CPU & Memory Subsystem

As we mentioned earlier, the Snapdragon 845 is the first SoC employing ARM’s new cluster technology DynamiQ. The different CPU cores are no longer hosted in dedicated cluster subsystems but are now integrated in a larger DynamiQ cluster. This change means that the cluster cache which to date was referred to as the L2 cache becomes the L3 cache and the individual CPU cores now receive a new private per-core L2. In the case of the Snapdragon 845, the Kryo 385 performance cores – derivative of ARM’s new A75 – are configured with 256KB L2 caches. The Kryo 845 efficiency cores, which are derivative of the A55, use 128KB L2 cache configurations.

The L3 cache of the DynamiQ Shared Unit (DSU) is configured at 2MB. At the launch of the Snapdragon 845 Qualcomm advertised three voltage and clock domains – unfortunately we haven’t had time to look deeper into the system of the QRD to find out how this is partitioned, however it is still my unconfirmed belief that the third clock/voltage domain is dedicated to the DSU and not part of the CPU cores. An important characteristic that is totally new to the Android SoC ecosystem is the introduction of a system cache – this 3MB cache seems to sit at the memory controller / interconnect level above the CPU subsystem, something we’ll get back to in just a bit.

Qualcomm’s performance projections seemed relatively conservative as they claimed a performance uplift of only 25-30% which seemed lower than ARM’s projections. We have to keep in mind that beyond the microarchitectural improvements expected from the transition from A73 to A75 based CPU cores we also have a flat 14% frequency increase from 2.47GHz to 2.8GHz on the side of the performance cores. Unfortunately in the limited testing time we had with the QRD we couldn’t make use of long-running CPU benchmarks such as our SPEC suite, so for the scope of this article we had to base our synthetic analysis on GeekBench4 results.

Geekbench 4 - Integer Performance
Single Threaded
  Snapdragon 845 Snapdragon 835 % Increase
 
AES 1160 MB/s 942.5 MB/s 23.1%
LZMA 4.15 MB/s 2.98 MB/s 39.3%
JPEG 20.8 Mpixels/s 16.6 Mpixels/s 25.2%
Canny 32.1 Mpixels/s 24.9 Mpixels/s 28.8%
Lua 2.18 MB/s 1.75 MB/s 24.0%
Dijkstra 1.90 MTE/s 1.62 MTE/s 16.9%
SQLite 70.3 Krows/s 53.4 Krows/s 31.8%
HTML5 Parse 12.9 MB/s 8.97 MB/s 44.1%
HTML5 DOM 3.00 Melems/s 2.27 Melems/s 31.9%
Histogram Equalization 67.3 Mpixels/s 52.5 Mpixels/s 28.2%
PDF Rendering 66.4 Mpixels/s 48.5 Mpixels/s 37.0%
LLVM 321.2 functions/s 257.3 functions/s 24.8%
Camera 7.96 images/s 5.64 images/s 40.9%

For the integer workload results we see a healthy performance across the various tests. Qualcomm’s 25-30% increase here seems to be justified as this is the most common increase in most tests. Workloads such as LZMA, HTML5 parsing, PDF rendering and the Camera substests see larger increases into the 40% range. The overall improvement in absolute performance for the integer tests is 31%.

Geekbench 4  (Single Threaded) Integer Score/MHz

If we revisit performance per clock across recent microarchitectures we see the Snapdragon’s A75 based cores increase by only a meagre 15% which is below our expectations. We move on to the floating point benchmarks to see if we see a similar story.

Geekbench 4 - Floating Point Performance
Single Threaded
  Snapdragon 845 Snapdragon 835 % Increase
SGEMM 16.6 GFLOPS 11.4 GFLOPS 45.1%
SFFT 4.23 GFLOPS 2.86 GFLOPS 47.9%
N-Body Physics 1400 Kpairs/s 872.2 Kpairs/s 60.5%
Rigid Body Physics 8524.2 FPS 6130.5 FPS 39.0%
Ray Tracing 354.0 Kpixels/s 232.7 Kpixels/s 52.1%
HDR 11.9 Mpixels/s 8.31 Mpixels/s 43.2%
Gaussian Blur 34.5 Mpixels/s 23.9 Mpixels/s 44.3%
Speech Recognition 17.9 Words/s 13.6 Words/s 31.6%
Face Detection 752.4 Ksubs/s 532.8 Ksubs/s 41.2%

The FP subtests of GB4 show a noticeably larger increase than the integer tests. Besides the switch from a 2-wide decode front-end to a 3-wide one, the largest changes of the A75 microarchitecture was found in the floating point execution pipelines and is likely the cause for the larger FP performance improvement. The boost here comes at an overall 45% in GB4.

Geekbench 4 (Single Threaded) Floating Point Score/MHz

In terms of performance per clock, the 45% overall boost translates into a much larger 26% increase in IPC which is nearer to what we had expected.

Revisiting the performance claims from ARM’s TechDay release of the A75 we notice that we had been promised larger improvements such as up to a 34% increase in GB4 performance per clock, which I interpreted with the frequency increase of the Snapdragon 845 to result in a 52% overall increase, which did not materialise. I reached out to ARM on the topic and got back several points of consideration: The projections ARM published were made on a A75 simulation with 512KB L2 caches and 2MB L3. The L3 matches the configuration of the Snapdragon 845 however Qualcomm’s choice of going with smaller L2 caches will have a certain performance hit. ARM didn’t have a number at hand for GB4 but quotes a 2% performance degradation for SPEC2000, and claims for GB4 it should be lower. Another consideration point is the memory subsystem of the SoC which ARM can’t control but heavily impacts the performance of the CPU, so let’s have a look at that.

Running our internal memory benchmark on the QRD we see several expected characteristics of the Snapdragon 845: Compared to the Snapdragon 835’s A73 based cores we see the shift from shared cluster L2’s to private ones as well as the integration of sort of an L3 and L4 cache. The new L2 caches are very visible in our benchmark as memory latency up to the 256KB barrier (or rather, the 320KB barrier as the L1D and L2 are exclusive) sees a vast reduction compared to the L2 region of the A73 cores. The A75 cores promise 8-cycle hits for the L2 versus 19 cycles on the A73. In our test the difference is far larger as see a reduction from ~30ns down to ~4.5ns (not forgetting a clock frequency increase of the new cache). After the 256/320KB test size boundary we enter the DSU’s L3 cache. ARM describes the L3 as pseudo-exclusive so the outer boundary should end around or shortly after the 2048KB mark, the transition here is much harder to make out in the limited data we had time to collect so hopefully we’ll get to revisit it on a commercial device.

On the Snapdragon 835 the transition between L2 cache and DRAM is very sharp and visible in the graph. On the Snapdragon 845 however we see a far more gradient latency transition stretching out to up to the 5MB test depth. This is confirmation that Qualcomm’s system cache is indeed applied to the CPU subsystem and acts as an exclusive L4 cache to the processors. I think this new system cache is a true SoC-wide cache lying high up at the interconnect or memory controller level.

One of the worries of such a configuration for the CPU subsystem was increased latency to DRAM and it seems my fears were realised as the Snapdragon 845 shows a 30% increase in main memory latency from the CPU subsystem. Previously the Snapdragon 835 seemed to have by far one of the best memory controller implementations which directly resulted in higher performance of memory latency sensitive workloads. The latency increase in the 845 thus must be counteracting some of the microarchitectural improvements on part of the CPU cores. For GB4 in particular I made a remark that I didn’t notice any performance impact at all on the part of the Kirin 970’s memory latency, however we’re talking about different platforms and CPUs so I can’t be certain.

We reserve final conclusion on synthetic benchmarks until we get more time with a Snapdragon 845 device and able to investigate more and run SPEC. For now it looks like the Snapdragon 845 does not reach ARM’s projected performance levels, and falls well short of the claims. Among one of the other performance claims was Octane. We retired Octane some years ago and Google shortly followed up with official retirement, but as an added data-point the Snapdragon 845 reached a score of 15969 versus the Snapdragon 835’s 11879, also well short of a 20000 target that a projected 1.48x per clock performance increase would have resulted in.

Introduction System Performance
Comments Locked

140 Comments

View All Comments

  • generalako - Monday, February 12, 2018 - link

    No, Qualcomm actually didn't list those charts themselves. Qualcomm, on their own site even, stated during and after the launch, that the SD845 will provide "up to 25% performance improvement" over the SD835. That was with the stated clockspeed of 2.8 GHz.
  • Reflex - Monday, February 12, 2018 - link

    Um, you just confirmed Stormy's point. Re-read Stormy's comment and then your reply...
  • tuxRoller - Tuesday, February 13, 2018 - link

    The Qualcomm slide said 25-30% increase@2.8GHz with the smaller cores increasing their performance by 15%@1.8GHz.
    The next slides looks like its from arm's a75 announcement (no mention of Qualcomm brand names and the graph is the same one that arm uses) not the Qualcomm presentation.

    From the article:
    "The Kryo 385 gold/performance cluster runs at up to 2.8GHz, which is a 14% frequency increase over the 2.45GHz of the Snapdragon 835's CPU core. But we also have to remember that given that the new CPU cores are likely based on A75's we should be expecting IPC gains of up to 22-34% based on use-cases, bringing the overall expected performance improvement to 39-52%. Qualcomm promises a 25-30% increase which is at the low-end of ARM's projections.'

    The author speculates about perf based on the arm graphs & the freq increase that Qualcomm announced but Qualcomm themselves didn't suggest such numbers.
    My guess as to why integer IPC didn't increase by much is the power issue Andrei had alluded to before. Iow, the freq scaling provided enough perf so that Qualcomm didn't have to employ the more expensive changes that would've been required for the IPC gains.
  • mfaisalkemal - Monday, February 12, 2018 - link

    @andrei_frumansunu and @ryan_smith
    any date when we can read review from iphone 8 and iphone x?
    i just curious about long term performance manhattan 3.1 compared ipghone 7 and snapdragon 835.

    any chance to compare gpu of android and ios device with real world benchmark games Tainted Keep? nvidia use that game when tegra x1 launch.
  • Andrei Frumusanu - Monday, February 12, 2018 - link

    The A11 generally throttles 35% from its peak performance figures while the 835 maintains full or 90%. I'll include the iPhones in the new full review.

    We don't have any way to benchmark iOS devices in real games.
  • Andrei Frumusanu - Monday, February 12, 2018 - link

    > I'll include the iPhones in the new full review.

    And by that, I meant general next full device reviews, not specifically iPhone reviews.
  • mfaisalkemal - Monday, February 12, 2018 - link

    sorry not explain you tell little detail, tainted keep have in-game benchmark offscreen 1080p in normal and extreme mode.
    nvida tegra x1 ultra score: http://www.legitreviews.com/wp-content/uploads/201...
    iphone 7+ extreme score: https://i.imgur.com/3v5Tgxt.jpg
    ipad pro(a9x) extreme score: https://i.imgur.com/9pKrQBE.jpg

    thanks andrei, i'm eagerly waiting for the review :)
  • Andrei Frumusanu - Monday, February 12, 2018 - link

    The game doesn't even allow me to use the extreme settings on Qualcomm devices and in the normal benchmark it's just vsync limited at 60 fps - so I don't think we'll do anything with it.
  • mfaisalkemal - Monday, February 12, 2018 - link

    because driver or bug maybe, what a pity!
    how about test peak GFLOPS with opencl for GPU on next benchmark?
    i found this benchmark application named CLPeak on playstore.

    https://play.google.com/store/apps/details?id=kr.c...

    Oneplus 5 (Adreno 540)
    Single-precision compute (GFLOPS)
    float : 294.65
    float2 : 285.81
    float4 : 311.02
    float8 : 265.02
    float16 : 308.34

    half-precision compute (GFLOPS)
    half : 570.72
    half2 : 539.62
    half4 : 610.79
    half8 : 314.82
    half16 : 313.73

    source: https://forum.beyond3d.com/posts/2011570/

    with alu result from your test about 50% improvement, i think adreno 640 is not far from tegra X1 (512GFLOPs) in smartphone device!!!
  • darkd - Monday, February 12, 2018 - link

    Note that Ravn is a studio Nvidia has contracted things like this out to multiple times. As with most Nvidia contractors, they tend to have features/optimizations that are Nvidia only, and tend to be almost pointedly unoptimized for tiling GPUs. You can see something to this effect in the Ravn supplied benchmark for Antutu v7.

Log in

Don't have an account? Sign up now