Deutsch
Español
Français
Italiano
Nederlands
Polski
Português
Русский
Türkçe
Svenska
Chinese
MagyarExclusive | First NVIDIA Ampere mobile benchmarks: Up to 18% faster than Turing mobile, but don't expect Ampere desktop-class performance
NVIDIA will take the wraps off its mobile Ampere GPUs in a few hours from now, and we expect many OEMs to announce laptops powered by these new GPUs. While NVIDIA's performance benchmarks are still under embargo, we have managed to gather some initial performance numbers from pre-sample notebooks.
RTX 30 Ampere mobile synthetic benchmarks
Unlike the Turing generation, Ampere mobile cards have significantly reduced CUDA core counts compared to their desktop counterparts, the result of which can be seen in the comparison below.
It will be some time before we get to know gaming performance, but these 3DMark numbers should provide a fair preliminary assessment of Ampere mobile's capabilities.
In 3DMark 11, we see the RTX 3080 Mobile on par with the RTX 3070 desktop but is about 18% slower than an average RTX 3080 desktop. A similar performance deficit is seen when comparing the RTX 3070 Mobile with its corresponding desktop variant. However, the RTX 3060 Mobile seems to be on par with an average RTX 2060 Super desktop but is a healthy 41% faster than the RTX 2060 Mobile. Compared to the RTX 2080 Super Mobile, the RTX 3070 Mobile and the RTX 3080 Mobile are about 6% and 24% faster, respectively.
3DMark Time Spy GPU test shows the RTX 3080 Mobile having a lead of 17% over the average RTX 2080 Super Mobile and about 14% faster than the RTX 3070 Mobile. We see that the RTX 3080 Mobile is a significant 31% slower than its desktop variant. The RTX 3070 Mobile is about 3% faster than the RTX 2080 Super Mobile in this test.
Coming to 3DMark Fire Strike Standard GPU, the RTX 3080 Mobile is behind both average values of the desktop variants of RTX 3070 and RTX 3080 by 13% and 31%, respectively. The RTX 3060 Mobile trails the average RTX 2080 Super Mobile by about 10% but is also 4% faster than a desktop RTX 2060 Super.
| Performance Rating - Percent | |
| Radeon RX 6800 XT | |
| Average NVIDIA GeForce RTX 3080 | |
| Average NVIDIA GeForce RTX 3070 | |
| GeForce RTX 3080 Laptop GPU | |
| Average NVIDIA GeForce RTX 2080 Super (Desktop) | |
| GeForce RTX 3070 Laptop GPU | |
| GeForce RTX 2080 Super Mobile | |
| Average NVIDIA GeForce RTX 2080 Super Mobile | |
| Average NVIDIA GeForce RTX 2070 Super (Desktop) | |
| GeForce RTX 3060 Laptop GPU | |
| Average NVIDIA GeForce RTX 2060 Super (Desktop) | |
| Average NVIDIA GeForce RTX 2070 Super Mobile | |
| GeForce RTX 2070 Mobile | |
| Average NVIDIA GeForce RTX 2070 Mobile | |
| Radeon RX 5600M | |
| Average NVIDIA GeForce RTX 2060 Mobile | |
| GeForce RTX 2060 Mobile | |
| 3DMark 11 - 1280x720 Performance GPU | |
| Average NVIDIA GeForce RTX 3080 (47287 - 57662, n=4) | |
| Radeon RX 6800 XT | |
| GeForce RTX 3080 Laptop GPU | |
| Average NVIDIA GeForce RTX 3070 (26812 - 46878, n=9) | |
| Average NVIDIA GeForce RTX 2080 Super (Desktop) (37958 - 41532, n=4) | |
| GeForce RTX 3070 Laptop GPU | |
| Average NVIDIA GeForce RTX 2080 Super Mobile (34021 - 36838, n=4) | |
| Average NVIDIA GeForce RTX 2070 Super (Desktop) (33257 - 37235, n=7) | |
| GeForce RTX 2080 Super Mobile | |
| GeForce RTX 3060 Laptop GPU | |
| Average NVIDIA GeForce RTX 2060 Super (Desktop) (29148 - 30099, n=6) | |
| Average NVIDIA GeForce RTX 2070 Super Mobile (28207 - 31050, n=8) | |
| GeForce RTX 2070 Mobile | |
| Average NVIDIA GeForce RTX 2070 Mobile (23335 - 27298, n=21) | |
| Radeon RX 5600M | |
| Average NVIDIA GeForce RTX 2060 Mobile (19143 - 24662, n=40) | |
| GeForce RTX 2060 Mobile | |
Overall, these preliminary benchmarks indicate that we can expect Ampere mobile to offer anywhere between 7% to 18% improved performance compared to average Turing mobile (RTX 2080 Super Mobile) scores. The exact benefits will vary from one laptop model to the other as NVIDIA allows OEMs a wide TDP tuning range in both Max-P and Max-Q variants.
We will be getting our hands on Ampere mobile-powered laptops soon, so stay tuned for full-length reviews and deeper performance analysis.
Source(s)
Own
