GPU Benchmarks
GFXBench Offscreen – T-Rex
The T-Rex test is based on OpenGL ES 2.0 and includes textures, material, geometry and particle effective that were highly detailed at the time of release. The graphics rendering engine features planar reflection, specular highlights and soft shadows, providing a good workout even for flagship smartphones and tablets. Offscreen tests are run at 1080p, regardless of the device’s native resolution, and are best used to compare the performance between competing silicon, not competing devices.
In this first test we see the Nexus 6 coming out on top with the OnePlus One not far behind, and it appears that the Adreno graphics in the Qualcomm chips outclasse the Zenfone 2's PowerVR G6430, with both ASUS phones at the bottom here – at least with OpenGL ES 2.0.
GFXBench Offscreen – Manhattan
Manhattan was the first benchmark to utilize OpenGL ES 3.0 features and uses a nighttime setting with a lot of external illumination to stress the GPU. It uses a deferred rending engine with multiple render targets for the geometry pass, includes both diffuse and specular lighting, uses depth shadow maps, bloom, depth of field and quite a bit more.
With OpenGL ES 3.0 the Zenfone 2 has a better showing, with both models moving up a step. Samsung's Exynos chip (with Mali T760 graphics) is the winner this time with an impressive showing.
GFXBench Offscreen – ALU Performance
This test measures the pure shader compute performance using a fragment shader and rending a single full-screen quad.
An odd result here considering the OnePlus One only has a Snapdragon 801, but the Nexus 6 result was repeatable. Again the Zenfone results are lower than the Snapdragon devices, but this time it's the Galaxy Note 4 at the bottom.
GFXBench Offscreen – Alpha Blending Performance
Rendering semi-transparent quads with uncompressed textures allows this test to measure the alpha performance of the GPU directly.
Alpha performance with the Zenfone 2 was nearly identical between the two, and again we see these phones struggling to keep up with a result that is only half that of the Nexus 6.
GFXBench Offscreen – Driver Overhead
This test renders a large number of very simple objects one-by-one, changing state with each item in a pattern consistent with real-world applications. This allows the benchmark to measure the CPU overhead of the OpenGL driver.
Samsung's international Galaxy Note 4 seems well optimized for this driver overhead test, and the results aren't close.
GFXBench Offscreen – Fill Rate Performance
The portion of the test measures texture fill rate performance by rendering four layers of compressed textures, a very common scenario in gaming.
Back to a more real-world test we see raw performance of the Snapdragon 805's Adreno 420 graphics is too much for the rest of the phones tested.
GFXBench – Native Resolution
All of the above tests were run at 1080p to compare the performance of the GPU systems on equal ground. But smartphones are not equal ground as the displays that are part of the platform differ from unit to unit, and resolution can and does directly impact perceived GPU performance. Both the Nexus 6 and Glaxy Note 4 have 2560×1440 native resolution displays, while the Zenfone 2 and OnePlus One each have native 1920×1080 displays.
Here's were things start to get interesting. It's not surprising that the three phones with native 1080p panels would win this test against a pair of QHD (2560×1440) devices, but the Zenfone 2's showing is very impressive here in both tests.
3DMark
Use 3DMark Ice Storm Unlimited for chip-to-chip comparisons of the hardware inside your device without vertical sync, display resolution scaling and other operating system factors affecting the result. In Unlimited mode the rendering engine uses a fixed time step between frames and renders exactly the same frames in every run on every device. The frames are rendered in 720p resolution "offscreen" while the display is updated with small frame thumbnails every 100 frames to show progress.
Ice Storm Graphics test 1 stresses the hardware’s ability to process lots of vertices while keeping the pixel load relatively light. Hardware on this level may have dedicated capacity for separate vertex and pixel processing. Stressing both capacities individually reveals the hardware’s limitations in both aspects. Pixel load is kept low by excluding expensive post processing steps, and by not rendering particle effects.
Graphics test 2 stresses the hardware’s ability to process lots of pixels. It tests the ability to read textures, do per pixel computations and write to render targets. The additional pixel processing compared to Graphics test 1 comes from including particles and post processing effects such as bloom, streaks and motion blur. The numbers of vertices and triangles are considerably lower than in Graphics test 1 because shadows are not drawn and the processed geometry has a lower number of polygons.
The purpose of the Physics test is to benchmark the hardware’s ability to do gameplay physics simulations on CPU. The GPU load is kept as low as possible to ensure that only the CPU’s capabilities are stressed. The test has four simulated worlds. Each world has two soft bodies and two rigid bodies colliding with each other. One thread per available CPU core is used to run simulations. All physics are computed on the CPU with soft body vertex data updated to the GPU each frame. The background is drawn as a static image for the least possible GPU load. The Ice Storm Physics test uses the Bullet Open Source Physics Library.
The overall result with 3DMark is always going to scale with CPU speed as well as GPU horsepower, and here the difference in the two Zenfone SoCs is very clear. The graphics edge still goes to the faster phone, a sign that the 1.83 GHz clock speed might be a bottleneck with the 16GB version. The Nexus 6 enjoys another win here, showing the performance of the Adreno GPU yet again. Finally in the physics test the Zenfone 2 64GB wins out by more than 200 points, making the $500+ smartphones tested look rather slow in comparison. This Atom SoC can be very impressive.
Basemark X
Basemark X is the world’s most popular benchmarking tool for evaluation and cross-platform comparison of gaming and graphics performance between Android, iOS and Windows Phone 8 smartphone and tablets.
Basemark X is the only vendor-independent benchmark that utilizes the real-world game engine Unity which is very popular among game developers. This means that it scores correlate exceptionally well with real-life gaming performance.
Basemark X includes two game-like graphics tests: Dunes and Hangar. Both tests contain heavy graphics content rendered with detail and complexity, thus pushing the measured device to the limit. The polygon counts in test sequences are up to 911,000.
The Nexus 6 and Note 4 both finished ahead by a wide margin with both medium and high settings, and in this "real-life" gaming simulation both Zenfone models struggle in comparison (though the results are right there with the Snapdragon 801 powered OnePlus One with high settings).
Basemark OS II – Graphics
Basemark OS II is a system-level All-In-One benchmarking tool designed for measuring overall performance of smartphones and tablets from all platforms, including Android, iOS and Windows phone 8.
The benchmark features a comprehensive suite of tests including system, internal and external memory, graphics, web browsing, camera, battery and CPU consumption.
The final synthetic graphics benchmark brings us full-circle with another result that heavily favors the more expensive phones tested. In this case the Nexus 6 dominated, with both Snapdragon devices finishing well above the rest.
Oh boy, now I can destroy my
Oh boy, now I can destroy my privacy for only $200! Thanks Pcper for letting me know about this amazing opportunity!
nobody cares about you so
nobody cares about you so your privacy is safe
Hi Sebastian,
thank you for
Hi Sebastian,
thank you for the detailed review! You might want to add that $199 model also supports Quick Charge technology, but doesn’t come with the Quick Charge adapter in the box.
Question, did you have a chance to use lower specs model (2GB/16GB Z5360 CPU)? I understand that it slower then $299 one, but how does it feel/respond during Lollipop normal usage?
Thank you!
Thanks, and the BoostMaster
Thanks, and the BoostMaster compatibilty is an important distinction. I didn't personally test the 16GB version with the Z3560, as the sample I was using was the $299 model. Given the way the SoCs are optimized I don't think there would be a significant UI/UX difference, but there's an obvious performance difference from our benchmarks in certain applications.
The best thing about this SOC
The best thing about this SOC is that it is not using Intel Graphics, so expect much better graphics from a GPU with the execution resources to do the JOB. I just wish that M$ could persuade Intel to use even higher end powerVR graphics and make a surface tablet useful for graphics uses. I hear that Intel is moving away from PowerVR to Mali graphics, but as long as its not Intel’s graphics.
Thank you Sebastian. I
Thank you Sebastian. I ordered $199 model for my dad and it should be here next week. It will be his first Android phone coming from candy-bar Nokia, so we’ll see how it’s going to be. But I didn’t feel he would need the higher spec mode and quick charger and microSDXC 64Gb card is around $40 from Amazon.
If Ryan was testing Z5360 model, could you ask him to chime in the comments with his experience of it?
Too bad that camera seems to be lacking a bit for daylight shots, though it seems that Asus releases firmware updates every couple weeks for this phone, so may be things will get better. If I like what I see, I might get myself a $299 one later this summer.Thanks again for the review!
I used the model with the
I used the model with the lower end processor but I will say that the experience and fluidity of the unit was just as Sebastian described in the review. It was smooth and I didn't know any performance penalties of the lower cost model even when I had both phones in my possession at the same time.
That’s good to know, thank
That’s good to know, thank you Ryan. I enjoy your site a lot and listen to the podcast on weekly basis. That’s actually how I remembered about ZenFone2 (even though I saw CES presentation about it), when Sebastian had as his pick couple weeks ago. Keep up a good work and thanks again!
Any chance you could find out
Any chance you could find out which benchmarks are x86 compiled and which ones are running ARM via translation layer? Or did you mention it and I missed it?
That’s a great question. I
That’s a great question. I need to do some research as I’m not sure off hand. The difference in performance in some of the benchmarks could easily be explained based on native/non-native code, though in the real world I do think it’s fair to judge it as-is considering one does have the option of downloading x86 apps from the Play store (not that devs would offer it at this point anyway).
bad.
bad.
Ryan..I really want to win
Ryan..I really want to win that car! It’s being given away on the next podcast I hear!
It has two 980ti’s installed
It has two 980ti’s installed as well I hear.
I would have guessed a pair
I would have guessed a pair of TITAN Z’s for the car. Gotta have that double precision for better handling