Google has launched its latest flagship Pixel 9 series, featuring the Tensor G4 chipset. The Google Tensor G4 is said to be a small upgrade over last year’s Tensor G3. To evaluate its performance gains, we have benchmarked the Tensor G4 chipset on Geekbench, 3DMark, AnTuTu, and many other tests in this guide.
Exploring the Tensor G4 Specifications
The Tensor G4 chipset is designed with advanced technology to enhance the performance of Google's Pixel 9 series. It features an eight-core CPU architecture, which includes a combination of high-performance and efficiency cores. This design aims to balance power consumption with performance, making it suitable for various tasks.
With a clock speed of up to 3.10 GHz for the prime core, the Tensor G4 aims to deliver robust performance. However, it's essential to note that the actual performance can vary based on thermal management and other factors. The integrated Mali-G715 GPU, upgraded for better graphical performance, is also a significant aspect of this chipset.
Moreover, the Tensor G4 supports Wi-Fi 7 and Bluetooth 5.3 connectivity, ensuring that users can enjoy faster and more stable connections. The inclusion of the Samsung Exynos 5400 modem allows for impressive 5G capabilities, although real-world performance may fluctuate.
Key Specifications
| Specs | Tensor G4 |
|---|---|
| CPU | Eight-core CPU (1+3+4) |
| CPU Cores | 1x 3.10GHz (Cortex-X4), 3x 2.60GHz (Cortex-A720), 4x 1.95GHz (Cortex-A520) |
| Process Technology | Samsung’s 4nm |
| GPU | Mali-G715 MC7 GPU (Up to 940MHz) |
| Connectivity | Wi-Fi 7, Bluetooth 5.3 |
| Modem | Samsung Exynos 5400 5G modem (Up to 14.79Gbps Peak Download) |
Performance Insights: Geekbench & AnTuTu Benchmarks
In the Geekbench 6 CPU test, the Google Tensor G4 scored 1,897 in single-core tasks and 3,721 in multi-core tasks. These scores highlight a performance level that appears to be lower than many competitors, including the Snapdragon 7+ Gen 3, which is a mid-range processor.
We ran the Geekbench benchmark multiple times to ensure consistency in our findings. Unfortunately, the Tensor G4's multi-core score failed to cross the 4,000 mark, which is surprising given its advanced architecture and cooling features in the Pixel 9 Pro XL. The performance deficiency seems to stem from the limitations of Samsung's older 4nm process node used in the manufacturing.
Moreover, in the AnTuTu benchmark test, the Tensor G4 scored 1,253,407, which is also lower than expected for flagship hardware. The CPU scored just 337,131 points, while the GPU achieved 444,488 points, indicating that the graphical performance is also not at the level one might anticipate from a leading smartphone chipset.
Performance Breakdown
| Benchmark | Tensor G4 |
|---|---|
| Geekbench 6 (Single-core) | 1,897 |
| Geekbench 6 (Multi-core) | 3,721 |
| AnTuTu Score | 1,253,407 |
| AnTuTu CPU | 337,131 |
| AnTuTu GPU | 444,488 |
Thermal Management and CPU Throttling
During our testing, we observed a notable trend in CPU performance related to thermal management. The Pixel 9 series experienced throttling, particularly during extended workloads. In our tests, the CPU throttled to 71% of its maximum performance in one instance and dropped to just 42% in another. This indicates that the performance is compromised significantly under sustained usage.
The design choice to prioritize thermals over peak performance is evident. Although the chipset can reach high clock speeds, it appears that Google has opted to underclock the CPU to maintain battery life and device temperatures. This trade-off means that users might not experience the full potential of the Tensor G4 during demanding tasks.
For users looking for sustained performance in gaming or professional applications, this could be a limiting factor. The Pixel 9 series is built with cooling mechanisms, yet the performance still suffers under heavy load, primarily due to the conservative tuning of the chip.
Graphics Performance: 3DMark Wild Life Stress Test
When it comes to graphics performance, the Tensor G4's Mali-G715 GPU was put through the 3DMark Wild Life Extreme Stress test. The results revealed a maximum loop score of 2,590 and a minimum score of 1,525, which indicates a concerning level of instability, with a stability rating of only 58.9% during testing.
This stress test showcases the challenges the Tensor G4 faces in maintaining consistent performance, particularly under heavy graphical loads. The GPU's upgraded frequency from 890MHz to 940MHz shows an attempt at improvement, but the results suggest that it still falls short of expectations.
For gaming enthusiasts or content creators, the weak stability and performance metrics may be disappointing. The GPU's limitations could hinder the overall experience, especially when compared to other flagship devices that offer superior graphics capabilities.
AI and Machine Learning Performance
The Tensor G4 also features a dedicated TPU (Tensor Processing Unit) for AI and machine learning tasks. However, in our Geekbench AI benchmark, the TPU scored only 289 points in the Single Precision test, which pales in comparison to competitors like Apple's A17 Pro, which scored 3,930 in the same benchmark.
This underwhelming performance in AI tasks indicates that despite having a specialized chip for machine learning, the Tensor G4 might not be leveraging its capabilities effectively. The TPU's performance, alongside the CPU and GPU benchmarks, paints a picture of a chipset that struggles to keep up with the competition.
For users who rely on AI-driven features in photography, voice recognition, or other applications, the Tensor G4's performance may not meet expectations compared to rival offerings. This could affect user satisfaction for those looking for advanced AI functionalities in their devices.
5G Connectivity and Real-World Performance
In terms of connectivity, the new Samsung Exynos 5400 modem integrated into the Tensor G4 achieved peak download speeds of 174.8 Mbps and upload speeds of 35.1 Mbps during our tests. However, we noted significant inconsistencies in performance, with speeds dropping drastically at times.
The modem's performance can vary greatly depending on network conditions and location. While the theoretical speeds are impressive, real-world usage may not reflect these capabilities consistently. This inconsistency can be a drawback for users who depend on reliable 5G connectivity for their daily tasks.
In summary, while the Tensor G4 shows promise with its theoretical specifications, the real-world performance raises concerns. Users may find that the connectivity and overall performance don't align with the expectations set by flagship pricing.
Summing Up the Tensor G4 Experience
In conclusion, the Google Tensor G4 presents itself as a capable chipset on paper, with advanced features and specifications aimed at enhancing the Pixel 9 series. However, as our benchmarks indicate, the reality may not match the expectations that come with a flagship device.
The performance across various tests reveals limitations, especially in CPU and GPU capabilities, which could affect user experience. Furthermore, issues with thermal management and inconsistent 5G performance may leave some users wanting more from their devices.
Ultimately, while the Tensor G4 is a step forward, it may not be the leap that many were hoping for. As Google continues to develop its silicon, we anticipate improvements in future iterations to better compete in the ever-evolving smartphone market.
