Tablets

Tegra 3 gets proper comparison to Apple’s new A5X chip, not a one-horse race

54

Independent benchmark testing is lending a bit of clarity to Apple’s claim that the new A5X chip found in the latest iPad model far exceeds the performance of NVIDIA’s quad-core Tegra 3 processor. Laptop Mag pitted the new iPad against the ASUS Transformer Prime and found that the results were’t as one-sided as Apple would have us believe.

Apple boasted that the A5X CPU offers four times the graphical processing power of Tegra 3, and an OpenGL 3D benchmark confirmed that the new chip outperformed NVIDIA’s platform in this area. The new iPad rendered about two times the frames per second and over four times as many texture pixels. In terms of computation power, however, Tegra 3 was far and away the top dog. The quad-core tablet outperformed the iPad in integer, floating point, and memory performance.

The tests did not rely solely on benchmarks, though. The two devices were pitted head-to-head running two graphically intense games, Riptide GP and Shadowgun. Here thing came down to a bit of a draw. The Retina display of the iPad went a long in way in boosting the appearance of on-screen graphics, but the Tegra 3 versions of the games featured extra graphical flourishes such as splashing water and billowing flags. It isn’t clear if this can be attributed solely to the Tegra 3 chipset or also to in-game optimizations made specifically for the platform.

While the comparison should only serve to spawn new arguments, it is clear that both processing platforms hold their advantages. We’d still like to see Apple’s official test results (and NVIDIA does, too), but this is the best we’ll get for now. Picking the obvious winner isn’t so obvious after all.

Kevin Krause
Pretty soon you'll know a lot about Kevin because his biography will actually be filled in!

New Motorola phone pictured, said to be Verizon-bound Droid Fighter

Previous article

HTC Droid Incredible 4G get benchmarked (plus a mystery device)

Next article

You may also like

54 Comments

  1. So what Apple claimed (the GPU being that much better than the Tegra 3 GPU) was correct…

    1. I thought they said graphic performance, in which case you are correct.

    2. Geez you didn’t even read i guess.. 

      In some functions yes, retina, FPS, but in terms of additional graphics (splashing water and billowing flags), or top dog in computation power (will really show sooner since multi-core apps are increasing).

      So it’s just quits so far.. unless your really a fanboy, I can’t do anything about it.

      1. These tests don’t prove anything about Tegra 3 being having better performance with extra visual bells and whistles such as physics processing and advanced shaders; it’s simply that the games used were not comparable across the two platforms and therefore said test was meaningless.

        Do you have any evidence to back up your claim that multi-core apps are increasing?  Practically every app is already capable of taking advantage of at least 2 cores since the UI and networking/processing will run in different threads if you want a reasonable user experience.

        1. Try TegraZone for starters, and i think it’s not impossible for developers to maximize more processors if their into hardcore apps right?

          “Tegra 3 being having better performance with extra visual bells and whistles”

          – yes. No tools to measure yet. So what other hardware again can produce those kinds of bells and whistles? 

          “Practically every app is already capable of taking advantage of at least 2 cores since the UI and networking/processing will run in different threads if you want a reasonable user experience”

          – Of course. So I assume your saying that in games, you’ll go for A5x and for everything else, Tegra. No?

          1. No, what I’m saying is:

            1) you are making conclusions based on assumption rather than documented evidence.  Do you know for sure that games advertised in the Tegra Zone app are really taking full advantage of the multi-core CPU and not just the enhanced GPU?
            2) it’s early days for games to take full advantage of the A5x’s improved GPU so you cannot imply lesser capability simply due to the non-existence of advanced graphical features in iPad games.  It took several months for games like Riptide and Shadowgun to start taking better advantage of the Tegra 3’s GPU and even now it’s still quite a spartan collection.

      2. bingo. people won’t let reality get in the way of them believing that apple’s chip doesnt’ compare to the tegra.

        Reality distortion field, goooo!

    3. Depending on the benchmark used. As a mobile SOC, the Tegra 3 is more power efficient, smaller, generates less heat, and has a better performance per watt.

      Even the upcoming Transformer Infinity which packs a 1920×1200 screen only uses a 25 Whr battery while the new iPad is packing a big heavy 42.5 Whr feedbag.

      nVidia seems to be better at getting devs to actually use the GPU in their SOC.

      The iPad 2’s A5 was getting better benchmarks for over a year now yet devs haven’t really put the spurs to it.

    4. Measuring FPS performance of a system which draws just one cube with a single color texture on its faces vs a system which draws complex figures and multi-texture won’t necessary show you which one is the most powerful one. In this case, iPad version of those 3D games lack some graphic effects. Saying that iPad has better 3D performance because FPS count was higher than Transformer’s while the iPad didn’t show all those extra effects, it’s just retarded.

    5. hey fanboy..get your glasses off…doesnt show one way or other..apple likes to fab their stories..as shown and heard…

  2. Apple being disingenuous because they know that their idiot users wont fact check them?

    Nawww.. I don’t believe it for a second.

    1. “We get a little water splash on the screen there with the Tegra”. How is this a benchmark?? Isn’t that something that the Game Developers choose to add into the game? SO, once the Game Developer adds a splash of water, then the Ipad wins? What kind of a lame comparison is this? How about frames per second, pixels we see, pixel processing power? This is “test” is ridiculous!

      1. Here’s a real test – not this wavy flag and water splash BS – 
        http://www.youtube.com/watch?feature=player_embedded&v=4jdbtiNnZzE

      2. Did you only read the second half of the article? Did you not see where they benchmarked both devices? Did you not see where it said the iPad outperformed the Transformer Prime in graphical processing? Learn to read before going on your wah spree.
        Your comment is ridiculous! 

  3. So what Apple claimed was true… Don’t get the point of this article… The tests show the Apple chip is better at graphics processing which was it’s claim. Pointless article.

    1. Actually the tests show nothing. Literally nothing. 

      The difference in color are likely due to the screen, not due to the chip, and the “extra effects” in the games on the Tegra 3 are due to developers building in extra graphics for the more powerful chip (since the previous iPad and other iOS devices don’t have chips powerful enough to show those effects without a performance loss.)

      The article and the video basically prove nothing. All they prove is that more testing must be done to decide what the real winner is. When/if they add in the extra effects for the new iPad, then, and only then, can you truly compare the two chips using those games. And even then, a benchmark that is designed to specifically target the chip, and not rely at all on the device that the chip is in, would be the only true test.

      I am a firm believer in the Tegra 3 (and also hate Apple with a passion,) but this video does not prove that it is either better or worse than the A5X chip in the new iPad.

    2. selective reading mate? smoke effect only vs actual splashes and details? still better for you? 

      they just have their pros and cons.. no one outclasses the other..

  4. The Bottom line is that both Riptide GP and Shadowgun THD are Tegra optimized. They have more details and some gimmicks that are not available for the non Tegra versions of these games.
    But this is hardly a comparison of the SoCs.
    The question would be: If you would manage to run the THD version with that gimmicks on the new iPad, would it run as smooth as on the Tegra 3?
    I believe that this would easily be possible if the version was optimized for the iPad3 SoC.

    TL;DR:
    This is a bad “test” in this video. It does not test the capabilities of the SoCs but compares what you get as a customer as of now.
    But what about future games? Will they run similar on both platforms?

    Another interesting thing is the CPU performance.
    The Tegra 3 utilizes 4 ARM A9 cores, The Apple A5X utilizes 2 ARM A9 cores.
    The architecture is not quite the same, for example memory interfaces are not the same, but nevertheless the Tegra should be much more powerful here.
    This does not matter for the iPad, since iOS runs apps suited for this dual core and weaker CPUs – period.

    The CPU in an Android Tablet is much more important. With the advent of Ubuntu for Android real computational power will be very welcomed. There are plenty of Programs written for the power of desktop CPUs which by far outperform any mobile CPU. Canonical aims at Tegra 3 as a SoC to work with for their Ubuntu for Android project. Interesting comparison will be between Tegra 2 and Quallcoms S4 dual core Krait which uses 2 heavily modified ARM A15 cores.

    Full desktop experience on a tablet or on a Phone connected to Monitor/TV, Keyboard and Mouse.
    Imagine walking into your workplace with your Phone in your pocket. You sit down at your desk with a Monitor, Keyboard and mouse but no computer.

    You start one App on the Phone and it connects to the Monitor via DLNA, to the Keyboard and mouse via Bluetooth and starts Ubuntu. You go home and have the same files, settings, browser history, all the things on your Phone.

    Imagine these Monitors, and controllers in public places like libraries. You approach a Terminal and have your personal desktop with you. The next person has theirs.

    This is a thing which is not available for Apple users. And won’t be, since iOS does not use a Linux kernel. They will come up with a similar thing at some point and will claim innovation though ;)

    1. What has been stated in the above article is actually misleading. If you read the original source of this, they point out that the apps tested were optimized for the iPad2 hardware and NOT the new chipset. The Tegra 3 apps are optimized for the Tegra 3. So it really isn’t a good demonstration.

      1. And to that point, if you run an app on your PC outside of it’s native resolution and without appropriate driver useage, you lose color/graphics/performance all day long.

    2. Your Phone-desktop vision isn’t very far off. Look up the Asus Padfone. We’re going to see the phone become the core of a user’s entire computing experience.

      Apple’s Software ecosystem would turn into a colossal clusterf$#k if they tried to shift to a hybrid device. Most iOS devs want users to buy different versions of apps for each device. Apple would have to force devs to consolidate their apps and that would take time and cause a massive riot when it came to changes in their pricing structure. Consumers would get confused. Backwards compatibility for non-hybrid iOS devices would be a disaster.

      Android is better structured for that.

      Hell, even Microsoft is better structured for that. 

      1. Here’s a real test – not this wavy flag and water splash BS – 
        http://www.youtube.com/watch?feature=player_embedded&v=4jdbtiNnZzE

    3.  I enjoyed reading this, but man that TL;DR is long as hell. haha

    4. Apple’s innovations actually work.

        1. Great people, great skills, great wealth, a great brand, great opportunities and they still fail at innovating. 

          http://blogs.hbr.org/schrage/2011/07/what-googles-quiet-failure-say.html

          1. I think something is wrong when in a race where you started first, someone else passes you and you say he would still fail.
            Talking about Androids Marketshare compared to iPhone here.
            Well in that link of yours an Analyst is talking about the chances for Google+ to succeed. He even admits that g+ does many things better and facebook has had to change lots of things.

            Google does a lot of risky start ups – Google Wave for example. They support good ideas. Some of them become big. Like Android.
            Android offers so much more than iOS.
            Among the plethora of manufacturers designing devices running Android are some of the most innovative companies. Sony invented the CD, the MiniDisc, the Walkman. ASUS invented the Netbook (with their first EeePC), the Eee Pad Transformer, the Padfone.
            There is much more to come.

            What was Apples innovation?
            The iPod? nope. There have been mp3 players around 1998 – the first one even 1995 (SanDisk/Siemens). What Apple did was good Marketing.
            The iPhone? Well one could argue here. There have been Smartphones with capacitive touch displays and without buttons before (LG Prada). What Apple did was to introduce some clever ideas to it – multi touch gestures, auto rotate – and again, good Marketing. One could call that innovative.

            The iPad? Nope. There have been tablets before. The market for Tablets was very small. Apple made people want to have them – again good Marketing but hardly innovation.

            I don’t say Apple is crap. They are not. I don’t like many things about the way they handle things. But they make good products.
            It’s just that I’d take an Android over an iPhone any day. And for Design and quality I’d go with ASUS over Apple any day.

          2. From your post, you would label Apple as a run-of-the-mill tech company with no imagination, who simply copied and never created anything innovative. The iPad, iPod, iPhone and Mac mean nothing to the tech industry.  Any tech company could have done it! Well, why didn’t they? 

            Why didn’t Sony create the iPod? Why didn’t IBM create the Mac? Why didn’t Microsoft create the iPad? Why didn’t Motorola or Nokia create the iPhone? After all, they had smart capable engineers, resources, access to the same technology as Apple and some were there before Apple. Why didn’t ASUS create a MacBook Air instead of the failed Netbook? Which company is likely to come out with the next big thing that everyone will copy? Apple. 

            Companies can throw a bunch of products at consumers, but only the ones that are really good survive and drive the industry. Like Apple products. Surely, Google can learn from its past failures, but has it? 

            On the issue of market share, it must be frustrating to Google and its partners to have such a small share of mobile profits. Apple has a 5% share of the world wide mobile phone market, but take in over 75% of the world wide mobile profits. That figure is growing. After a while the field of Android manufacturers will get smaller and I could see iOS eventually taking a larger share of the market.

          3. The way the company is built up – with the late Steve Jobs ruling it’s company as a aristocrat would. Combine that with really good marketing and there is your success formula.

            In Europe we call this the BMW effect. Some people are so focussed on BMW that everything the company designs or produces is an immediate must have. Even though the competitors have a cheaper and/or better equivalent.

          4. @Darktanone:disqus Sony created the iPod before iPod was cool. And they called it Walkman.

            Without the Walkman – portable music players would not exist. Not even the iPod.

            LG brought the candybar, capacitive touchscreen, no button smartphone which could run J2ME midlets (Java apps) Prada phone to the market before the iPhone exited conceptual phase.

            Microsoft and Samsung created the iPad before the iPad was even on the backside of a beer coaster. It was called the UMPC – but because it’s UI was pretty much unchanged with the original Windows it failed.

            If you notice carefully Apple basically waited until their predecessors failed. Noted every thing they failed at – improved on that and copied the rest. Added some of their own idea’s (innovation) et voila – Apple’s new success story.

            I’m not saying Apple is a bad company – in contrary. But it’s not the innovation miracle everyone expects it to be. I guess it takes a lot years of marketing lectures to see how they really work.

      1.  I can see how the new iPad works – 47°C on the Backside – and we don’t even have summer yet. Yellow displays, WiFi and 3G Problems – fantastic product.
        “It just works!”

        1. Just like with the iPhone 4, these issues will work themselves out. User satisfaction results show iDevices dominate and well ahead of the competition. That tells me these early production problems become non-issues in the long run. 3 million New iPads sold in three days…That’s SOLD to consumers and not SHIPPED to resellers to be sold to consumers.

    5. Here’s a real test – not this wavy flag and water splash BS – 
      http://www.youtube.com/watch?feature=player_embedded&v=4jdbtiNnZzE

      1. Thanks for posting the same thing multiple times. Probably would not have found this otherwise. Thanks for not reading what others have written. Geez you must be a frustrated being, coming here, spamming hysterically and probably hyperventilating while at it.

    6. first time seeing a TL;DR bigger than the opening comment

  5. The bottom line is it really doesn’t matter.   Apple could say Android on a new Galaxy S III runs about as fast as the old Newton and fanbois and media will take it as gold.  In the eyes of the world, Apple can do no wrong. 
      It’s not like Nvidia is going to produce a series of commercials showing how much better Tegra3 is compared to Apple’s “new” ipad.  

    The point is, the damage is done.  Any noise from the Nvidia camp is perceived as just that.. noise. I still think that’s where Apple has an advantage some times.  They have a unifying voice for their products.   Android and PCs have Microsoft, Nvidia, AMD, Dell, Asus, HP and whomever trying to promote their own agenda and stepping on each other’s toes in the process.I hate Apple as much, or more than the next person here (except for Julian, who teet-leeches off apple), but I’m just laying out an opinion why things are the way they are.

    1. amen brother..apple guys who think they know what they are talking about is so funny though.

  6. fail reply, ignore this one

  7. Typo: “Here thing came down to a bit of a draw.”

  8. Its not a proper test for one reason.. The transformer prime was running the THD version and so was doing much more processing then the new iPad. How would the new iPad fair if it were running the THD version? Perhaps not a well since the guys showing off the comparison were saying they were pretty equal minus some color differences and that the Tegra 3 had more eye candy.

  9. If you go to the laptop mag website and read the source article you get to see the actual benchmark results. From what I gather at them, the A5X bests the Tegra 3 in everything GPU based, by a factor of approximately 2 to over 4 depending on what is being observed. However the Tegra 3, having 2x as many Cortex A9 cores manages to always win in CPU based benchmarks, except on one subset of a test which the article doesn’t go into detail on.

    However I have to wonder, if in certain tests like the Egypt benchmark, both benchmarks run at native resolution. If they do then we have some overall GPU benchmark that agrees with apple. The egypt test shows apple getting 6718 frames processed, to tegra 3’s 5,939. However if we calculate the difference in resolution, apple has to push abour 3.0x as many pixels. So the a5x gpu is getting 13% better benchmarks at 3.0x the resolution, thus being about 3.475x more powerful in regards to its gpu compared to tegra. I cannot be sure if this is the case for the egypt benchmark, but if it is, then apple wasn’t off by a significant amount in their claim. However if egypt has a maximum resolution of 720p or so, then it is just scaled to both displays and apple would be wrong to claim 4x performance.

  10. Overall the Tegra 3 is still the top dog overall.  Graphics performance while slightly better on the new iPad doesn’t trounce the Tegra 3 as claimed.  In real world performance and use the Tegra 3 shines.  Once apps are written to take real advantage of multi core units like the Tegra 3 then you will see an even larger gap in performance.  Wondering what the performance difference would be had the games on the new iPad been “optimized” for its’ new display?  Granted it may have looked slightly better but the performance would probably be lacking due to the increased resources needed.

  11. Still would prefer t3 since nvidia has many partnerships with game makers as we can see, better effects on t3 = better experience in games. If one doesn’t play game t3 has better computational power. Sorry Apple you lost again.

  12. If you notice apple never gives specs in commercials or anywhere they just say the new iPad is amazing the scree. Looks so good you can do anything…. apple sheep buy it and buy the products I’m not saying it sucks it has great screen but android usually states the specs and we make our choice based on these specs in the end android usually comes out with better screens processors etc and apple fan bois never admit it

  13. Just give us tegra4 already, ipad has a better gpu but same cpu as ipad 2 last year, tegra 3 has a better cpu than a5x.

    Ipad needs the extra perfomance with the bigger screen resolution.

    I cant wait for android to start popping out higher resolution screens as well!

  14. this test is total BS marketing crap!! Nvidia or whomever promoted this should be embarrassed of themselves..

    So what the new Apple Ipad has a better GPU.. benchmark proves it.. Does that mean I am going to run out and return my Prime (hell no)

  15. I have an idea for a device!!! An iPad 3 on one side and a Transformer Prime on the back. PERFECT!!!

  16. To me, this just came down as Apple Fanboyism.

    Quote #1: “Now if you look at the Tegra 3, the smoke off the fire is just a little more intense than the iPad. But the text… the TEXT is a little bit better on the iPad!”

    Quote #2: You see there on your right, the Tegra has a water splash affect when you hit the water, but this is non existent on the iPad. But again, JUST LOOK AT THE TEXT ON THE IPAD!”

    Obviously not his exact words, but I’m sitting here gob-smacked watching this video.

    1. subliminal stuff there

  17. Omg haha the text…. But the text….. But the text… LOL

  18. the a5x is the EXACT same as the a5 cpu wise the x is for the gpu which was changed which is what was compared to the tegra 2. Of course a quadcore would beat a dual core in a cpu test thats simple common sense

    its completely ignorant and stupid to not believe the ipad 3 which features the SAME GPU as the PS VITA would not destroy the Tegra 3 graphically

    cpu is the same as the ipad 2

  19. I wonder if you can see the marks on the jet ski when I run into the wall? If not then that is a dealbreaker

  20. Although still dual core, the A5X pushes graphics cores further than the A5, while the Tegra is an actual quad core. I doubt there will many conclusive tests or benchmarks to determine which chip is better. Also, keep in mind that the A5X on the new iPad is certainly pushing far more pixels out than any other tablet processor out there. Resolution is one of the bigger factors in performance.

Leave a reply

Your email address will not be published. Required fields are marked *

More in Tablets