Google Tensor vs A15 Bionic
We compared the 8-core Google Tensor (Mali-G78 MP20) with the older 6-core Apple A15 Bionic (Apple A15 GPU) SoC. Here you will find the pros and cons of each chip, technical specs, and comprehensive tests in benchmarks, like AnTuTu and Geekbench.
Review
CPU Performance
Single and multi-core processor tests
Gaming Performance
GPU performance in games and OpenCL/Vulkan
Battery life
Efficiency of battery consumption
NanoReview Score
Overall chip score
Key Differences
Pros of Google Tensor
- Supports 50% higher memory bandwidth (51.2 against 34.1 GB/s)
- Has 2 more cores
Pros of Apple A15 Bionic
- Higher GPU frequency (~57%)
- Performs 58% better in floating-point computations
- Shows better (up to 40%) AnTuTu 11 score β 1413K vs 1010K
- 15% higher CPU clock speed (3230 vs 2800 MHz)
- Better instruction set architecture
Benchmarks
Performance tests in popular benchmarksAnTuTu 11
| CPU | 365546 | 410975 |
| GPU | 169803 | 370837 |
| Memory | 178763 | 164151 |
| UX | 295913 | 467312 |
| Total score | 1010025 | 1413275 |
| Samples | [51 tests] | [69 tests] |
AnTuTu results from iOS and Android are not directly comparable! They won't affect NanoReview scores.
GeekBench 6
Single-Core Score
Multi-Core Score
| Asset compression | 159.8 MB/sec | 208.1 MB/sec |
| HTML 5 Browser | 97.5 pages/sec | 122.9 pages/sec |
| PDF Renderer | 138.9 Mpixels/sec | 139.1 Mpixels/sec |
| Image detection | 83.4 images/sec | 125.1 images/sec |
| HDR | 119.7 Mpixels/sec | 179.4 Mpixels/sec |
| Background blur | 14.1 images/sec | 22.3 images/sec |
| Photo processing | 30.2 images/sec | 53.2 images/sec |
| Ray tracing | 3.44 Mpixels/sec | 6.26 Mpixels/sec |
| Samples | 26 tests | 37 tests |
Compute Score (GPU)
| Background blur | 13.2 images/sec | 35.4 images/sec |
| Face detection | 7.31 images/sec | 44.8 images/sec |
| Horizon detection | 0.34 Gpixels/sec | 0.62 Gpixels/sec |
| Edge detection | 0.5 Gpixels/sec | 0.82 Gpixels/sec |
| Gaussian blur | 0.55 Gpixels/sec | 0.84 Gpixels/sec |
| Feature matching | 0.09 Gpixels/sec | 0.3 Gpixels/sec |
| Stereo matching | 39.4 Gpixels/sec | 71.1 Gpixels/sec |
| Particle physics | 1368.9 FPS | 2198.2 FPS |
| API | Vulkan | Metal |
| Samples | 11 tests | 19 tests |
Compute results use different APIs (Metal vs Vulkan/OpenCL) and are not directly comparable!
3DMark
3DMark Steel Nomad Light
| Score | 704 | - |
| Stability | 63% | - |
| Graphics test | 5 FPS | - |
3DMark Solar Bay
| Score | - | 3765 |
| Stability | - | 71% |
| Graphics test | - | 14 FPS |
| Section 1 | - | 19 FPS |
| Section 2 | - | 14 FPS |
| Section 3 | - | 9 FPS |
3DMark Wild Life Extreme
| Score | 1899 | 2854 |
| Stability | 59% | 76% |
| Graphics test | 11 FPS | 17 FPS |
PCMark 3.0
PCMark 3.0
| Web score | 7105 | - |
| Video editing score | 5884 | - |
| Data manipulation score | 10310 | - |
| Photo editing score | 19072 | - |
| Writing score | 15550 | - |
| Phones tested | 3 models | - |
Recent User Tests
Google Tensor
| Date | Benchmark | Result |
|---|---|---|
| π 2026-09-13 (VladiβοΈSlaveβοΈ) | AnTuTu 12 | 1281085 |
| π 2026-08-19 (duwi) | AnTuTu 11 | 774697 |
| π 2026-08-17 -> Paranoid _Antost | AnTuTu 11 | 930801 |
| π 2026-08-04 (nguyen van nguyen) | AnTuTu 11 | 990021 |
| π 2026-07-29 (Mimo) | AnTuTu 11 | 1117985 |
Apple A15 Bionic
| Date | Benchmark | Result |
|---|---|---|
| π 2026-09-23 -> HotPower | AnTuTu 11 | 1184759 |
| π 2026-09-19 -> HotPower | Geekbench 6 GPU | 21284 |
| π 2026-09-19 (Vova) | AnTuTu 11 | 1297615 |
| π 2026-09-19 -> HotPower | Geekbench 6 CPU | 2423 / 5985 |
| π 2026-09-18 (charlesddd) | AnTuTu 11 | 1461749 |
Specifications
Full list of technical specifications of Google Tensor and A15 BionicCPU
| Architecture | 2x 2.8 GHz β Cortex-X1 2x 2.25 GHz β Cortex A76 4x 1.8 GHz β Cortex A55 |
2x 3.23 GHz β Avalanche 4x 2.02 GHz β Blizzard |
| Cores | 8 | 6 |
| Frequency | 2800 MHz | 3230 MHz |
| Instruction set | ARMv8.2-A | ARMv8.5-A |
| L1 cache | - | 256 KB |
| L2 cache | 3 MB | 12 MB |
| L3 cache | 4 MB | - |
| Process | 5 nanometers | 5 nanometers |
| Transistor count | - | 15 billion |
| TDP (Sustained Power Limit) | 5.6 W | 6 W |
| Manufacturing | Samsung | TSMC |
Graphics
| GPU name | Mali-G78 MP20 | Apple A15 GPU |
| Architecture | Valhall 2nd gen | Apple Bionic GPU |
| GPU frequency | 850 MHz | 1338 MHz |
| Pipelines | 20 | 5 |
| Shading units | 32 | 128 |
| Total shaders | 640 | 640 |
| FLOPS | 1085.4 Gigaflops | 1712.6 Gigaflops |
| Vulkan version | 1.3 | - |
| OpenCL version | 2.0 | - |
AI Accelerator
| Neural processor (NPU) | Tensor Processing Unit | Apple Neural Engine |
| NPU theoretical performance | - | 15.8 TOPS |
Memory
| Memory type | LPDDR5 | LPDDR4X |
| Memory frequency | 3200 MHz | 2133 MHz |
| Bus | 4x 16 Bit | 4x 16 Bit |
| Max bandwidth | 51.2 Gb/s | 34.1 Gb/s |
| Max size | 12 GB | 6 GB |
Multimedia (ISP)
| Storage type | UFS 3.1 | NVMe |
| Max display resolution | 3840 x 2160 | 2732 x 2048 |
| Max camera resolution | 1x 200MP, 2x 32MP | - |
| Video capture | 4K at 60FPS | 4K at 60FPS |
| Video playback | 4K at 60FPS | 4K at 60FPS |
| Video codecs | - H.264 - H.265 - AV1 - VP9 |
- H.264 - H.265 - VP8 - VP9 - Motion JPEG |
| Audio codecs | - AAC - AIFF - CAF - MP3 - MP4 - WAV |
- AAC - AAX - AAX+ - AC-3 - AIFF - CAF - E-AC-3 - MP3 - MP4 - WAV |
Connectivity
| Modem | Exynos 5300b | - |
| 4G support | LTE Cat. 24 | LTE Cat. 24 |
| 5G support | Yes | Yes |
| Download speed | Up to 7350 Mbps | Up to 7500 Mbps |
| Upload speed | Up to 3670 Mbps | Up to 3000 Mbps |
| Wi-Fi | 6 | 6 |
| Bluetooth | 5.2 | 5.0 |
| Navigation | GPS, GLONASS, Beidou, Galileo, QZSS | GPS, GLONASS, Beidou, Galileo, QZSS |
Info
| Announced | October 2021 | September 2021 |
| Class | Flagship | Flagship |
| Model number | GS101 (S5P9845) | APL1W07 |
| Official page | Google Tensor official site | - |
| Successor | - Google Tensor G5 | - Apple A19 Pro |
Cast your vote
Total votes: 2028
