Early Google Pixel 4a performance review compares the 2020 mid-range Pixel to the Pixel 4, Pixel 3a, and Pixel 3 XL
If there wasn’t a global pandemic to worry about, then Google would have held its annual developer conference, Google I/O, in Mountain View, California this week. At last year’s I/O, Google unveiled its first-ever mid-range Pixel devices, the Pixel 3a and 3a XL. This year, we’re expecting Google to follow-up with a new mid-range Pixel smartphone called the Pixel 4a. We don’t know when exactly this 2020 mid-range Pixel will be announced, but thanks to leaks, we know pretty much all there is to know about it. Ahead of the Google Pixel 4a announcement, we can detail the phone’s performance thanks to benchmarks performed on pre-release hardware.
To date, the most substantive leaks of the Pixel 4a have come from Cuban YouTuber Julio Lusson who runs the TecnoLike Plus channel. Last week, he shared multiple photos he took from his pre-release Pixel 4a, giving us an early look at the camera performance of Google’s 2020 mid-range Pixel. Today, he has published a video on his YouTube channel showcasing the performance of this upcoming smartphone in various benchmarking applications. He shared the raw results with us before publishing the video, allowing us to compile and analyze the data into the below tables and charts. Take note that since he performed these benchmarks on pre-release hardware running pre-release software, there’s a chance that the retail units may perform slightly better in benchmarks if Google has further optimized the performance.
Julio’s video below shows him running the Pixel 4a through multiple benchmarks, and it even offers a few glimpses at games like The Legend of Zelda: The Wind Waker for the Nintendo GameCube (via Dolphin Emulator) and PUBG Mobile running on the device. The video is in Spanish, though, so keep reading below if you’re interested in the benchmark results we put together.
Test Devices – Pixel 4a, 4, 3 XL, 3a XL, and the latest QRD
The Google Pixel 4a is powered by the Qualcomm Snapdragon 730 mobile platform, which is fabricated by Samsung using an 8nm LPP process. The Snapdragon 730 has an octa-core CPU consisting of 2 clusters: 2 ARM Cortex-A76-based CPU cores clocked at up to 2.2GHz and 6 ARM Cortex-A55-based CPU cores clocked at up to 1.8GHz. The GPU is Qualcomm’s Adreno 618.
The Snapdragon 730 is no longer Qualcomm’s best mid-range processor as that crown goes to the Snapdragon 765, but the 730 is still a significant jump up from the Snapdragon 670 found in the Pixel 3a and 3a XL. We can expect noticeable bumps in CPU performance because of the difference in the architecture between the ‘Performance’ CPU cores. We can also expect the Snapdragon 730 in the Pixel 4a to be more power-efficient than the Snapdragon 670 in the Pixel 3a because of the more modern manufacturing process, though our benchmarks won’t reflect this improvement. In terms of real-world performance, the Google Pixel 4a should noticeably outperform the Pixel 3a because of the better processor, better GPU, higher memory capacity, and faster storage technology.
For good measure, we also added benchmark results from the Google Pixel 4 (powered by the Qualcomm Snapdragon 855), the Google Pixel 3 XL (powered by the Qualcomm Snapdragon 845), and the latest Qualcomm Reference Device (powered by the Qualcomm Snapdragon 865). We collected most of these benchmark results back in December when we benchmarked the Snapdragon 865. We’re fairly certain the upcoming Google Pixel 5 won’t feature the Snapdragon 865, but we thought it would still be interesting to see how much of a performance gap there is between Google’s 2020 mid-range Pixel and the best hardware you can find on any Android device. The Pixel 5 is instead expected to feature the Snapdragon 765, but we don’t have a device with this processor at our disposal yet since few devices with this platform are available outside of China.
|Qualcomm Reference Device (QRD)||Google Pixel 4||Google Pixel 3 XL||Google Pixel 4a||Google Pixel 3a XL|
|Device Name||Qualcomm Snapdragon 865||Qualcomm Snapdragon 855||Qualcomm Snapdragon 845||Qualcomm Snapdragon 730||Qualcomm Snapdragon 670|
|Software||Android 10||Android 10||Android 10||Android 10||Android 10|
|Display||2880×1440 @ 60Hz||2280×1080 @ 60Hz||2960×1440 @ 60Hz||2340×1080 @ 60Hz||2160×1080 @ 60Hz|
|Memory||12GB LPDDR5||6GB LPDDR4X||4GB LPDDR4X||6GB LPDDR4X||4GB LPDDR4X|
|Storage||128GB UFS 3.0||64GB UFS 2.1||64GB UFS 2.1||64GB UFS 2.1||64GB eMMC 5.1|
Here is an overview of the specifications for the Qualcomm Snapdragon 865, Qualcomm Snapdragon 855, Qualcomm Snapdragon 845, Qualcomm Snapdragon 730, and Qualcomm Snapdragon 670.
Qualcomm Snapdragon 730, 865, 855, 845, and 670 Specifications
|Qualcomm Snapdragon 865||Qualcomm Snapdragon 855||Qualcomm Snapdragon 845||Qualcomm Snapdragon 730||Qualcomm Snapdragon 670|
|GPU||Adreno 650||Adreno 640||Adreno 630||Adreno 618||Adreno 615|
|Memory||4x 16bit, 2133MHz LPDDR4X|
4x 16bit, 2750MHz LPDDR5
|4x 16bit, 2133MHz LPDDR4X||4x 16-bit, 1866MHz LPDDR4X||2x 16-bit, 1866MHz LPDDR4X||2x 16-bit, 1866MHz LPDDR4X|
|Manufacturing Process||7nm (TSMC N7P)||7nm (TSMC)||10nm LPP (Samsung)||8nm LPP (Samsung)||10nm LPP (Samsung)|
Quick Overview of Each Benchmark
With input from Mario Serrafero
- AndroBench: AndroBench is a fairly old benchmark with an equally dated design, but it’s still the go-to for storage testing. It tests the speed of sequential read/write, random read/write, and SQLite insert, update, and delete operations. A sequential read/write is an operation that involves reading/writing storage blocks that are contiguous, while a random read/write involves reading/writing randomly scattered storage blocks. SQLite describes a type of database management system; developers dealing with large databases often have to make SQLite calls to retrieve or modify the database. We can get a good idea of the storage performance of an Android device with AndroBench. By default, the benchmark writes a 64MP file with either 32MB or 4KB buffer sizes for sequential and random read/writes respectively, and an SQLite transaction size of 1. The speed of the former operation is measured in MB/s while the latter in Queries Per Second (QPS).
- AnTuTu: This is a holistic benchmark. AnTuTu tests the CPU, GPU, and memory performance, while including both abstract tests and, as of late, relatable user experience simulations (for example, the subtest which involves scrolling through a ListView). The final score is weighted according to the designer’s considerations.
- GeekBench: A CPU-centric test that uses several computational workloads including encryption, compression (text and images), rendering, physics simulations, computer vision, ray tracing, speech recognition, and convolutional neural network inference on images. The score breakdown gives specific metrics. The final score is weighted according to the designer’s considerations, placing a large emphasis on integer performance (65%), then float performance (30%), and finally, crypto (5%).
- GFXBench: Aims to simulate video game graphics rendering using the latest APIs. Lots of onscreen effects and high-quality textures. Newer tests use Vulkan while legacy tests use OpenGL ES 3.1. The outputs are frames during test and frames per second (the other number divided by the test length, essentially), instead of a weighted score. Aztec Ruins: These tests are the most computationally heavy ones offered by GFXBench. Currently, top mobile chipsets cannot sustain 30 frames per second. Specifically, the test offers really high polygon count geometry, hardware tessellation, high-resolution textures, global illumination and plenty of shadow mapping, copious particle effects, as well as bloom and depth of field effects. Most of these techniques will stress the shader compute capabilities of the processor.
- PCMark 2.0: Tests the device as a complete unit. It simulates everyday use cases that can implement abstract algorithms and a lot of arithmetic; the difference is that these are dispatched within an application environment, with a particular practical purpose, and handled by API calls and Android libraries common to multiple applications. The test will output a variety of scores corresponding to the various subtests, which will be detailed below; the composite, Work 2.0 score is simply the geometric mean of all of these scores, meaning all tests are weighted equally.
PCMark 2.0 Subscore Explanations. Click to expand.
- Web browsing 2.0 simulates browsing social media: rendering the web page, searching for the content, re-rendering the page as new images are added, and so on. This subtest uses the native Android WebView to render (WebKit) and interact with the content, which is locally stored — this means you can run it offline, but it does not simulate web browsing fully as it rules out internet connection factors (latency, network speed). It is specifically tracking frame rates and completion time across seven tasks, with their score being a multiple of their geometric mean.
- Video Editing simulates video editing performance: applying effects to a video using OpenGL ES 2.0 fragment shaders, decoding video frames (sent to an Android GLSurfaceView), and rendering/encoding the video in H.264/MPEG-4AVC at several frame rates and resolutions up to 4K. It is specifically tracking frame rates on the UI, except for a final test tracking the completion time of a video editing pipeline.
- Writing simulates general document and text editing work: adding or editing texts and images within a document, copying and pasting text, and so on. It uses the native Android EditText view as well as PdfRenderer and PdfDocument APIs. It will open compressed documents, move text bodies, insert images in the document, then save them as a PDF, to then encrypt and decrypt them (AES). It specifically tracks task completion times for the processes of opening and saving files, adding images and moving text bodies, encrypt/decrypt the file, and render the PDF pages on ImageViews.
- Photo Editing simulates photo-editing performance: opening images, applying different effects via filters (grains, blurs, embossing, sharpening, and so on) and saving the image. It uses 4MP JPEG source images and manipulates them in bitmap format using the android.media.effect API, android.renderscript API’s RenderScript Intrinsics, android-jhlabs, and the native android.graphics API for drawing the process on the screen. This is an extremely comprehensive test in that it will be impacted by storage access, CPU performance, GPU performance, and it is dependent on many different Android APIs. The test specifically measures memory and storage access times, encoding and decoding times, task completion times. The various filters and effects come from different APIs.
- Data manipulation simulates database management operations: parsing and validating data from files, interacting with charts, and so on. It will open (date, value) tuples from CSV, XML, JSON files, and then render animated charts with the MPAndroidChart library. It specifically tracks data parsing times as well as draws per second of each chart animation (similar to frame rate, but specific to the updating chart).
AnTuTu isn’t my preferred benchmark, especially after it got booted off the Play Store, but there’s no denying it’s one of the most popular benchmarks for Android devices. In this test, the Pixel 4a scores overall about ~48% as high as the Qualcomm Reference Device and ~70% as high as the Pixel 4, but it does about as well as the Pixel 3 XL and substantially better than the Pixel 3a XL. When we looked at AnTuTu’s subscores, we can see that the Pixel 4a scores fairly well in the CPU, Memory, and UX tests but falls significantly behind all the Snapdragon 8-series processors we tested when it comes to the GPU tests. In fact, the Pixel 4a outperformed the Pixel 3 XL in most of AnTuTu’s tests except for the GPU ones, where the Pixel 4a performed 50% to 60% as well as the Pixel 3 XL. The Pixel 4a’s performance in AnTuTu’s Memory tests is about on par with that of the Pixel 4 and Pixel 3 XL—no surprise, though, since all of these devices have similar memory configurations. The Pixel 4a’s overall UX score in AnTuTu is also about on par with that of the Pixel 4, but it’s about 35-36% higher than the scores for the Pixel 3 XL and Pixel 3a XL. Out of all of these devices, however, the Pixel 4 will still offer the best real-world UI performance since we can’t discount the fact that it’s the only Pixel device with a 90Hz refresh rate panel. Overall, the Pixel 4a outperforms the Pixel 3a XL in nearly every test in AnTuTu while it unsurprisingly underperforms the QRD in every test.
PCMark is one of my favorite benchmarks because of its emphasis on real-world performance. While there’s a big gap in the scores between the Pixel 4a and the Pixel 3a XL, there’s a much smaller gap in the scores between the former device and the Pixel 4 or Pixel 3 XL. The QRD predictably blows the competition out of the water because of its superior hardware across the board, so it’s not even worth analyzing its subscores. The Pixel 4a’s subscores for PCMark’s Writing 2.0 and Photo Editing 2.0 tests are much higher (44% and 56% respectively) than the Pixel 3a XL’s, which is good news for users looking to perform basic document and image editing tasks on their device.
Julio was only able to share a single result from GFXBench with us, but it clearly demonstrates what I was talking about earlier in regards to the GPU performance: The Adreno 618 in the Snapdragon 730 is majorly outclassed by the Adreno GPUs in Qualcomm’s Snapdragon 8-series. GFXBench’s Aztec Ruins test is Kishonti’s most computationally intensive graphics test by far but it’s not representative of most Android games, so don’t let these results discourage you from trying to play games on the Google Pixel 4a. I would bet that most games will run just fine on the device, even the notoriously performance-intensive ones like Fortnite Mobile, provided you’re willing to turn down some of the graphical settings. For what it’s worth, Julio told me that he played a round of PUBG Mobile just fine at the game’s “High” graphics preset.
If you’re into retro gaming via emulators, though, then you’ll be plenty happy with the Google Pixel 4a’s CPU performance. While the Pixel 4a’s multi-core Geekbench 5.0 scores are significantly lower than that of the Pixel 4 and Pixel 3 XL (again, the QRD is so far ahead it’s not even worth discussing), the Pixel 4a’s single-core score is much closer to the scores from the Pixel 4 and Pixel 3a XL. In a single-core score comparison, the Pixel 4a actually outperforms the Pixel 3 XL, which we can likely attribute to the fact that the Snapdragon 730’s performance cluster is comprised of 2 ARM Cortex-A76-based CPU cores compared to the Snapdragon 845’s older ARM Cortex-A75-based CPU cores. When looking at the Geekbench 5.0 subscores for the multi-core scores, we can see that the Pixel 4a’s Snapdragon 730 mostly underperforms the Pixel 4’s Snapdragon 855 and Pixel 3 XL’s Snapdragon 845 when it comes to cryptographic calculations.
Here’s a table summarizing the overall benchmark scores charted on these bar graphs. To keep this table from becoming too massive, we didn’t include the subscores for each test—if you’re interested in seeing those results, then feel free to reach out to me.
|Benchmark||Version||QRD – Snapdragon 865||Google Pixel 4 – Snapdragon 855||Google Pixel 3 XL – Snapdragon 845||Google Pixel 4a – Snapdragon 730||Google Pixel 3a XL – Snapdragon 670|
|AnTuTu||8.0.4 & 8.3.2||565,384||386,499||278,647||268,973||192,779|
|GFXBench 1440p Aztec Ruins OpenGL (High Tier) Offscreen IFH||5.00||20||16||14||6||4.5|
|PCMark – Work 2.0||2.0.3716||12,626||9,311||8,988||8,687||6,881|
|Androbench Sequential Read (MB/s)||5.0.1||1,459||873||659||509||301|
|Androbench Sequential Write (MB/s)||5.0.1||225||189||231||188||237|
|Androbench Random Read (IOPS)||5.0.1||50,378||37,600||32,376||33,422||16,226|
|Androbench Random Write (IOPS)||5.0.1||48,410||41,340||37,417||39,053||25,522|
|Androbench Random Read (MB/s)||5.0.1||195||147||126||131||63|
|Androbench Random Write (MB/s)||5.0.1||189||161||146||153||100|
|Androbench SQLite Insert||5.0.1||3,705||3,207||2,627||1,914||1,712|
|Androbench SQLite Update||5.0.1||4,014||3,996||3,333||2,458||2,080|
|Androbench SQLite Delete||5.0.1||5,037||4,558||4,081||2,826||2,471|
Conclusion – The Google Pixel 4a is a solidly performing mid-range smartphone
The Google Pixel 4a is shaping up to be an excellent mid-range smartphone. If the rumored $399 starting price is correct, that would make the Pixel 4a a direct competitor to the Apple iPhone SE (2020) and Samsung Galaxy A51. While Google may not be packing the greatest hardware in this device, we can be sure that the phone will be capable of taking excellent photos. Google’s software will also be a huge advantage over other mid-range offerings. While some companies were basically goaded into committing to just providing 2 years of Android updates, Google offers 3 years on all Pixel devices. What’s more is that the Pixel 4a will feature most Pixel software features that have been announced already, including Now Playing, Live Caption, and the new Google Assistant (which Julio confirmed to us is supported). All that’s left is for Google to just announce the phone already, but we may have to wait until June 3rd for that to happen.
Rumored Google Pixel 4a Specifications
- Processor: Qualcomm Snapdragon 730
- GPU: Adreno 618
- RAM: 6GB LPDDR4X
- Internal storage: 64GB UFS 2.1
- Display: Single hole-punch 5.81-inch display, 2,340 x 1,080 resolution, 443 dpi, 60Hz refresh rate
- Rear camera: 12.2 MP Sony IMX363, f/1.73 aperture, 1.4µm pixels, OIS, EIS, LED flash, 4K video recording, Autofocus
- Front camera: 8.0 MP Sony IMX355, f/2.0 aperture, 1.14µm pixels, EIS, Fixed focus
- Connectivity: 4G, Dual SIM, GPS, WiFi 5, Bluetooth, GLONASS, NFC
- Ports: USB Type-C, 3.5mm headphone jack
- Security: Rear fingerprint sensor
- Battery : 3,080 mAh
- Software: Android 10