Follow us
Share link
Submit benchmark
VS
Google Tensor G4 vs NVIDIA N1X
Last updated:
Benchmarks
Performance Overview
Average performance across multiple benchmarks
Single-core performance
Multi-core performance
Geekbench 6 Single-Core
Single-core performance
Google Tensor G4
59 %
8C / 8T · 3.10 GHz
NVIDIA N1X
100 %
20C / 20T · 4.05 GHz
Geekbench 6 Multi-Core
Multi-core performance
Google Tensor G4
25 %
8C / 8T · 3.10 GHz
NVIDIA N1X
100 %
20C / 20T · 1.50 GHz
iGPU - FP32 Performance (Single-precision GFLOPS)
Google Tensor G4
0 %
ARM Immortalis-G715 MP10
NVIDIA N1X
100 %
NVIDIA N1X (GB20B)
More benchmarks
Specifications
At a glance
|
Wischen
|
Google Tensor G4 | NVIDIA N1X |
|---|---|---|
| Family | Google Tensor (5) | NVIDIA N1 (1) |
| CPU group | Google Tensor G4 (1) | NVIDIA N1X (1) |
| Architecture | G4 | GB10 Grace / Blackwell |
| Technology | 4 nm | 3 nm |
| Segment | Smartphone / Tablet | Notebook |
| Socket | ||
| Predecessor | Google Tensor | |
| Successor | Google Tensor G5 |
CPU Cores and Base Frequency
|
Wischen
|
Google Tensor G4 | NVIDIA N1X |
|---|---|---|
| CPU Cores / Threads | 8 / 8 | 20 / 20 |
| Hyperthreading / SMT | ✗ | ✗ |
| Core architecture | hybrid (Prime / big.LITTLE) | hybrid (big.LITTLE) |
| Core Cluster 1: | 1x Cortex-X4 3.10 GHz |
10x Cortex-X925 1.50 - 4.05 GHz |
| Core Cluster 2: | 3x Cortex-A720 2.60 GHz |
10x Cortex-A725 1.80 - 2.80 GHz |
| Core Cluster 3: | 4x Cortex-A520 1.95 GHz |
|
| L2-Cache | 24.00 MB | |
| L3-Cache | 32.00 MB | |
| Overclocking | ✗ | ✗ |
Integrated graphics (iGPU)
|
Wischen
|
Google Tensor G4 | NVIDIA N1X |
|---|---|---|
| GPU name | ARM Immortalis-G715 MP10 | NVIDIA N1X (GB20B) |
| GPU frequency | 0.89 GHz | 1.05 - 2.42 GHz |
| CUs / Shader | 10 / | 48 / 6144 |
| Raytracing | ✗ | ✓ |
| Max. displays | 2 | 4 |
| Max. GPU Memory | 128 GB | |
| Technology | 4 nm | 3 nm |
| Release date | Q4/2022 | Q2/2026 |
NPU AI performance
|
Wischen
|
Google Tensor G4 | NVIDIA N1X |
|---|---|---|
| AI hardware | Google Tensor AI | NVIDIA Blackwell Tensor Engine |
| AI specifications | Google Edge TPU | 384 Tensor Kerne (5. Gen Blackwell) |
| NPU + CPU + iGPU | FP4: 1000 TOPS. FP8: 500 TOPS |
Memory & PCIe
|
Wischen
|
Google Tensor G4 | NVIDIA N1X |
|---|---|---|
| Memory | LPDDR5-6400 (51.2 GB/s) | LPDDR5X-9400 (301.0 GB/s) |
| Max. Memory | 16 GB | 128 GB |
| Memory channels | 4 | 4 |
| ECC | No | No |
| PCIe | 5.0 x 8 | |
| PCIe Bandwidth | 31.5 GB/s |
Thermal Management
|
Wischen
|
Google Tensor G4 | NVIDIA N1X |
|---|---|---|
| TDP | 10 W | 140 W |
| TDP (PL2) | ||
| TDP up | 240 W | |
| TDP down | ||
| T. junction max. |
Technical details
|
Wischen
|
Google Tensor G4 | NVIDIA N1X |
|---|---|---|
| Chip design | Chiplet | Monolithic |
| AES-NI | ✗ | ✓ |
| Operating systems | Android | Windows 11, Linux |
| Instruction set | Armv8.7-A (64 bit) | Armv9.2-A (64 bit) |
| ISA extensions | SVE2, SME, MTE, VIRT | |
| Release date | Q4/2024 | Q2/2026 |
| Release price | ||
| Documents |
User Reviews
Your rating
Here you can rate one or both of the processors. The average rating of Google Tensor G4 is 3.8 stars (175 ratings), while NVIDIA N1X has received 3.3 stars (4 ratings).
Click on the stars to rate one or both processors.
Most popular CPUs in both categories
1
Apple M5 Max (18-CPU 32-GPU)
18C / 18T · 4.61 GHz
2
Apple M5 Max (18-CPU 40-GPU)
18C / 18T · 4.61 GHz
3
Apple M2 Ultra (76-GPU)
24C / 24T · 3.50 GHz
4
Apple M1 Ultra (64-GPU)
20C / 20T · 3.20 GHz
5
Apple M1 Ultra (48-GPU)
20C / 20T · 3.20 GHz
6
Apple M4 (8-CPU)
8C / 8T · 4.46 GHz
7
Apple M2 Ultra (60-GPU)
24C / 24T · 3.50 GHz
8
Google Tensor G4
8C / 8T · 0.00 GHz
9
NVIDIA N1X
20C / 20T · 4.05 GHz
Popular Comparisons
