Google Tensor vs HiSilicon Kirin 9000
Last updated:
CPU comparison with benchmarks
|
 |
|
| Google Tensor | HiSilicon Kirin 9000 | |
CPU comparisonGoogle Tensor or HiSilicon Kirin 9000 - which processor is faster? In this comparison we look at the differences and analyze which of these two CPUs is better. We compare the technical data and benchmark results.
The Google Tensor has 8 cores with 8 threads and clocks with a maximum frequency of 2.80 GHz. Up to 12 GB of memory is supported in 2 memory channels. The Google Tensor was released in Q4/2021. The HiSilicon Kirin 9000 has 8 cores with 8 threads and clocks with a maximum frequency of 3.13 GHz. The CPU supports up to GB of memory in 4 memory channels. The HiSilicon Kirin 9000 was released in Q4/2020. |
||
| Google Tensor (3) | Family | HiSilicon Kirin (29) |
| Google Tensor (1) | CPU group | HiSilicon Kirin 9000 (2) |
| 1 | Generation | 9 |
| G1 | Architecture | Cortex-A77 / Cortex-A55 |
| Mobile | Segment | Mobile |
| -- | Predecessor | -- |
| Google Tensor G2 | Successor | -- |
|
|
||
CPU Cores and Base FrequencyThe Google Tensor has 8 CPU cores and can calculate 8 threads in parallel. The clock frequency of the Google Tensor is 2.80 GHz while the HiSilicon Kirin 9000 has 8 CPU cores and 8 threads can calculate simultaneously. The clock frequency of the HiSilicon Kirin 9000 is at 3.13 GHz. |
||
| Google Tensor | Characteristic | HiSilicon Kirin 9000 |
| 8 | Cores | 8 |
| 8 | Threads | 8 |
| hybrid (Prime / big.LITTLE) | Core architecture | hybrid (Prime / big.LITTLE) |
| No | Hyperthreading | No |
| No | Overclocking ? | No |
| 2.80 GHz 2x Cortex-X1 |
A-Core | 3.13 GHz 1x Cortex-A77 |
| 2.25 GHz 2x Cortex-A76 |
B-Core | 2.54 GHz 3x Cortex-A77 |
| 1.80 GHz 4x Cortex-A55 |
C-Core | 2.05 GHz 4x Cortex-A55 |
Artificial Intelligence and Machine LearningProcessors with the support of artificial intelligence (AI) and machine learning (ML) can process many calculations, especially audio, image and video processing, much faster than classic processors. Algorithms for ML improve their performance the more data they have collected via software. ML tasks can be processed up to 10,000 times faster than with a classic processor. |
||
| Google Tensor | Characteristic | HiSilicon Kirin 9000 |
| Google Tensor AI | AI hardware | -- |
| Google Edge TPU @ 1.6 TOPS | AI specifications | -- |
Internal GraphicsThe Google Tensor or HiSilicon Kirin 9000 has integrated graphics, called iGPU for short. The iGPU uses the system's main memory as graphics memory and sits on the processor's die. |
||
| ARM Mali-G78 MP20 | GPU | ARM Mali-G78 MP24 |
| 0.76 GHz | GPU frequency | 0.76 GHz |
| -- | GPU (Turbo) | -- |
| Vallhall 2 | GPU Generation | Vallhall 2 |
| 5 nm | Technology | 5 nm |
| 1 | Max. displays | 1 |
| 20 | Compute units | 24 |
| 320 | Shader | 384 |
| No | Hardware Raytracing | No |
| No | Frame Generation | No |
| -- | Max. GPU Memory | -- |
| 12 | DirectX Version | 12 |
Hardware codec supportA photo or video codec that is accelerated in hardware can greatly accelerate the working speed of a processor and extend the battery life of notebooks or smartphones when playing videos. |
||
| ARM Mali-G78 MP20 | GPU | ARM Mali-G78 MP24 |
| Decode / Encode | Codec h265 / HEVC (8 bit) | Decode / Encode |
| Decode / Encode | Codec h265 / HEVC (10 bit) | Decode / Encode |
| Decode / Encode | Codec h264 | Decode / Encode |
| Decode / Encode | Codec VP9 | Decode / Encode |
| Decode / Encode | Codec VP8 | Decode / Encode |
| Decode | Codec AV1 | Decode |
| Decode / Encode | Codec AVC | Decode / Encode |
| Decode / Encode | Codec VC-1 | Decode / Encode |
| Decode / Encode | Codec JPEG | Decode / Encode |
Memory & PCIeThe Google Tensor can use up to 12 GB of memory in 2 memory channels. The maximum memory bandwidth is 53.0 GB/s. The HiSilicon Kirin 9000 supports up to GB of memory in 4 memory channels and achieves a memory bandwidth of up to --. |
||
| Google Tensor | Characteristic | HiSilicon Kirin 9000 |
| LPDDR5-5500 | Memory | LPDDR5-2750, LPDDR4X-2133 |
| 12 GB | Max. Memory | |
| 2 (Dual Channel) | Memory channels | 4 (Quad Channel) |
| 53.0 GB/s | Max. Bandwidth | -- |
| No | ECC | No |
| 8.00 MB | L2 Cache | -- |
| -- | L3 Cache | -- |
| -- | PCIe version | -- |
| -- | PCIe lanes | -- |
| -- | PCIe Bandwidth | -- |
Thermal ManagementThe thermal design power (TDP for short) of the Google Tensor is 10 W, while the HiSilicon Kirin 9000 has a TDP of --. The TDP specifies the necessary cooling solution that is required to cool the processor sufficiently. |
||
| Google Tensor | Characteristic | HiSilicon Kirin 9000 |
| 10 W | TDP (PL1 / PBP) | -- |
| -- | TDP (PL2) | -- |
| -- | TDP up | -- |
| -- | TDP down | -- |
| -- | Tjunction max. | -- |
Technical detailsThe Google Tensor is manufactured in 5 nm and has 8.00 MB cache. The HiSilicon Kirin 9000 is manufactured in 5 nm and has a 0.00 MB cache. |
||
| Google Tensor | Characteristic | HiSilicon Kirin 9000 |
| 5 nm | Technology | 5 nm |
| Unknown | Chip design | Chiplet |
| Armv8-A (64 bit) | Instruction set (ISA) | Armv8-A (64 bit) |
| -- | ISA extensions | -- |
| -- | Socket | -- |
| None | Virtualization | None |
| No | AES-NI | No |
| Android | Operating systems | Android |
| Q4/2021 | Release date | Q4/2020 |
| -- | Release price | -- |
| show more data | show more data | |
|
8C 8T @ 2.80 GHz buy now on Amazon and save!
8C 8T @ 3.13 GHz buy now on Amazon and save!
|
||
Rate these processors
Average performance in benchmarks
⌀ Single core performance in 1 CPU benchmarks
⌀ Multi core performance in 1 CPU benchmarks
Geekbench 5, 64bit (Single-Core)
Geekbench 5 is a cross plattform benchmark that heavily uses the systems memory. A fast memory will push the result a lot. The single-core test only uses one CPU core, the amount of cores or hyperthreading ability doesn't count.
|
Google Tensor
8C 8T @ 2.80 GHz |
||
|
HiSilicon Kirin 9000
8C 8T @ 3.13 GHz |
||
Geekbench 5, 64bit (Multi-Core)
Geekbench 5 is a cross plattform benchmark that heavily uses the systems memory. A fast memory will push the result a lot. The multi-core test involves all CPU cores and taks a big advantage of hyperthreading.
|
|
Google Tensor
8C 8T @ 2.80 GHz |
||
|
|
HiSilicon Kirin 9000
8C 8T @ 3.13 GHz |
||
iGPU - FP32 Performance (Single-precision GFLOPS)
The theoretical computing performance of the internal graphics unit of the processor with simple accuracy (32 bit) in GFLOPS. GFLOPS indicates how many billion floating point operations the iGPU can perform per second.
|
|
Google Tensor
ARM Mali-G78 MP20 @ 0.76 GHz |
||
|
|
HiSilicon Kirin 9000
ARM Mali-G78 MP24 @ 0.76 GHz |
||
AnTuTu 8 Benchmark
The AnTuTu 8 Benchmark measures the performance of a SoC. AnTuTu benchmarks the CPU, GPU, Memory as well as the UX (User Experience) by simulating browser and app usage. AnTuTu can benchmark any ARM CPU that runs under Android or iOS. Devices may not be directly compareable if the benchmark has been performed under different operating systems.
In the AnTuTu 8 benchmark, the single-core performance of a processor is only slightly weighted. The evaluation consists of the multi-core performance of the processor, the speed of the RAM and the performance of the internal graphics.
In the AnTuTu 8 benchmark, the single-core performance of a processor is only slightly weighted. The evaluation consists of the multi-core performance of the processor, the speed of the RAM and the performance of the internal graphics.
|
|
Google Tensor
8C 8T @ 2.80 GHz |
||
|
|
HiSilicon Kirin 9000
8C 8T @ 3.13 GHz |
||
Geekbench 6 (Single-Core)
Geekbench 6 is a benchmark for modern computers, notebooks and smartphones. What is new is an optimized utilization of newer CPU architectures, e.g. based on the big.LITTLE concept and combining CPU cores of different sizes. The single-core benchmark only evaluates the performance of the fastest CPU core, the number of CPU cores in a processor is irrelevant here.
|
|
Google Tensor
8C 8T @ 2.80 GHz |
||
|
|
HiSilicon Kirin 9000
8C 8T @ 3.13 GHz |
||
Geekbench 6 (Multi-Core)
Geekbench 6 is a benchmark for modern computers, notebooks and smartphones. What is new is an optimized utilization of newer CPU architectures, e.g. based on the big.LITTLE concept and combining CPU cores of different sizes. The multi-core benchmark evaluates the performance of all of the processor's CPU cores. Virtual thread improvements such as AMD SMT or Intel's Hyper-Threading have a positive impact on the benchmark result.
|
|
Google Tensor
8C 8T @ 2.80 GHz |
||
|
|
HiSilicon Kirin 9000
8C 8T @ 3.13 GHz |
||
AnTuTu 9 Benchmark
The AnTuTu 9 benchmark is very well suited to measuring the performance of a smartphone. AnTuTu 9 is quite heavy on 3D graphics and can now also use the "Metal" graphics interface. In AnTuTu, memory and UX (user experience) are also tested by simulating browser and app usage. AnTuTu version 9 can compare any ARM CPU running on Android or iOS. Devices may not be directly comparable when benchmarked on different operating systems.
In the AnTuTu 9 benchmark, the single-core performance of a processor is only slightly weighted. The rating is made up of the multi-core performance of the processor, the speed of the working memory, and the performance of the internal graphics.
In the AnTuTu 9 benchmark, the single-core performance of a processor is only slightly weighted. The rating is made up of the multi-core performance of the processor, the speed of the working memory, and the performance of the internal graphics.
|
|
Google Tensor
8C 8T @ 2.80 GHz |
||
|
|
HiSilicon Kirin 9000
8C 8T @ 3.13 GHz |
||
Performance for Artificial Intelligence (AI) and Machine Learning (ML)
Processors with the support of artificial intelligence (AI) and machine learning (ML) can process many calculations, especially audio, image and video processing, much faster than classic processors. The performance is given in the number (trillions) of arithmetic operations per second (TOPS).
|
|
Google Tensor
8C 8T @ 2.80 GHz |
||
|
|
HiSilicon Kirin 9000
8C 8T @ 3.13 GHz |
||
Devices using this processor |
|
| Google Tensor | HiSilicon Kirin 9000 |
| Google Pixel 6 Google Pixel 6 Pro |
Unknown |
Popular comparisons containing this CPUs
back to index
