify your latest mandatory BOGU-purchase* from Apple...

    *: Bend Over, Grease Up, here comes this year's 'must-have' Apple (tax)...
  • nitussi - Sunday, April 4, 2010 - link

    The Nexus one is a phone which has to do a lot more than the ipad does..

    The problem to me is, with todays technology I expect to have control over how much ram is in my device and being able to change the battery without sending it in for service when it dies in two to three years. I prefer not to have to spend an extra $100 for 16GB.. $200 for 32GB.. and a wopping $300 for 64GB for memory. No usb port! This is a oversize ipod and peeps are being played with this ipad just like back in the day.
  • Some1ne - Sunday, April 4, 2010 - link

    While I understand that the iPad is little more than a super-sized iPhone, it is still marketed as a tablet PC. And I would fully expect a tablet PC to trounce a smartphone in terms of performance. If anything, I think the fact that the numbers are reasonably close in most cases shows that the iPad is woefully underpowered given that it's a completely different class of device. It should be able to handily outperform any smartphone, and not just by 10-20%. It really should be more like 100-200% faster, all things considered. Why would I want a tablet PC that can only just barely outperform my phone?
  • bsoft16384 - Monday, April 5, 2010 - link

    Exactly. The iPad is not being sold as a smartphone with a bigger screen. It's being sold as an alternative to a netbook or tablet PC, both of which are dramatically faster.

    I just ran SunSpider on an old Athlon XP 1.2GHz laptop (with 512MB of SDRAM and Windows XP). It completed in 2550ms, which is about four times as fast as the iPad.

    No one expects the iPad to be as fast as a netbook. I think that you can legitimately argue that the iPad's performance is perfectly acceptable. What you can't argue is that a device that takes 10 seconds to load Digg (or 14 to load Gizmodo) is "super fast".

    The iPad is faster than a smartphone, but slower than the products (netbooks) it competes with.
  • gubatron - Monday, April 5, 2010 - link

    And why would you benchmark the CPU of a tablet with the CPU of a smartphone?

    It's like comparing the engine of a motorcycle with the engine of a car.

    It beats me how these articles make it to the homepage of techmeme.

    Sorry, but this is trash.
  • dugbug - Monday, April 5, 2010 - link

    It is in context for the chipsets they have in common, and if you read the article you would understand that.
  • joe_dude - Monday, April 5, 2010 - link

    I think the ease of use with the iPad is a given. All I want to know is if it can play games, not Scrabble. Is this a serious gaming platform that it was so hyped to be?

    The big two (personally) are Internet and games. We know it can't do Flash, which is irritating, but I don't play Farmville.

    But games so far suck on the iPad. Plants vs. Zombies? Vector Wars (1980's gaming)? "HD" racing games with washed out backgrounds? I'm not paying $500 to $800 to play with gadgets.
  • Extremophile - Monday, April 5, 2010 - link

    First, Anand, I believe that your test is fair enough and interesting.

    I have read the negative comments, and they can all make their point. But - honestly - comparing apples and oranges is still better than not comparing at all. And no item was downgradede to become a "lemon" in your test.

    Your test gives a "quick and dirty" preview of what we have to expect from coming hardware. Even if we accept the influence of the different hardware and software environmentsof the SoCs, results are quite conclusive.

    Second:

    What I missed: Whenever I open a web page on a smartphone, the next step is to enlarge it so that I can read the text. This is at least true for normal HTML pages.

    If you add the time needed for this manual "rework" after loading, how would the three appliances compare?

    Maybe, IPad can justify its existence mainly by saving this process step?
  • dentaldoc - Monday, April 5, 2010 - link

    Listening to so many apoplectic responses makes me ponder. Just imagine if the test results had been just slightly different: Nexus 1.5% faster than either of the (bad, mean, corrupt, greedy) Apple products.

    Comment 1: Ah....

    Comment 2: What a great article

    Comment 3: What a fair comparison

    Comment 4: You folks at Anand are AWESOME!
  • longhorntrojan - Tuesday, April 6, 2010 - link

    Are you planning on doing a battery assessment?

    The iPad does not seem to allow you to loop video. :(
  • AdamBv1 - Wednesday, April 7, 2010 - link

    Some other things to take note, the iPhone 3GS, Moto Droid and Palm Pre all have the same processor at 600MHz but when you look at sunspider tests you see something very interesting.

    iPhone: 17k
    Palm Pre: 20k
    Droid: 34k

    The nexus one has a snapdragon (another ARM Cortex A8 running at 1GHz) and gets 14k. Take that 17k the iPhone gets and multiply by .6 and you get the 10k the iPad scores. Run a palm pre at 800MHz and you get 14k for the sunspider test.

    At least to me this says that Apple has a much better mobile browser than Google or Palm, not that the processor is that much better.
  • manicfreak - Wednesday, April 7, 2010 - link

    If one is going to compare a chip from a smartphone and a chip from a tablet, then why not compare it to Pine Trail?

    And I hope Anand stays true to his word from a previous article... about the Cortex A8, which is used in the Apple A4, is dead to him.
  • aka2 - Wednesday, April 7, 2010 - link

    This is the worst review that I saw on Anandtech. This looks like a paid Apple advertisement. Comparing a netbook sized device with a mobile phone sized device and being “surprised” that the netbook sized device is so fast.
    This is like a joke.
  • TheHolyLancer - Wednesday, April 7, 2010 - link

    don't know if this was said or not

    why not compare the perf to a netbook, or a culv notebook?

    this may well be a phone derivative, but to me it is at least partly competing in the netbook/culv segment.
  • hrishivb - Wednesday, April 7, 2010 - link

    Hi,
    I did a simple maths to compare the A4 vs iPhone 3gs CPU. Theoretically, if the iPhone CPU could run at 1Ghz speed, then the performance difference between the A4 and iPhone 3gs CPU according to the test performed by Anandtech should be negligible.
    For example the WebKit SunSpider JavaScript Benchmark numbers for A4 and iPhone 3gs CPU are 10475 and 17360 respectively. If we were to run the iPhone CPU at 1 Ghz, then the difference is negligible and for iPhone CPU at 1Ghz it comes around- 10416 (actually better than the A4) !!! Ofcourse I understand that the increase in the performance of the CPU may not be linear to the increase in its speed.
  • Kidster3001 - Wednesday, April 7, 2010 - link

    The reason it's compared to smart-phones is because it has the same hardware as smart-phones. It's not a laptop or a netbook. It's a great big iPod. Why would you compare it to anything else?
  • Zoomer - Thursday, April 8, 2010 - link

    I'm suspecting that the A4 is nothing but the Hummingbird. Basically, a A8 ported to domino logic (from CMOS).

    Note that only Samsung can make Hummingbird. :)
  • jconan - Saturday, April 10, 2010 - link

    When will Anand do an A4 vs Tegra2 for comparison?
  • douglaswilliams - Monday, April 12, 2010 - link

    No kidding!

    Anand seems to think that the A4 might make it into the iPhone 4G (and it might), but i would much rather see the Tegra 2 in it.

    Based on Anand's expressed excitement for the Cortex A9 in his NVIDIA Tegra 2 preview on January 7th, I would have thought he would have mentioned it in this article - at least as a one-liner.
  • xenios - Tuesday, April 13, 2010 - link

    I have a Nokia N900 overclocked to 900Mhz (same Cortex-A8 as in iPad`s 1 GHz one), and running the completely unoptimized Linux Chromium I score around 11000 ms in Sunspider, so I am really not impressed
  • nexusforce - Monday, June 7, 2010 - link

    I own a nexus one updated to froyo 2.2 and it is significantly faster than the iphone 3gs after I compared it with my old 3gs and scored better marks on sunspider benchmark than the ipad in this article. The comparisons and benchmarks should all be redone and this article updated, to show a more contemporary state of performance between the snapdragon SOC and apple's A4.
  • gilbs72 - Wednesday, September 7, 2011 - link

    yeah it's difficult to rely on such a test. even among android phones using the same chipsets, there are noticeable variations. of course, apple only has iphone4 so it's easier to to see how the a4 performs. if indeed the a4 is a just an higher-clocked cortex-a8, aren't there cortex-a8 androids running at 1ghz to compare with snapdragons?

Log in

Don't have an account? Sign up now