so the conclusion looks like nexus 9 cpu is weaker than ipad air 2 but gpu is equal on the ipad air 2. battery life cannot be compared to ipad air 2 since anandtech still did not review it.
Yep, but keep in mind Apple is leveraging 20nm and 1Bn extra transistors on their A8X, which almost perfectly translates into 50% more CPU (1 extra core from 2 in A8) and 50% more GPU (2 extra Rogue GPU clusters from 4 in A8). So the disparity in CPU perf is easy to explain in any case.
I think the Denver K1 did better than expected given early benchmarks had it losing to A8X virtually across the board (except for single-threaded).
Also, you can compare battery life because this Preview is basically an iPad Air 2 preview too as it has all the iPad 2 Air data points including battery life. That's probably the most impressive aspect actually, Google/Nvidia was able to beat iPad Air 2 battery life in a smaller form factor.
According to ZD the iPad Air 2's battery is 27.62Wh (actually smaller than iPad Air) while the Nexus 9's battery is 25.46Wh, so yeah while the Air has the bigger screen, it also has the bigger battery and benefit of 20nm on its SoC.
I guess it is pretty safe at this point to say, all the rumors about Tegra K1 not being able to do well in a thin tablet form factor while retaining its performance were wholly unfounded. Certainly bodes well for Erista with 20nm and Maxwell next year!
It's weird to say it's a smaller form factor to claim the battery life is therefore impressive, when the display size difference is double that of the battery capacity difference.
"This build of Android clearly has AArch64 active, which means that we should be able to directly compare the Nexus 9 to the iPad Air 2 for performance." — Um, no. System is using AArch64, but apps aren't.
Native third party applications won't be, but the included applications should be for the most part and so will the applications purely run on ART bytecode. Granted, benchmarks will be the most likely to be native and still AARCH32, but some of the tests will be quite fair.
Preinstalled Google Apps maybe AArch64, but all third-party applications are AArch32. ART bytecode doesn't make them 64 bit. BTW, there is no Dalvik in Android 5, so all apps are purely run on ART.
ART bytecode most certainly IS optimized for 64-bit assuming they didn't throw in native methods or odd languages. "Apps built in Java will automatically gain these benefits, with no changes to existing code. Apps built on other languages, built with the Android NDK r10b [0], can compile for 64-bit architectures to access the features listed above." https://plus.google.com/u/0/+AndroidDevelopers/pos...
ART compiles java bytecode to the appropriate native bytecode: it should generate AArch64 instructions if the hardware supports it. The java bytecode is platform-independent so it doesn't care about AArch64, so purely Java apps don't need to do anything special to be 64-bit, just be run by a 64-bit interpreter.
And while there is no Dalvik, there is still native code: these are the apps which don't "run purely on ART bytecode". This native code has to be recompiled by the developers to include AArch64 insruction set.
Well yeah, but how many Apple apps are 64-bit? How many Windows apps?
Apple is going to have the short-term advantage in just a few months, though, as they are requiring their devs to publish 64-bit versions of apps when they update them after February...
I think the timeframe Apple is giving is a little aggressive, but commendable. Google should consider doing the same thing, but because of the way Play Store works (no screening) I doubt they will.
How so? That's exactly the point, Apple had the opportunity to go with a significantly bigger battery with their larger screen size but didn't, in fact the battery on the Air 2 is smaller than the Air 1. We also know for a fact it is harder to get bigger batteries into devices the smaller they get, and large tablets can scale battery much higher due to size/surface area.
Since the screen area difference is significantly larger than the battery size difference and both tablets give the same battery life, one can hypothesize that the A8X is more energy efficient than the Tegra k1.
Again, can't make any such assumptions based on screen size alone, given early reviews show the Nexus 9 screen is brighter. Apple could've certainly put in a bigger battery or at the very least left the same size as the iPad Air 1, but I guess they wanted to trim weight and cut costs above all else.
Nexus 9 has higher single-threaded CPU performance than iPad air 2 in Geekbench 3, and has higher performance in Kraken too (and possibly some other tests such as Octane). It depends on the workload and where the CPU can take advantage of dynamic code optimization.
based on some broken down geekbench single core, the Nvidia chip has about the same IPC in integer (~2800/2.2 Divided by ~1800/1.4 == ~99%) but, it has a huge deficit in FP (~1200/2.2 divided by ~1700/1.4) of only about 45%.
Numbers ran in 32 bit for K1 as the 64b Android version that supports the K1 appears to not be functioning as it should. NVidia's numbers should improve a bit, probably to 100-115% of Int and 50-65% of Floating Point.
(The output is Tegra K1 Denver core having 99% Integer and 45% based on multiple-core scores PER CORE of Apple's tri-core A8 chip (~5200/2300 used for Tegra, ~5100/4900 used for tri-core A8)
I am pretty sure that the author is referencing the build of the software whenever he reminds us that this is not a final build. The hardware probebly is final, but the software is still on a testing build. Each compilation of a software packet gives
... Sorry hit submit by accident. So I was saying: Each compilation gives you a build which is usually given a build number. In Android you can check your exact build number in the about phone menu in the last menu entry. I bet you can di the same in iOS
looking at some numbers broken down in other places, the Denver CPU appears to be about the same IPC with integer code (~99.5% in the only numbers I have seen), and, about 69.1% of the IPC in Floating point code.
That is per core. Also, I would say that the GPU in the K1 is faster than Apple's as Nvidia's top bin is 951Mhz according to Nvidia. Apple's is likely lower =] (the performance as a given clockspeed should be equal)
Very impressive performance from Nvidia's Denver K1, their 1st 64-bit effort basically ties Apple's 2nd-gen 64-bit SoC without even the benefit of 20nm. Expecting big things from Erista next year with 20nm and Maxwell GPU.
Overall this device I think missed on the lack of expandable storage and low entry level 16GB storage. Personally I think it is priced too close to iPad to do well, but we shall see. If they went with 32GB or even expandable storage I think it would have been a compelling alternative on the higher-end of tablet market.
The question is simple, at what power denver matches the A8X ? If we take the K1 GPU story, we see that Apple has a "bit" lower target for TDP at which the SoC starts to throttle.
Even my two year old laptop is faster than A8X but I didn't saw anyone being interested in that.
Is your two year old laptop 6-8mm deep and last ~10 hours on 1 battery charge? Then why bother mentioning it.
We see in the review, Nexus 9 actually has slightly better battery life than A8X, so I'd say it matches A8X now at current TDP which is estimated on all these SoCs to be about 4.5W. With the benefit of 20nm it would undoubtedly beat A8X either via more transistors (as Apple did) or higher clockspeeds.
Yeah, what he's saying is that the raw performance doesn't matter if it's not taken in context with the size and battery life and such. He's arguing the same thing as you.
Uh, no except he was comparing two devices in the same class with similar form factors with a 3 year old laptop that is probably 5x thicker with 1/5th the battery life. If he was trying to draw parallels he did an exceptionally poor job at doing so. How is comparing the performance of the Nexus 9 to iPad Air 2 SoCs the same as trying to draw a similar comparison with a 3 year old laptop? Everything that makes the tablet interesting goes out the window on something in that huge clunky form factor.
He was clearly exaggerating to make a point. Performance doesn't everything by itself, especially in situations like this. The A8X is a good chip, and it's more efficient than Denver. Give Denver a die shrink and the tables should turn.
He made a pointless point I guess then, because the A8X and K1 are very close in terms of performance and efficiency to the point I'm not sure anyone could make an accurate assessment based on what we know.
Air 2 has a bigger screen, but the N9's screen is brighter. Air 2 has a bigger battery, and the N9 has better battery life.
So again, how exactly do you know the A8X is more efficient than Denver?
Just look at manufacturing norms. Wonders belong to magic farytales, hard reality tells that 20 nm is much more energy efficient than 28 nm.
As result, on heavy load NVidia K1 Denver will fall dramatically, if the load is prolonged. There is no other way, because otherwise battery life under load would be drastically lower than on iPad Air 2.
The N9 has essentially the same battery life as the larger screened (larger power draw), smaller battery ipad. The ipad is also using much lower clocks on their cpu, and their gpu looks to be at least a match to the tk1. Apple is on 20nm, and that's a big deal, but performance doesn't change things significently.
Pretty dissapointed with this nvidia effort. Hoping from more with the upcoming cortex a57.
Pardon, that should've been relatively (as a linear function of screen area it comes to 176 mAh/in2 vs 163 for the ipad). As for the N4, well, I'd simply forgotten about it. However, we still don't have benchmarks for it in 64bit mode. As it is, it seems to be a bit slower than the a15 which shouldn't be the case.
But the N9 has the brighter screen and a smaller battery, and yet N9 has slightly battery life. Not sure how you can be disappointed in this Nvidia effort when they have matched Apple's performance without needing to leverage 20nm and they have in most cases, doubled performance relative to the next closest Android ARM-based competitor and left Intel's Atom based offerings in the rear view as well. Cherry Trail might have proved an interesting part here, but its been delayed, so right now its clear there are really only two choices in the tablet SoC market for pure performance: A8X and Denver K1.
because the CPU and GPU use all the power in battery life tests, right? The brightness is not equalized, wifi doesn't draw power. Just the SoC.
Oh, wait, the screens are standardized to 200nits brightness. And the bightlight is added to power, the wifi is adding to power, the digitizer is adding to power... Wifi tests using are not demanding, so, the SoC is more likely in a low power state, which, makes everything else besides the backlight take a lot more in terms of percentage than normally they would (backlight is still the largest power usage after SoC iirc, with a notable gap between those two and everything else)
Could it be possible that when you equalize the screen brightness (should be able equal power draw) that perhaps a larger screen matters more? Everything else should be about equal.
Low power usage for low power states of the SoCs, wifi radios about the same power, except, the iPad screen is ~29000mm^2, The Nexus 9 is about ~25500mm^2 on its screen.
Do you think that having about a 13% larger area to backlight would reasonably give under 1% less battery life given the battery is only 7% larger?
Apple's chip is likely more effective. The margin by which it is more effective is hard to tell without knowing what percentage of the power draw the tablets each had from the SoC. I would guess at least 15-20%. Which, would mean WORST CASE for Apple, Nvidia's chip's power consumption would be equal at the same node (more or less).
On that note, I think that a K1 A15r3 would likely be faster absolutely as a SoC. Granted, most things don't use 4 threads =]
The brighter screen doesn't matter as the test is done at the same brightness, and the runtimes differed by less than a minute. 20nm doesn't make the performance any better unless you actually clock your cpus higher. Apple, iirc, only moved their clocks up 100MHz, to 1.4, while denver is running at 2.3. I'm dissapointed b/c it looks like denver CAN be fast, but only if running one or two apps for long periods (filling the cache with that vliw optimized code). When multitasking it apparently has serious issues. Hopefully these can be addressed in future revisions to denver. I'm not comparing denver to atom or other current android offerings as none of them have attempted to go use a radically different arch from everyone else.
"But the N9 has the brighter screen and a smaller battery, and yet N9 has slightly battery life" -- battery life is measured with equal brightness, so K1's longer battery life is result of smaller screen and dramatical throttling slowdown which is unavoidable since it is 28 nm process (versus 20 nm in A8X).
in a web browsing test? The CPU isn't high enough to throttle... Or, if it is the webpage is more intensive than an 'average' mobile web page, OR, the SoC is terribly designed.
While I am very down on Nvidia's mobile efforts due to peak power draws and throttling, it shouldn't be throttling anymore than the A8X in sane web browsing tests...
We were being told that ARM could never compete with x86, but the A8x is stronger than most i3 low power chips. I assume this new Tegra is somewhere around that as well. The fact that it does compare with a three year old i5 notebook is something that I find incredible.
Considering that Apple and possibly others, will continue to upgrade performance in their ARM offerings by substantially greater amounts than Intel is able to with x86, for at least the next few years, means that Intel really does have something to worry about.
ipad has some pretty serious music making apps. Both have some cad programs from autodesk. Lastly, browsers themselves can actually use that power. For instance, firefox via webgl + asm.js. This is ignoring the coming push into servers.
The A8X is ridiculously more impressive than any Core chip because of giving comparable processing power at about 30% of power, and probably far less than 50% of price.
And here's the thing - Apple AX chips are a prime candidate for future MacBook successors.
"The A8X is ridiculously more impressive than any Core chip because of giving comparable processing power at about 30% of power, and probably far less than 50% of price."
You heard of a thing called Core M? Same TDP as A8x, single core geekbench 33% higher. GPU performance 2.5x? And on silicon that is smaller than the A8 let alone the A8X, so cost of manufacture is cheaper.
Core M is definitely in a class of its own in terms of performance, but I agree its not applicable in this discussion because of its price tag. We won't see it in a sub $600 device competing with these parts.
Please back that claim up. From what we have seen, Core M in the Yoga 3 Pro (high end core M variant!) fails to match the Core i5 4200U, and .. Apple A8X in fact matches it in raw performance!
My bet is that Apple A9X on 14nm should allow the MacBook Air succesor to outperform windows competition in every way, and actually cost less.
Oh and also, Core M doesn't have the same TDP as A8X, not even close!
The A8X has, according to battery endurance measurements, has lower TDP than the A7..you know, the chip that runs in the iPad Air and..iphone 5S. Think about that for a moment.
The only reason why A8X isn't in new iPhones is because the A8 is sufficient for iPhones while A8X(along with the 2GB RAM) turns the iPad into a bit more than the big iPhone.
It's not about power, it's about how you can use that power. On Windows there are many ways to use power of any Core chip, on iOS there is pretty much no way you can use power of A8X.
@kron, you wrote "But pretty much no way you can use power of A8X", which directly wrong. By now there are already few dozens of heavy load apps that load things (many graphics-related, medical processing pro apps, and so on).
"Overall this device I think missed on the lack of expandable storage and low entry level 16GB storage. "
Yup. This one boggles the mind. They did 32gb on the Nexus 6. WTF is with 16gb as a start point for a high end tablet? Other than that, its a pretty darn good tablet though. Amazing GPU.
Denver looks like some good design choices. 2 big cores are better than 4 scrawny ones, glad to see Android SoC makers learn that, ironically it's the same one that sent everyone chasing 4 cores. The VLIW+Code morphing architecture is interesting, in thoery it would take a hit from spaghetti code, but with 128MB of far caching it shouldn't really run into situations where it's reoptimizing the same code a lot.
Maybe I spoke too soon, seeing the benchmarks, it seems to ping between great and average. Joshua, think what I said is coming into effect with spaghetti code?
The final firmware builds were supposedly pushed Sunday afternoon via OTA. Can't comment on what other sites do but our review will be on that firmware, and understandably it's impossible to do a review within 24 hours.
So go read those reviews then? Wait, you still came here to complain about anandtech's review not being up?
I'm going to guess it's the same reason I came here to see if anandtech's review was up. Because I wanted a more in-depth review. Depth takes time obviously so I guess we'll have to wait.
True, some reviews from other sites contains even less information than this preview. If Anandtech posts a review too quickly without fully testing the hardware, a lot more people will complain.
I guess it also really depends what you consider a "full review". Personally I think AnandTech is the only site that actually tries to do a comprehensive performance review from a true tech enthusiasts point of view.
Now compare with these other consumer tech gadget sites that throw up 3 benchmarks for 3 devices with another 2500 words of subjective filler regarding aesthetics that anyone who walked into a Best Buy could ascertain on their own.
I'll wait for quality analysis. As the article stated the final firmware has only been available for a day, so that should tell you something about the "full reviews" you love so much.
I wish Anandtech would stop focusing so much on Sunspider. Google said it 2-3 years ago that they aren't focusing Chrome on Sunspider performance anymore because it's a very simplistic test without much relation to real-world web app performance. They even did a 50xSunspider test to prove how hopefully obsolete it is, if some of you remember.
The Basemark OS benchmarks are all over the place. In some it tests twice as good as the iPad, in others half as flow. I don't think Basemark OS has its algorithms well optimized, and it may give some subcomponents way more score than they deserve, or way too little in others - or Basemark OS may simply be acting weirdly with this new processor.
Kraken seems to give the same relative performance results with the same ordering of devices, so it doesn't seem to make much difference which one they use.
It's not just benchmarks g hardware. It's benching the OSs and the browser engine too. So it does matter. Besides, as I said, since Kracken scores seem to follow the Sunspider scores in ranking, it doesn't matter than much.
Are we reading the same chart? In Sunspider chart, Nexus 9 is well below iPad Air, whereas in Kraken benchmark, Nexus 9 shows a slightly better score than than iPad Air 2.
Sunspider results are far too easy to optimize for in the browser, but it will differ for different architectures. Krait could for instance not really distance itself from cortex a9 in sunspider for a good while on Android.
What I'd like to know - is the tablet actually taking advantage of the 64-bit/Aarch64 mode? Or is it running in 32-bit mode still? Because I think it was still running in 32-bit mode in early Geekbench tests. I don't know why Google would release it like that, but it would be disappointing if they did, since that could be cutting its CPU performance by 10-20 percent.
Is it really a guarantee that 64-bit performance will be better than 32-bit performance with a VLIW architecture?
The performance is going to be highly dependent on how well the code morpher can pack instructions into a VLIW instruction. So, if the VLIW instruction is 256 bits wide, does that mean it can only pack half the 64-bit instructions?
1) It's not final firmware on Nexus 9. 2) It's pre-alpha (or something) build of 64-bit Geekbench. We'll see when the final version of 64-bit Geekbench will be released.
Gfxbench's results posted here were very generous. Nexus 9's Gfxbench onscreen results has only 2048x1440. The results will be worse when Gfxbench update the software to process 2048x1536 (196608 more pixels). The same happened with Iphone 6
Seems pretty competitive with the Air 2 all around, what about some display calibration details? Those would be greatly apprecited
Now, the real catch will be next year, when we get quad core Denver + a Maxwell GPU on a 20nm process, they were able to match the A8X in pretty much everything but some CPU bound tests with a bigger process
Still, after seeing these results the display calibration and speaker sound quality will determine if I get it or not
When did the benchmarks get run? Shouldn't the final build be available today? Or will it be over the air when the first tablets reach people in the mail?
Wow, the GPU perfomance of the Samsung Galaxy Tab S is utter trash. I didn't know the GPU performance of the Mali-T628 would be THAT bad compared to the K1 and A8X. Thank god I didn't buy one.
Quite often Nexus devices have a close sibling from the same manufacturer but with extra hardware niceties, such as gaining a memory card slot, different or removable battery, USB on the go etc.
So it will be interesting to see if HTC come out with a very similar table to this Nex9 but with a memory car slot and some premium features?
Can you guys do a comparison article between apple a8x, tegra k1 and intel core m ? would love to understand your thoughts on the tablet/hybrid market as a whole and if there is a potential here.
The 20nm 1.9GHz A57 is already out in the Galaxy Note 4. Scores around 1250 in single core Geekbench. But power draw in the octa core setup should be incredibly low considering that the 20nm Exynos 5430(Cortex A15/Galaxy Alpha) doesn't even cross 3W under maximum load!
Also,if the rumors are true that Snapdragon 810 will have it clocked at 2.7GHz, the end result should be quite incredible.
I bought a Nexus 7 2013 from Walmart last year just as it was released. Battery life was about 6 hours of screen time with low brightness and no background syncing, doing mostly web browsing or local video playback. It was nowhere near Anandtech's 12 hours, not even the advertised 9 hours. Ever since then I'm skeptical battery life claims for android devices. This Nexus 9 review sample was probably cherry-picked.
I have had similar experience with android devices, I have owned nexus 7 2013,LG G pad 8.3 and Galaxy pro tab 10.1. Had IPad Air and the battery would last a long time mostly with web browsing, much more than the review sites.
between my 5.7" Note 3 phablet, my 13" ultrabook, and my 27" desktop beast, what exactly do u need a tablet for? All my friends that bought one say it just sits there unused. Seems like a niche market or just a gimmick.
One solution to your question: replace the phablet with a small, basic phone plus a 8" wide screen tablet and you get something that's more convenient for actual phone calls, provides more screen space for everything else, has a lower total cost and can still be carried in your pockets. Not sure about tablets that need bags myself, but lots of other people seem to like them.
I disagree that 8" wide screen tablet is more convenient than a 5.7" palmtop (I happen to own both, palmtop being the Galaxy Note 3). The Note 3 is simply much easier to hold. And the screen is still perfectly big for reading and even movies while laying in bed. Handling Web pages and gaming is also much easier.
While my tablet is perfectly okay for what it is, I find it using it mostly when my Note 3 is charging.
So I tend to agree with the above post. Phablets (or palmtops as I prefer to call them) are the reason for tablet market stagnation, and are rightfully booming right now.
Just look at Apple's reaction if you need further evidence - neutering the iPad mini a release of the iPhone 6 plus.
I think Google's point is that everyone has a slightly different idea of an ideal size. That's why they didn't make a nexus 7 (2014) this time.
I bought an LG G2 last year because I wanted a screen big enough do some actual reading on without the size of a Note. But now that I've got a Shield Tablet, I'm seriously thinking about getting a smaller phone. The G2 isn't bulky in my pocket, it's just a hair too big to be used comfortably as a phone.
Why do sites like anand insist on publishing pre-reviews like this? dpreview is the same for cameras, several first impressions/almost there but not quite complete/final view type things.
Why don't you wait until you have release software? No-one is going to be using the pre-release build used in this review so why review it?
I normally appreciate Anandtech reviews and it's the first place I go to but this sort of thing is just annoying.
Joshua, please read Futuremark's comments on Apple's CPU architecture and why 3DMark Physics (which is a CPU test and perhaps should be in the CPU benchmark section?) performs relatively badly on Cyclone CPU cores: http://www.futuremark.com/pressreleases/understand...
My guess is that Denver has similar limitations - it's exceptionally good at certain types of task (ie the simple tests found in typical benchmark scenarios) but isn't as capable on other tasks. Or maybe drop Futuremark's a line for further information? Based on the in-depth Cyclone analysis I just linked to, I am sure they will be happy to clarify Denver's disappointing performance in their benchmark.
The text suffers from a great deal of awkward and even contradictory wording to the point that it detracts from the content. Hopefully an editor will work harder on the full review.
I have a simple rule. No MicroSD slot means no purchase. There is a fixation shared by most tablet designers that everything is in the cloud and therefore there is no need for a decent amount of storage space.
There is a simple message to the designers - grow up and get out more. There are plenty of parts of the world where internet access is "patchy" even at international hotels or expensive or both. Then there are those parts of the world which have no access - aircraft. I am in the market to replace my 3 year old Nook HD+, Nexus 9 will not be on the shortlist
I'd have loved to have seen how well this compared to Intel's second gen Atom, had the licensing worked out. Presumably the GPU would be better if nothing else...maybe Nvidia could have shipped a Windows version of their Shield products, which would be awesome!
Still tempted by this since there's no Windows version of Marvel Unlimited...
Been playing with mine all morning. I don't really see the performance issues everyone is talking about. My "Recent Screens" button does not take 2-3 seconds to pop up, it's quite responsive, actually. I'm not a huge gamer, so I can't comment there, but just in general use as a media consumption device / large screen mobile browser / productivity device, I couldn't be happier.
As a side note, anyone care to explain to me why Samsung Galaxy Tab S routinely scores worse than the Tab Pro even though they are pushing the same number of pixels? Based on the graphs, Samsung's own Exynos 5420/Mali-T628 combo is inferior to last year's SD800+Adreno 330?
What is with the sunspider regression? It seems to indicate the need for further code optimization. Perhaps try chrome beta and update the firmware. Ars had the newer firmware for their review.
Google said a while back they could basically care less about Sunspider because it's a poor representation of actual web browsing, which, any extended use of Safari vs. Chrome (on a droid) bears out. There is no appreciable difference in the web experience on either device, although I prefer Chrome just for the universal tabs, etc.
AnandTech should start testing interface responsiveness quantitatively. E.g. when I swipe or click, I want the device to react instantaneously - 200 milliseconds is way too long, I need 1 millisecond. A high-speed camera and a piece of software to analyse the video recording should do the trick.
In my opinion, this is the main difference between Apple and Android world. I always feel almost ashamed when I use my Android phone next to someone with an iPhone. My phone seems to always do something important in the background exactly when I need it to do something for me fast.
100% agree with the principle of this test. I've definitely noticed my Note 3 a bit laggy at time. I expect the device to work for my needs and not the other way around. I also like the idea of doing some 'real world' tests with a whole load of apps open, various browser tabs open, then switch tasks, ring the phone etc. and check responsiveness. However, maybe I would have same issue with Apple.
When is it going to happen to have the complete recension of both devices for having a clear idea? I think preliminary results in this article only cause flames between users. Bit disappointed by Anand sincerely.
I purchased one for my daughter which we received on Tuesday, previously she had an iPad 2, three and a half years ago. She's happy with it (it plays games and videos and surfs the internet etc.) and I don't need to install iTunes. So, a win-win situation. HOWEVER, we still cant buy ANY cover for it on Google Play. WTFlip is that about. Everyone here is talking about what SOC and how many cores and blah blah, but they can't even sort out the cover for the device. I had the same issue last year with the 7 inch Nvidea tablet. Launched and we couldn't buy the great cover that was designed for it. So, in this regards Apple wins by a long way. The iPad 2 was engraved on the back for free and it arrived with the cover at the same time the iPad came. By the way, the cover protects the screen and allows you to stand the tablet up etc. to watch movies. Pretty essential for most people I would think... (and yes, I know I could get a generic cover or put it in a paper bag or something but really, I would like the cover designed for the device that gets advertised...)
when is the review coming??? I was going to get a Nexus 9 but now aftet bad reviews am seriously thinking about an ipad mini 2 - waiting for the always great anandtech review before making final decision!
I gave up waiting for the review to decide between the Nexus 9 and the iPad Air 2. I just bought the Nexus 9 a few days ago and I am waiting for it to ship.
I think the full review is delayed so much because Anandtech is currently consulting lawyers fearing nVidia's wrath. The K1 Denver is the biggest benchmark cheater, Anandtech found it out, and....... you know the story between 3DMark and nVidia.
To be clear: the K1 Denver is an *emulator*, NOT a custom designed ARM. It's so easy to write an emulator that cheats primitive, predictable, synthetic benchmarks.
For example, write a function that does hundreds and thousands times memcpy on same addresses. With the -O3 option, the compiler optimizes the function so that it does memcpy only once.
Benchmarks are written in a way that the resulting machine codes actually do all the iterations, but with an emulator that translates these machine codes to a different one, it can easily mimic the compiler's behavior mentioned above. Of course, it would be "optimization" by nVidia's definition. This kind of "optimizations" were done before by Transmeta, and K1 Denver is a reincarnation of Transmeta's Crusoe/Efficeon. And we all know that nVidia .......... well, the rest is up to you guys.
I hope someone will see this and help me. I ran those benchmarks on my Nexus 9 and mine is almost 200% slower, i.e., double sunspider score. Did I get a fake Nexus 9?
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
146 Comments
Back to Article
shashiiishot2007 - Monday, November 3, 2014 - link
so the conclusion looks like nexus 9 cpu is weaker than ipad air 2 but gpu is equal on the ipad air 2. battery life cannot be compared to ipad air 2 since anandtech still did not review it.chizow - Monday, November 3, 2014 - link
Yep, but keep in mind Apple is leveraging 20nm and 1Bn extra transistors on their A8X, which almost perfectly translates into 50% more CPU (1 extra core from 2 in A8) and 50% more GPU (2 extra Rogue GPU clusters from 4 in A8). So the disparity in CPU perf is easy to explain in any case.I think the Denver K1 did better than expected given early benchmarks had it losing to A8X virtually across the board (except for single-threaded).
Also, you can compare battery life because this Preview is basically an iPad Air 2 preview too as it has all the iPad 2 Air data points including battery life. That's probably the most impressive aspect actually, Google/Nvidia was able to beat iPad Air 2 battery life in a smaller form factor.
blackcrayon - Monday, November 3, 2014 - link
Do we know what the relative sizes of the batteries are? Obviously the iPad is larger but is also thinner.chizow - Monday, November 3, 2014 - link
According to ZD the iPad Air 2's battery is 27.62Wh (actually smaller than iPad Air) while the Nexus 9's battery is 25.46Wh, so yeah while the Air has the bigger screen, it also has the bigger battery and benefit of 20nm on its SoC.I guess it is pretty safe at this point to say, all the rumors about Tegra K1 not being able to do well in a thin tablet form factor while retaining its performance were wholly unfounded. Certainly bodes well for Erista with 20nm and Maxwell next year!
solipsism - Monday, November 3, 2014 - link
It's weird to say it's a smaller form factor to claim the battery life is therefore impressive, when the display size difference is double that of the battery capacity difference.kron123456789 - Monday, November 3, 2014 - link
"This build of Android clearly has AArch64 active, which means that we should be able to directly compare the Nexus 9 to the iPad Air 2 for performance." — Um, no. System is using AArch64, but apps aren't.frostyfiredude - Monday, November 3, 2014 - link
Native third party applications won't be, but the included applications should be for the most part and so will the applications purely run on ART bytecode. Granted, benchmarks will be the most likely to be native and still AARCH32, but some of the tests will be quite fair.kron123456789 - Monday, November 3, 2014 - link
Preinstalled Google Apps maybe AArch64, but all third-party applications are AArch32. ART bytecode doesn't make them 64 bit. BTW, there is no Dalvik in Android 5, so all apps are purely run on ART.toyotabedzrock - Monday, November 3, 2014 - link
I think the android Dev blog said art would compile any non native apps to aarch64 automatically.Samus - Monday, November 3, 2014 - link
They ditched Dalvik? wow, there goes like half of the Play Store ;)jhenzie - Tuesday, November 11, 2014 - link
Final compilation has always been on device and ART can consume Dexfrostyfiredude - Monday, November 3, 2014 - link
ART bytecode most certainly IS optimized for 64-bit assuming they didn't throw in native methods or odd languages. "Apps built in Java will automatically gain these benefits, with no changes to existing code. Apps built on other languages, built with the Android NDK r10b [0], can compile for 64-bit architectures to access the features listed above." https://plus.google.com/u/0/+AndroidDevelopers/pos...buttdill - Tuesday, November 4, 2014 - link
ART compiles java bytecode to the appropriate native bytecode: it should generate AArch64 instructions if the hardware supports it. The java bytecode is platform-independent so it doesn't care about AArch64, so purely Java apps don't need to do anything special to be 64-bit, just be run by a 64-bit interpreter.And while there is no Dalvik, there is still native code: these are the apps which don't "run purely on ART bytecode". This native code has to be recompiled by the developers to include AArch64 insruction set.
Zarsus - Wednesday, November 5, 2014 - link
ART will compile any non-native Android app to 64-bit when viable just as Java VM will compile any java code to 64-bit when viable.What makes you think ART doesn't do that when it makes sense anyway?
Samus - Monday, November 3, 2014 - link
Well yeah, but how many Apple apps are 64-bit? How many Windows apps?Apple is going to have the short-term advantage in just a few months, though, as they are requiring their devs to publish 64-bit versions of apps when they update them after February...
I think the timeframe Apple is giving is a little aggressive, but commendable. Google should consider doing the same thing, but because of the way Play Store works (no screening) I doubt they will.
darwinosx - Monday, November 3, 2014 - link
Apple has been 64 bit for a year and ART is still a runtime. Many iOS apps are 64 bit and they are complied for the shot part.kron123456789 - Monday, November 3, 2014 - link
"Well yeah, but how many Apple apps are 64-bit?" — I think almost all of them. All the benchmarks, at least.noelbonner - Tuesday, November 11, 2014 - link
There are a number of tablets that are higher ranked in CONSUMER BASED rankings (like http://is.gd/yWOP3B for example).craighamilton - Saturday, December 6, 2014 - link
Not a big fan of Nexus 9... there are plenty of tablets that have higher user satisfaction(see http://www.topreport.org/tablets/ for example...)
chizow - Monday, November 3, 2014 - link
How so? That's exactly the point, Apple had the opportunity to go with a significantly bigger battery with their larger screen size but didn't, in fact the battery on the Air 2 is smaller than the Air 1. We also know for a fact it is harder to get bigger batteries into devices the smaller they get, and large tablets can scale battery much higher due to size/surface area.Ppietra - Monday, November 3, 2014 - link
Since the screen area difference is significantly larger than the battery size difference and both tablets give the same battery life, one can hypothesize that the A8X is more energy efficient than the Tegra k1.kron123456789 - Monday, November 3, 2014 - link
It's no surprise since A8X is 20nm and Tegra K1 is 28nm.chizow - Tuesday, November 4, 2014 - link
Again, can't make any such assumptions based on screen size alone, given early reviews show the Nexus 9 screen is brighter. Apple could've certainly put in a bigger battery or at the very least left the same size as the iPad Air 1, but I guess they wanted to trim weight and cut costs above all else.ams23 - Monday, November 3, 2014 - link
Nexus 9 has higher single-threaded CPU performance than iPad air 2 in Geekbench 3, and has higher performance in Kraken too (and possibly some other tests such as Octane). It depends on the workload and where the CPU can take advantage of dynamic code optimization.testbug00 - Tuesday, November 4, 2014 - link
based on some broken down geekbench single core, the Nvidia chip has about the same IPC in integer (~2800/2.2 Divided by ~1800/1.4 == ~99%) but, it has a huge deficit in FP (~1200/2.2 divided by ~1700/1.4) of only about 45%.Numbers ran in 32 bit for K1 as the 64b Android version that supports the K1 appears to not be functioning as it should. NVidia's numbers should improve a bit, probably to 100-115% of Int and 50-65% of Floating Point.
(The output is Tegra K1 Denver core having 99% Integer and 45% based on multiple-core scores PER CORE of Apple's tri-core A8 chip (~5200/2300 used for Tegra, ~5100/4900 used for tri-core A8)
Ratman6161 - Monday, November 3, 2014 - link
Also don't forget it's Android 5 so besides being an early build of the tablet, the OS may not be final either.Leosch - Tuesday, November 4, 2014 - link
I am pretty sure that the author is referencing the build of the software whenever he reminds us that this is not a final build. The hardware probebly is final, but the software is still on a testing build. Each compilation of a software packet givesLeosch - Tuesday, November 4, 2014 - link
... Sorry hit submit by accident. So I was saying:Each compilation gives you a build which is usually given a build number. In Android you can check your exact build number in the about phone menu in the last menu entry. I bet you can di the same in iOS
testbug00 - Tuesday, November 4, 2014 - link
looking at some numbers broken down in other places, the Denver CPU appears to be about the same IPC with integer code (~99.5% in the only numbers I have seen), and, about 69.1% of the IPC in Floating point code.That is per core. Also, I would say that the GPU in the K1 is faster than Apple's as Nvidia's top bin is 951Mhz according to Nvidia. Apple's is likely lower =] (the performance as a given clockspeed should be equal)
chizow - Monday, November 3, 2014 - link
Very impressive performance from Nvidia's Denver K1, their 1st 64-bit effort basically ties Apple's 2nd-gen 64-bit SoC without even the benefit of 20nm. Expecting big things from Erista next year with 20nm and Maxwell GPU.Overall this device I think missed on the lack of expandable storage and low entry level 16GB storage. Personally I think it is priced too close to iPad to do well, but we shall see. If they went with 32GB or even expandable storage I think it would have been a compelling alternative on the higher-end of tablet market.
GC2:CS - Monday, November 3, 2014 - link
The question is simple, at what power denver matches the A8X ?If we take the K1 GPU story, we see that Apple has a "bit" lower target for TDP at which the SoC starts to throttle.
Even my two year old laptop is faster than A8X but I didn't saw anyone being interested in that.
chizow - Monday, November 3, 2014 - link
Is your two year old laptop 6-8mm deep and last ~10 hours on 1 battery charge? Then why bother mentioning it.We see in the review, Nexus 9 actually has slightly better battery life than A8X, so I'd say it matches A8X now at current TDP which is estimated on all these SoCs to be about 4.5W. With the benefit of 20nm it would undoubtedly beat A8X either via more transistors (as Apple did) or higher clockspeeds.
mkozakewich - Monday, November 3, 2014 - link
Yeah, what he's saying is that the raw performance doesn't matter if it's not taken in context with the size and battery life and such. He's arguing the same thing as you.chizow - Monday, November 3, 2014 - link
Uh, no except he was comparing two devices in the same class with similar form factors with a 3 year old laptop that is probably 5x thicker with 1/5th the battery life. If he was trying to draw parallels he did an exceptionally poor job at doing so. How is comparing the performance of the Nexus 9 to iPad Air 2 SoCs the same as trying to draw a similar comparison with a 3 year old laptop? Everything that makes the tablet interesting goes out the window on something in that huge clunky form factor.Alexvrb - Monday, November 3, 2014 - link
He was clearly exaggerating to make a point. Performance doesn't everything by itself, especially in situations like this. The A8X is a good chip, and it's more efficient than Denver. Give Denver a die shrink and the tables should turn.chizow - Tuesday, November 4, 2014 - link
He made a pointless point I guess then, because the A8X and K1 are very close in terms of performance and efficiency to the point I'm not sure anyone could make an accurate assessment based on what we know.Air 2 has a bigger screen, but the N9's screen is brighter.
Air 2 has a bigger battery, and the N9 has better battery life.
So again, how exactly do you know the A8X is more efficient than Denver?
DERSS - Wednesday, November 5, 2014 - link
Just look at manufacturing norms. Wonders belong to magic farytales, hard reality tells that 20 nm is much more energy efficient than 28 nm.As result, on heavy load NVidia K1 Denver will fall dramatically, if the load is prolonged. There is no other way, because otherwise battery life under load would be drastically lower than on iPad Air 2.
Alexvrb - Friday, November 7, 2014 - link
So, before you evangelize Nvidia products, do you actually read the articles?"As always, our battery life tests are all run with the display calibrated to 200 nits."
Anandtech has been doing this very same procedure for a long time.
tuxRoller - Tuesday, November 4, 2014 - link
The N9 has essentially the same battery life as the larger screened (larger power draw), smaller battery ipad.The ipad is also using much lower clocks on their cpu, and their gpu looks to be at least a match to the tk1.
Apple is on 20nm, and that's a big deal, but performance doesn't change things significently.
Pretty dissapointed with this nvidia effort.
Hoping from more with the upcoming cortex a57.
kron123456789 - Tuesday, November 4, 2014 - link
Well, iPad's battery is larger. And Cortex A57 is already here - Galaxy Note 4.tuxRoller - Tuesday, November 4, 2014 - link
Pardon, that should've been relatively (as a linear function of screen area it comes to 176 mAh/in2 vs 163 for the ipad).As for the N4, well, I'd simply forgotten about it. However, we still don't have benchmarks for it in 64bit mode. As it is, it seems to be a bit slower than the a15 which shouldn't be the case.
chizow - Tuesday, November 4, 2014 - link
But the N9 has the brighter screen and a smaller battery, and yet N9 has slightly battery life. Not sure how you can be disappointed in this Nvidia effort when they have matched Apple's performance without needing to leverage 20nm and they have in most cases, doubled performance relative to the next closest Android ARM-based competitor and left Intel's Atom based offerings in the rear view as well. Cherry Trail might have proved an interesting part here, but its been delayed, so right now its clear there are really only two choices in the tablet SoC market for pure performance: A8X and Denver K1.testbug00 - Tuesday, November 4, 2014 - link
because the CPU and GPU use all the power in battery life tests, right? The brightness is not equalized, wifi doesn't draw power. Just the SoC.Oh, wait, the screens are standardized to 200nits brightness. And the bightlight is added to power, the wifi is adding to power, the digitizer is adding to power... Wifi tests using are not demanding, so, the SoC is more likely in a low power state, which, makes everything else besides the backlight take a lot more in terms of percentage than normally they would (backlight is still the largest power usage after SoC iirc, with a notable gap between those two and everything else)
Could it be possible that when you equalize the screen brightness (should be able equal power draw) that perhaps a larger screen matters more? Everything else should be about equal.
Low power usage for low power states of the SoCs, wifi radios about the same power, except, the iPad screen is ~29000mm^2, The Nexus 9 is about ~25500mm^2 on its screen.
Do you think that having about a 13% larger area to backlight would reasonably give under 1% less battery life given the battery is only 7% larger?
Apple's chip is likely more effective. The margin by which it is more effective is hard to tell without knowing what percentage of the power draw the tablets each had from the SoC. I would guess at least 15-20%. Which, would mean WORST CASE for Apple, Nvidia's chip's power consumption would be equal at the same node (more or less).
On that note, I think that a K1 A15r3 would likely be faster absolutely as a SoC. Granted, most things don't use 4 threads =]
tuxRoller - Tuesday, November 4, 2014 - link
The brighter screen doesn't matter as the test is done at the same brightness, and the runtimes differed by less than a minute.20nm doesn't make the performance any better unless you actually clock your cpus higher. Apple, iirc, only moved their clocks up 100MHz, to 1.4, while denver is running at 2.3.
I'm dissapointed b/c it looks like denver CAN be fast, but only if running one or two apps for long periods (filling the cache with that vliw optimized code). When multitasking it apparently has serious issues. Hopefully these can be addressed in future revisions to denver.
I'm not comparing denver to atom or other current android offerings as none of them have attempted to go use a radically different arch from everyone else.
DERSS - Wednesday, November 5, 2014 - link
"But the N9 has the brighter screen and a smaller battery, and yet N9 has slightly battery life" -- battery life is measured with equal brightness, so K1's longer battery life is result of smaller screen and dramatical throttling slowdown which is unavoidable since it is 28 nm process (versus 20 nm in A8X).kron123456789 - Wednesday, November 5, 2014 - link
What "dramatical throttling slowdown" are you talking about? Do you know what "throttling" means?DERSS - Thursday, November 6, 2014 - link
Yes, clock drops down to SoC not to overheat and to save battery life.testbug00 - Thursday, November 6, 2014 - link
in a web browsing test? The CPU isn't high enough to throttle... Or, if it is the webpage is more intensive than an 'average' mobile web page, OR, the SoC is terribly designed.While I am very down on Nvidia's mobile efforts due to peak power draws and throttling, it shouldn't be throttling anymore than the A8X in sane web browsing tests...
melgross - Monday, November 3, 2014 - link
We were being told that ARM could never compete with x86, but the A8x is stronger than most i3 low power chips. I assume this new Tegra is somewhere around that as well. The fact that it does compare with a three year old i5 notebook is something that I find incredible.Considering that Apple and possibly others, will continue to upgrade performance in their ARM offerings by substantially greater amounts than Intel is able to with x86, for at least the next few years, means that Intel really does have something to worry about.
kron123456789 - Monday, November 3, 2014 - link
Yeah, but here's what i think — iPad Air 2 and Nexus 9(and all Tegra K1 devices as well) has so much power and that power is used ONLY in benchmarks.tuxRoller - Thursday, November 6, 2014 - link
ipad has some pretty serious music making apps. Both have some cad programs from autodesk.Lastly, browsers themselves can actually use that power. For instance, firefox via webgl + asm.js.
This is ignoring the coming push into servers.
toyotabedzrock - Monday, November 3, 2014 - link
Not when I can buy an i5 laptop for the price of the tablet.beginner99 - Tuesday, November 4, 2014 - link
No, an i3 runs circles around A8x or Tegra K1. It's just a stupid comparison because they also target a different market.darkich - Tuesday, November 4, 2014 - link
You're trolling??The A8X is ridiculously more impressive than any Core chip because of giving comparable processing power at about 30% of power, and probably far less than 50% of price.
And here's the thing - Apple AX chips are a prime candidate for future MacBook successors.
Speedfriend - Tuesday, November 4, 2014 - link
"The A8X is ridiculously more impressive than any Core chip because of giving comparable processing power at about 30% of power, and probably far less than 50% of price."You heard of a thing called Core M? Same TDP as A8x, single core geekbench 33% higher. GPU performance 2.5x? And on silicon that is smaller than the A8 let alone the A8X, so cost of manufacture is cheaper.
darkich - Tuesday, November 4, 2014 - link
Yeah I heard of Core M, that Uber fail chip that perfectly shows how badly screwed Intel is against Apple.It throttles tragically, has 30% lower performance than a Haswell U counterpart, and costs over $200.
darkich - Tuesday, November 4, 2014 - link
^ at least that is what the first actual Core M product reveals, that ischizow - Tuesday, November 4, 2014 - link
Core M is definitely in a class of its own in terms of performance, but I agree its not applicable in this discussion because of its price tag. We won't see it in a sub $600 device competing with these parts.darkich - Wednesday, November 5, 2014 - link
Definitely in a class of its own??Please back that claim up.
From what we have seen, Core M in the Yoga 3 Pro (high end core M variant!) fails to match the Core i5 4200U, and .. Apple A8X in fact matches it in raw performance!
My bet is that Apple A9X on 14nm should allow the MacBook Air succesor to outperform windows competition in every way, and actually cost less.
darkich - Tuesday, November 4, 2014 - link
Oh and also, Core M doesn't have the same TDP as A8X, not even close!The A8X has, according to battery endurance measurements, has lower TDP than the A7..you know, the chip that runs in the iPad Air and..iphone 5S.
Think about that for a moment.
The only reason why A8X isn't in new iPhones is because the A8 is sufficient for iPhones while A8X(along with the 2GB RAM) turns the iPad into a bit more than the big iPhone.
kron123456789 - Tuesday, November 4, 2014 - link
It's not about power, it's about how you can use that power. On Windows there are many ways to use power of any Core chip, on iOS there is pretty much no way you can use power of A8X.DERSS - Wednesday, November 5, 2014 - link
Easy; Apple has shown heavy load iPad Air 2 applications during presentation.kron123456789 - Wednesday, November 5, 2014 - link
Yup. It's like...two of them, yeah? That's definitely makes a difference.ins1dious - Wednesday, November 5, 2014 - link
Have to start somewhere... I was more impressed with the pixelmator demohttps://www.youtube.com/watch?v=sBfvJn-fpnc#t=3138
I think these ARM chips will start to become more useful once more and more developers take advantage of their power.
DERSS - Thursday, November 6, 2014 - link
@kron, you wrote "But pretty much no way you can use power of A8X", which directly wrong. By now there are already few dozens of heavy load apps that load things (many graphics-related, medical processing pro apps, and so on).kron123456789 - Thursday, November 6, 2014 - link
I bet those apps are working well with A7 chip. Correct me if i'm wrong.ins1dious - Thursday, November 6, 2014 - link
@kron - but with Pixelmator as example... you see 2x performance. Isn't that good?testbug00 - Tuesday, November 4, 2014 - link
The Tegra chip should be some where around equal on Integer and about a half(!) the FP performance of the A8X.retrospooty - Monday, November 3, 2014 - link
"Overall this device I think missed on the lack of expandable storage and low entry level 16GB storage. "Yup. This one boggles the mind. They did 32gb on the Nexus 6. WTF is with 16gb as a start point for a high end tablet? Other than that, its a pretty darn good tablet though. Amazing GPU.
tipoo - Monday, November 3, 2014 - link
Denver looks like some good design choices. 2 big cores are better than 4 scrawny ones, glad to see Android SoC makers learn that, ironically it's the same one that sent everyone chasing 4 cores. The VLIW+Code morphing architecture is interesting, in thoery it would take a hit from spaghetti code, but with 128MB of far caching it shouldn't really run into situations where it's reoptimizing the same code a lot.tipoo - Monday, November 3, 2014 - link
Maybe I spoke too soon, seeing the benchmarks, it seems to ping between great and average. Joshua, think what I said is coming into effect with spaghetti code?abhaxus - Monday, November 3, 2014 - link
Well, in the article he points out that the firmware is nearly two months old. I imagine there are more optimizations in the release FW.tipoo - Monday, November 3, 2014 - link
Yeah, hopefully it's all in the software. The CPU looks great when it does well, and the GPU absolutely is killer.Dribble - Monday, November 3, 2014 - link
When other sites are getting full reviews up anandtech can only manage a half hearted preview using an ancient build?What's happened?
Andrei Frumusanu - Monday, November 3, 2014 - link
The final firmware builds were supposedly pushed Sunday afternoon via OTA. Can't comment on what other sites do but our review will be on that firmware, and understandably it's impossible to do a review within 24 hours.melgross - Monday, November 3, 2014 - link
I wouldn't expect to much of a gain. A few percent here and there most likely.tential - Monday, November 3, 2014 - link
So go read those reviews then?Wait, you still came here to complain about anandtech's review not being up?
I'm going to guess it's the same reason I came here to see if anandtech's review was up. Because I wanted a more in-depth review. Depth takes time obviously so I guess we'll have to wait.
edwpang - Monday, November 3, 2014 - link
True, some reviews from other sites contains even less information than this preview. If Anandtech posts a review too quickly without fully testing the hardware, a lot more people will complain.chizow - Monday, November 3, 2014 - link
I guess it also really depends what you consider a "full review". Personally I think AnandTech is the only site that actually tries to do a comprehensive performance review from a true tech enthusiasts point of view.Now compare with these other consumer tech gadget sites that throw up 3 benchmarks for 3 devices with another 2500 words of subjective filler regarding aesthetics that anyone who walked into a Best Buy could ascertain on their own.
dragonsqrrl - Monday, November 3, 2014 - link
I'll wait for quality analysis. As the article stated the final firmware has only been available for a day, so that should tell you something about the "full reviews" you love so much.tipoo - Monday, November 3, 2014 - link
This has always been the way, never the first review out, but almost always the most in depth.esterhasz - Monday, November 3, 2014 - link
OK, preliminary numbers, no fast conclusions. But I can't help feeling that the shield tablet looks actually pretty attractive in that lineup.melgross - Monday, November 3, 2014 - link
I really had expected a typical full iPad Air2 review before this. I can't understand what's taking so long.It would be more interesting to see full reviews of both already. Include the latest Full sized Samsung model.
Krysto - Monday, November 3, 2014 - link
I wish Anandtech would stop focusing so much on Sunspider. Google said it 2-3 years ago that they aren't focusing Chrome on Sunspider performance anymore because it's a very simplistic test without much relation to real-world web app performance. They even did a 50xSunspider test to prove how hopefully obsolete it is, if some of you remember.The Basemark OS benchmarks are all over the place. In some it tests twice as good as the iPad, in others half as flow. I don't think Basemark OS has its algorithms well optimized, and it may give some subcomponents way more score than they deserve, or way too little in others - or Basemark OS may simply be acting weirdly with this new processor.
melgross - Monday, November 3, 2014 - link
Kraken seems to give the same relative performance results with the same ordering of devices, so it doesn't seem to make much difference which one they use.danbob999 - Monday, November 3, 2014 - link
javascript benchmarks are all useless to benchmark hardware. They were created to benchmark software (javascript engines)melgross - Monday, November 3, 2014 - link
It's not just benchmarks g hardware. It's benching the OSs and the browser engine too. So it does matter. Besides, as I said, since Kracken scores seem to follow the Sunspider scores in ranking, it doesn't matter than much.darkich - Tuesday, November 4, 2014 - link
Exactly.My Note 3 scores 570ms in Samsung browser, while Chrome manages only around 1100ms
arsjum - Monday, November 3, 2014 - link
Are we reading the same chart? In Sunspider chart, Nexus 9 is well below iPad Air, whereas in Kraken benchmark, Nexus 9 shows a slightly better score than than iPad Air 2.eiriklf - Monday, November 3, 2014 - link
Sunspider results are far too easy to optimize for in the browser, but it will differ for different architectures. Krait could for instance not really distance itself from cortex a9 in sunspider for a good while on Android.eiriklf - Monday, November 3, 2014 - link
That is, the optimizations seem to be architecture specific.Krysto - Monday, November 3, 2014 - link
What I'd like to know - is the tablet actually taking advantage of the 64-bit/Aarch64 mode? Or is it running in 32-bit mode still? Because I think it was still running in 32-bit mode in early Geekbench tests. I don't know why Google would release it like that, but it would be disappointing if they did, since that could be cutting its CPU performance by 10-20 percent.Aenean144 - Monday, November 3, 2014 - link
Is it really a guarantee that 64-bit performance will be better than 32-bit performance with a VLIW architecture?The performance is going to be highly dependent on how well the code morpher can pack instructions into a VLIW instruction. So, if the VLIW instruction is 256 bits wide, does that mean it can only pack half the 64-bit instructions?
Andrei Frumusanu - Monday, November 3, 2014 - link
It's running on a 64 bit kernel.kron123456789 - Monday, November 3, 2014 - link
But apps(benchmarks included) aren't 64 bit.Marcos Stein - Monday, November 3, 2014 - link
64bit optimization will not save it.http://arstechnica.com/gadgets/2014/11/nexus-9-rev...
kron123456789 - Monday, November 3, 2014 - link
1) It's not final firmware on Nexus 9.2) It's pre-alpha (or something) build of 64-bit Geekbench.
We'll see when the final version of 64-bit Geekbench will be released.
AnandTechUser99 - Monday, November 3, 2014 - link
I would wait until these benchmarks can test AArch64 for Android before jumping to a conclusion on performance.Any chances of seeing performance in GFXBench 4.0? Or is that not ready yet?
All of these GPU benchmarks test OpenGL ES 2.x or 3.x performance, which misses the point of Tegra K1.
Marcos Stein - Monday, November 3, 2014 - link
Gfxbench's results posted here were very generous.Nexus 9's Gfxbench onscreen results has only 2048x1440. The results will be worse when Gfxbench update the software to process 2048x1536 (196608 more pixels).
The same happened with Iphone 6
Fidelator - Monday, November 3, 2014 - link
Seems pretty competitive with the Air 2 all around, what about some display calibration details? Those would be greatly apprecitedNow, the real catch will be next year, when we get quad core Denver + a Maxwell GPU on a 20nm process, they were able to match the A8X in pretty much everything but some CPU bound tests with a bigger process
Still, after seeing these results the display calibration and speaker sound quality will determine if I get it or not
andrewaggb - Monday, November 3, 2014 - link
Personally, I thought it looked like the ipad air 2 was the clear all around winner. Hopefully the final build will do better.darwinosx - Monday, November 3, 2014 - link
Plus the Nexus 9 screen is crap.icebox - Monday, November 3, 2014 - link
As always Google hopes to push cloud storage to everybody with a max. of 32GB and no expansion.jjj - Monday, November 3, 2014 - link
Funny how you don't include TK1 results in phone reviews graphs but you include the iphone in tablet reviews. Intentional or not it's still unethical.melgross - Monday, November 3, 2014 - link
That's a very discordant statement. How could it be unethical if it's done unintentionally?Ryan Smith - Monday, November 3, 2014 - link
We don't have a TK1-equipped phone. Right now the only devices it appears in are tablets.ruggia - Monday, November 3, 2014 - link
think he's upset about 5s being in the benchmarkBarilla - Monday, November 3, 2014 - link
I'm hoping the full review will dive deeply into the cpu design, I'm really interested what Nvidia has actually cooked up with their project Denver.toyotabedzrock - Monday, November 3, 2014 - link
When did the benchmarks get run? Shouldn't the final build be available today? Or will it be over the air when the first tablets reach people in the mail?Liveartonline - Monday, November 3, 2014 - link
Wow, the GPU perfomance of the Samsung Galaxy Tab S is utter trash. I didn't know the GPU performance of the Mali-T628 would be THAT bad compared to the K1 and A8X. Thank god I didn't buy one.speculatrix - Monday, November 3, 2014 - link
Quite often Nexus devices have a close sibling from the same manufacturer but with extra hardware niceties, such as gaining a memory card slot, different or removable battery, USB on the go etc.So it will be interesting to see if HTC come out with a very similar table to this Nex9 but with a memory car slot and some premium features?
arkhamasylum87 - Monday, November 3, 2014 - link
Can you guys do a comparison article between apple a8x, tegra k1 and intel core m ? would love to understand your thoughts on the tablet/hybrid market as a whole and if there is a potential here.varad - Wednesday, November 5, 2014 - link
+1agoyal - Monday, November 3, 2014 - link
I think the real question is, how will it fair in comparison with A57 which should be coming soon!!darkich - Tuesday, November 4, 2014 - link
The 20nm 1.9GHz A57 is already out in the Galaxy Note 4.Scores around 1250 in single core Geekbench.
But power draw in the octa core setup should be incredibly low considering that the 20nm Exynos 5430(Cortex A15/Galaxy Alpha) doesn't even cross 3W under maximum load!
Also,if the rumors are true that Snapdragon 810 will have it clocked at 2.7GHz, the end result should be quite incredible.
NotLupus - Monday, November 3, 2014 - link
I bought a Nexus 7 2013 from Walmart last year just as it was released. Battery life was about 6 hours of screen time with low brightness and no background syncing, doing mostly web browsing or local video playback. It was nowhere near Anandtech's 12 hours, not even the advertised 9 hours. Ever since then I'm skeptical battery life claims for android devices. This Nexus 9 review sample was probably cherry-picked.agoyal - Monday, November 3, 2014 - link
I have had similar experience with android devices, I have owned nexus 7 2013,LG G pad 8.3 and Galaxy pro tab 10.1. Had IPad Air and the battery would last a long time mostly with web browsing, much more than the review sites.poohbear - Monday, November 3, 2014 - link
between my 5.7" Note 3 phablet, my 13" ultrabook, and my 27" desktop beast, what exactly do u need a tablet for? All my friends that bought one say it just sits there unused. Seems like a niche market or just a gimmick.WarlockOfOz - Tuesday, November 4, 2014 - link
One solution to your question: replace the phablet with a small, basic phone plus a 8" wide screen tablet and you get something that's more convenient for actual phone calls, provides more screen space for everything else, has a lower total cost and can still be carried in your pockets. Not sure about tablets that need bags myself, but lots of other people seem to like them.darkich - Tuesday, November 4, 2014 - link
I disagree that 8" wide screen tablet is more convenient than a 5.7" palmtop (I happen to own both, palmtop being the Galaxy Note 3).The Note 3 is simply much easier to hold. And the screen is still perfectly big for reading and even movies while laying in bed.
Handling Web pages and gaming is also much easier.
While my tablet is perfectly okay for what it is, I find it using it mostly when my Note 3 is charging.
So I tend to agree with the above post. Phablets (or palmtops as I prefer to call them) are the reason for tablet market stagnation, and are rightfully booming right now.
Just look at Apple's reaction if you need further evidence - neutering the iPad mini a release of the iPhone 6 plus.
konondrum - Tuesday, November 4, 2014 - link
I think Google's point is that everyone has a slightly different idea of an ideal size. That's why they didn't make a nexus 7 (2014) this time.I bought an LG G2 last year because I wanted a screen big enough do some actual reading on without the size of a Note. But now that I've got a Shield Tablet, I'm seriously thinking about getting a smaller phone. The G2 isn't bulky in my pocket, it's just a hair too big to be used comfortably as a phone.
probedb - Tuesday, November 4, 2014 - link
Why do sites like anand insist on publishing pre-reviews like this? dpreview is the same for cameras, several first impressions/almost there but not quite complete/final view type things.Why don't you wait until you have release software? No-one is going to be using the pre-release build used in this review so why review it?
I normally appreciate Anandtech reviews and it's the first place I go to but this sort of thing is just annoying.
OrphanageExplosion - Tuesday, November 4, 2014 - link
Joshua, please read Futuremark's comments on Apple's CPU architecture and why 3DMark Physics (which is a CPU test and perhaps should be in the CPU benchmark section?) performs relatively badly on Cyclone CPU cores: http://www.futuremark.com/pressreleases/understand...My guess is that Denver has similar limitations - it's exceptionally good at certain types of task (ie the simple tests found in typical benchmark scenarios) but isn't as capable on other tasks. Or maybe drop Futuremark's a line for further information? Based on the in-depth Cyclone analysis I just linked to, I am sure they will be happy to clarify Denver's disappointing performance in their benchmark.
DBissett - Tuesday, November 4, 2014 - link
The text suffers from a great deal of awkward and even contradictory wording to the point that it detracts from the content. Hopefully an editor will work harder on the full review.cjs150 - Tuesday, November 4, 2014 - link
I have a simple rule. No MicroSD slot means no purchase. There is a fixation shared by most tablet designers that everything is in the cloud and therefore there is no need for a decent amount of storage space.There is a simple message to the designers - grow up and get out more. There are plenty of parts of the world where internet access is "patchy" even at international hotels or expensive or both. Then there are those parts of the world which have no access - aircraft. I am in the market to replace my 3 year old Nook HD+, Nexus 9 will not be on the shortlist
odedia - Tuesday, November 4, 2014 - link
Are there any plans for an iMac 5k review?Wolfpup - Tuesday, November 4, 2014 - link
I'd have loved to have seen how well this compared to Intel's second gen Atom, had the licensing worked out. Presumably the GPU would be better if nothing else...maybe Nvidia could have shipped a Windows version of their Shield products, which would be awesome!Still tempted by this since there's no Windows version of Marvel Unlimited...
soccerballtux - Tuesday, November 4, 2014 - link
HI DOES THIS HAVE THE GOOGLE ON IT?/varad - Wednesday, November 5, 2014 - link
NO, THE GOOGLE IS STILL IN MOUNTAIN VIEW. /SPwnstar - Tuesday, November 18, 2014 - link
LOLnodecodiver - Tuesday, November 4, 2014 - link
Been playing with mine all morning. I don't really see the performance issues everyone is talking about. My "Recent Screens" button does not take 2-3 seconds to pop up, it's quite responsive, actually. I'm not a huge gamer, so I can't comment there, but just in general use as a media consumption device / large screen mobile browser / productivity device, I couldn't be happier.Liveartonline - Tuesday, November 4, 2014 - link
As a side note, anyone care to explain to me why Samsung Galaxy Tab S routinely scores worse than the Tab Pro even though they are pushing the same number of pixels? Based on the graphs, Samsung's own Exynos 5420/Mali-T628 combo is inferior to last year's SD800+Adreno 330?toyotabedzrock - Tuesday, November 4, 2014 - link
What is with the sunspider regression? It seems to indicate the need for further code optimization. Perhaps try chrome beta and update the firmware. Ars had the newer firmware for their review.nodecodiver - Tuesday, November 4, 2014 - link
Google said a while back they could basically care less about Sunspider because it's a poor representation of actual web browsing, which, any extended use of Safari vs. Chrome (on a droid) bears out. There is no appreciable difference in the web experience on either device, although I prefer Chrome just for the universal tabs, etc.raghwendra123 - Wednesday, November 5, 2014 - link
What I don't understand is how is Basemark OS II graphics score double that of ipad air 2 yet its performs almost the same in offsceen benchmarks?kf27fix - Wednesday, November 5, 2014 - link
AnandTech should start testing interface responsiveness quantitatively. E.g. when I swipe or click, I want the device to react instantaneously - 200 milliseconds is way too long, I need 1 millisecond. A high-speed camera and a piece of software to analyse the video recording should do the trick.In my opinion, this is the main difference between Apple and Android world. I always feel almost ashamed when I use my Android phone next to someone with an iPhone. My phone seems to always do something important in the background exactly when I need it to do something for me fast.
Desusenam - Friday, November 7, 2014 - link
100% agree with the principle of this test. I've definitely noticed my Note 3 a bit laggy at time. I expect the device to work for my needs and not the other way around. I also like the idea of doing some 'real world' tests with a whole load of apps open, various browser tabs open, then switch tasks, ring the phone etc. and check responsiveness. However, maybe I would have same issue with Apple.tipoo - Sunday, November 9, 2014 - link
You "need" 1, I assume you don't use any touch devices? I think ~47 was the fastest tested, the One M8 being one of the fastest.lucam - Thursday, November 6, 2014 - link
When is it going to happen to have the complete recension of both devices for having a clear idea? I think preliminary results in this article only cause flames between users. Bit disappointed by Anand sincerely.Desusenam - Friday, November 7, 2014 - link
I purchased one for my daughter which we received on Tuesday, previously she had an iPad 2, three and a half years ago.She's happy with it (it plays games and videos and surfs the internet etc.) and I don't need to install iTunes. So, a win-win situation.
HOWEVER, we still cant buy ANY cover for it on Google Play.
WTFlip is that about. Everyone here is talking about what SOC and how many cores and blah blah, but they can't even sort out the cover for the device.
I had the same issue last year with the 7 inch Nvidea tablet. Launched and we couldn't buy the great cover that was designed for it.
So, in this regards Apple wins by a long way. The iPad 2 was engraved on the back for free and it arrived with the cover at the same time the iPad came.
By the way, the cover protects the screen and allows you to stand the tablet up etc. to watch movies. Pretty essential for most people I would think... (and yes, I know I could get a generic cover or put it in a paper bag or something but really, I would like the cover designed for the device that gets advertised...)
baycorn - Thursday, November 27, 2014 - link
when is the review coming??? I was going to get a Nexus 9 but now aftet bad reviews am seriously thinking about an ipad mini 2 - waiting for the always great anandtech review before making final decision!jji7skyline - Sunday, November 30, 2014 - link
I gave up waiting for the review to decide between the Nexus 9 and the iPad Air 2. I just bought the Nexus 9 a few days ago and I am waiting for it to ship.Maleficum - Tuesday, December 30, 2014 - link
I think the full review is delayed so much because Anandtech is currently consulting lawyers fearing nVidia's wrath. The K1 Denver is the biggest benchmark cheater, Anandtech found it out, and....... you know the story between 3DMark and nVidia.To be clear: the K1 Denver is an *emulator*, NOT a custom designed ARM. It's so easy to write an emulator that cheats primitive, predictable, synthetic benchmarks.
For example, write a function that does hundreds and thousands times memcpy on same addresses. With the -O3 option, the compiler optimizes the function so that it does memcpy only once.
Benchmarks are written in a way that the resulting machine codes actually do all the iterations, but with an emulator that translates these machine codes to a different one, it can easily mimic the compiler's behavior mentioned above. Of course, it would be "optimization" by nVidia's definition.
This kind of "optimizations" were done before by Transmeta, and K1 Denver is a reincarnation of Transmeta's Crusoe/Efficeon. And we all know that nVidia .......... well, the rest is up to you guys.
http://www.vanshardware.com/articles/2003/07/03071...
PS: Sunspider is one of very few "uncheatable" benchmarks, and you know how Denver fares in that.
dragonsqrrl - Thursday, January 15, 2015 - link
Soo... I know I said I'd wait for a quality analysis, but isn't this a bit much? Are we going to get a Nexus 9 review Josh?snoukkis - Friday, January 23, 2015 - link
Yup, still waiting for the review with the final shipping product...Vinny DePaul - Saturday, August 22, 2015 - link
I hope someone will see this and help me. I ran those benchmarks on my Nexus 9 and mine is almost 200% slower, i.e., double sunspider score. Did I get a fake Nexus 9?