Snapdragon 810 Performance Preview

Testing Snapdragon 810

Benchmark Suite

For our first look at Snapdragon 810 we will be examining CPU, Web, and GPU performance.

Swipe to scroll horizontally
HTML5 And JavaScript BenchmarksBrowsermark 2.1, Google Octane 2.0, JSBench, Peacekeeper 2.0, WebXPRT 2013
CPU Core BenchmarksAndEBench Pro v2.0.2124, AnTuTu v5.6, Basemark OS II Full (Anti-Detection) v2.0, Geekbench 3 Pro v3.3.1, MobileXPRT 2013
GPU Core Benchmarks3DMark (Anti-Detection) v1.2, Basemark X 1.1 Full (Anti-Detection) v1.1, GFXBench 3.0 Corporate v3.0.0

Test Methodology

All handsets are benchmarked on a fully updated copy of the device's stock software. The table below lists other common device settings that we standardize to before testing.

Swipe to scroll horizontally
BluetoothOff
Brightness200 nits
CellularSIM card removed
Display ModeDevice Default (non-adaptive)
Location ServicesOff
PowerBattery
SleepNever (or longest available interval)
VolumeMuted
Wi-FiOn

Furthermore, for browser-based testing on Android, we're employing a static version of the Chromium-based Opera in order to keep the browser version even across all devices. Due to platform restrictions, Safari is the best choice for iOS-based devices, while Internet Explorer is the only game in town on Windows RT.

Comparison System Specs

For this analysis we will be focusing on SoCs rather than devices, specifically those found in tablets. To get a feel for Qualcomm's progress, we'll compare Snapdragon 810 to its predecessors—Snapdragon 805, represented by the Nexus 6, and 801, represented by the Sony Z3. Since we're not comparing displays, battery life, or onscreen GPU performance, mixing smartphones with tablets in our benchmark charts won't matter. The Nexus 6 was chosen because it displays less thermal throttling than the Galaxy Note 4 and has a more complete and up-to-date set of benchmark data than our other devices. For Snapdragon 801 we're using the Z3 because it's a top performer with consistent results free of thermal throttling. The Galaxy S5 actually has slightly better performance in multi-core CPU tests, but the Z3 pulls ahead in graphics, memory, and NAND.

We also include four other current tablet architectures for comparison: the 3-core A8X in Apple's iPad Air 2, an Intel Atom Z3580 in Dell's Venue 8 7000 (which is still stuck on Android 4.4, limiting it to 32-bit mode), an Nvidia Tegra K1 with Denver cores in HTC's Nexus 9, and Tegra K1 with ARM cores in Nvidia's own Shield Tablet.

We're also including some results for Nvidia's new Tegra X1 SoC with its Maxwell based GPU. Be advised however, that these numbers came from Tegra X1 development boards running Android Lollipop. Although we couldn't run our own versions of the benchmarks and couldn't verify clock speeds, we did see the benchmarks running. The SoC itself had a small heatsink (no fan), which Nvidia said represented the typical thermal dissipation capability of a tablet chassis.

Snapdragon 810 MDP Tablet

The data for Snapdragon 810 comes from a Mobile Development Platform (MDP) tablet supplied by Qualcomm at its benchmarking workshop. Obviously, this tablet isn't shipping hardware (it's noticeably bulkier than a production device, which could improve its ability to deal with heat buildup) and the devices were running on AC power, so these results should be considered preliminary. We did however, install and run our own benchmarking software. As a point of reference, the Snapdragon 805 MDP tablet generally scored worse than shipping products.

Unfortunately, we'll have to wait for products to ship before we can get a feel for the 810's effect on battery life, since there wasn't enough time to collect battery drain data.

The table below contains all the pertinent technical specifications for today’s comparison units:

Swipe to scroll horizontally
Header Cell - Column 0 Qualcomm Snapdragon 810 MDP TabletMotorola Nexus 6Sony Z3Apple iPad Air 2Dell Venue 8 7000HTC Nexus 9Nvidia Shield Tablet
SoCQualcomm Snapdragon 810 (MSM8994)Qualcomm Snapdragon 805 (APQ8084)Qualcomm Snapdragon 801 (MSM8974AC)Apple A8XIntel Z3580Nvidia Tegra K1Nvidia Tegra K1
CPU4x ARM Cortex-A57 @ 2.0GHz + 4x ARM Cortex-A53 @ 1.5GHz (big.LITTLE)Qualcomm Krait 450 (4 Core) @ 2.65GHzQualcomm Krait 400 (4 Core) @ 2.45GHzApple Cyclone+ (3 Core) @ 1.5GHzIntel Atom (4 Core) @ 2.33GHzNvidia Denver (2 Core) @ 2.30GHzARM Cortex-A15 r3p3 (4+1 Cores) @ 2.2GHz
GPUAdreno 430 @ 600MHzQualcomm Adreno 420 @ 600MHzQualcomm Adreno 330 @ 578MHzApple/PowerVR GX6450 x2 (8 clusters)PowerVR G6430Kepler (192 Core) @ 950MHzKepler (192 Core) @ 950MHz
Memory4GB LPDDR4 (25.6GB/s)3GB LPDDR3 (25.6GB/s)3 GB LPDDR3 (14.9GB/s)2GB LPDDR3 (25.6GB/s)2GB LPDDR3 (12.8GB/s)2GB LPDDR3 (14.9GB/s)2GB DDR3L (14.9GB/s)
Display10.1-inch @ 3840×2160 (436 PPI)5.96-inch AMOLED @ 2560x1440 (493 PPI)5.2-inch IPS @ 1920x1080 (423 PPI)9.7-inch IPS @ 2048x1536 (264 PPI)8.4-inch OLED @ 2560x1600 (359 PPI)8.9-inch IPS @ 2048x1536 (288 PPI)8-inch IPS @ 1920x1200 (283PPI)
Storage64 GB32 GB, 64 GB16 GB, microSD (up to 128 GB)16 GB, 64 GB, 128 GB16 GB, microSD (up to 512 GB)16 GB, 32 GB16GB (Wi-Fi) / 32GB (LTE), microSD (up to 128GB)
Operating SystemAndroid 5.0.2Android 5.0Android 4.4.4iOS 8.1.2Android 4.4.4Android 5.0.1Android 5.0.1

We should definitely see some interesting results considering how many different CPU and GPU architectures we're comparing. Will the 810 or Nvidia's unique Denver CPUs be able to outperform Apple's improved Cyclone CPUs? How close will the new Adreno 430 get to Nvidia's Kepler based GPU?

  • realjjj
    In Geekbench your result is by far the highest in the database , something is off there, it's overclocked or you are testing in a fridge or you got some new revision.What's certain is that something is way off.
    Reply
  • MobileEditor
    In Geekbench your result is by far the highest in the database , something is off there, it's overclocked or you are testing in a fridge or you got some new revision.What's certain is that something is way off.

    The 810 scores lower than both the Tegra K1 (Denver) and A8X in Geekbench single-core and 8% better than the A8X in multi-core. Looking at the table for the individual Integer test results shows that most of the 810's advantage is in the AES and SHA1 encryption tests, which AArch64 targets with new instructions, as I noted in the article.

    I used Geekbench 3 Pro v3.3.1 (as noted on the "Testing" page) and I definitely didn't test it in a fridge :) It was standing on a conference room table like shown in the picture on the "Testing" page.

    - Matt
    Reply
  • MrCommunistGen
    "With only half as many cores, both A8X and Tegra K1 (Denver) see their IPC advantage diminish in the multi-core tests"

    A8X is a tri-core CPU.
    Reply
  • airborn824
    This does not seem very promising at all. At this rate i am stuck with my S4 this year. I wont upgrade with such a small performance increase.
    Reply
  • MobileEditor
    "With only half as many cores, both A8X and Tegra K1 (Denver) see their IPC advantage diminish in the multi-core tests"

    A8X is a tri-core CPU.

    Doh! I was still thinking A8. That's what happens when writing at 4am with no sleep. I'll fix that.

    - Matt
    Reply
  • nebun
    tegra is a monster of a processor....nice numbers
    Reply
  • nebun
    battery life?????
    Reply
  • nebun
    how efficient is it?
    Reply
  • MobileEditor
    how efficient is it?

    If you're referring to Tegra X1, Nvidia claims between 5W-10W depending on application (tablet or car). We can't verify these claims since there aren't any shipping products yet and nobody outside of Nvidia has even been able to touch it.

    For the 810, we weren't given enough time to test battery life. We need to wait until products ship. With the 810 moving to 20nm and the fact that the Krait CPUs were pushed to their max frequency, I wouldn't be surprised to see the 810 use less power than the 805 for average tablet workloads. The Adreno 430 might use a little more power than 420 though.

    - Matt
    Reply
  • JeanLuc
    Are you checking to see if any of the devices are deliberately 'turboing/boosting' clockrates when certain benchmarks are run?
    Reply