NVIDIA GeForce RTX 2070 vs NVIDIA GeForce RTX 3070 Mobile Comparison

The NVIDIA GeForce RTX 3070 Mobile and the NVIDIA GeForce RTX 2070 represent two distinct generations of NVIDIA’s laptop and desktop graphics, respectively. The data shows a clear generational shift, with the RTX 3070 Mobile winning 7 of 10 head-to-head benchmarks, while the RTX 2070 retains wins in 3 legacy-oriented tests. This analysis breaks down the benchmark results, architectural differences, and use-case scenarios strictly from the provided data.

Head-to-Head Benchmarks

The most decisive victory for the RTX 3070 Mobile comes in the 3dmark_3dmark_steel_nomad_dx12 test, where it scores 2380 versus the RTX 2070’s 1569. That is a 51.7% delta, the largest margin in the entire comparison. This result indicates a substantial advantage in modern DirectX 12 workloads, which typically scale with raw compute and newer architectural features.

The RTX 3070 Mobile also leads in geekbench_opencl, scoring 92939 against 79966, a 16.2% advantage. Similarly, in geekbench_vulkan it wins with 86768 versus 82521, a narrower 5.1% margin. These results suggest the Ampere-based mobile chip holds a solid lead in general-purpose compute and Vulkan API performance.

In the Passmark suite, the RTX 3070 Mobile wins four tests: passmark_directx_10 (113 vs 111, +1.8%), passmark_directx_11 (138 vs 129, +7.0%), passmark_directx_12 (64 vs 60, +6.7%), and passmark_gpu_compute (6827 vs 6411, +6.5%). While these margins are modest, they show consistent superiority across compute and modern graphics APIs.

However, the RTX 2070 strikes back in three tests. The largest reversal is in passmark_directx_9, where it scores 211 versus 160, a 24.2% lead for the older card. It also wins passmark_g2d (819 vs 641, +21.7%), a 2D graphics test, and passmark_g3d (16094 vs 15309, +4.9%). These wins are notable because they suggest the RTX 2070 retains an edge in legacy DirectX 9 workloads and certain 3D rendering scenarios, despite being an older architecture.

Overall, the RTX 3070 Mobile’s average benchmark score is 20534, placing it in the 65th percentile of all GPUs. The RTX 2070 averages 18789, sitting in the 63rd percentile. The 9.3% gap in average score aligns with the RTX 3070 Mobile’s wins in the more modern and compute-heavy tests.

Architecture Differences

The two GPUs are built on fundamentally different architectures and process nodes. The RTX 3070 Mobile uses the GA104 chip on Ampere architecture, fabricated on an 8 nm process by Samsung. The RTX 2070 uses the TU106 chip on Turing architecture, built on a 12 nm process by TSMC. This process shrink allows the RTX 3070 Mobile to pack 17,400 million transistors into a 392 mm² die, yielding a transistor density of 44.4M / mm². The RTX 2070 has 10,800 million transistors on a larger 445 mm² die, with a density of just 24.3M / mm².

The Ampere chip also doubles down on compute resources. The RTX 3070 Mobile has 5120 shading units, 160 TMUs, and 80 ROPs, compared to the RTX 2070’s 2304 shading units, 144 TMUs, and 64 ROPs. This results in significantly higher theoretical throughput: 15.97 TFLOPS FP32 for the mobile chip versus 7.465 TFLOPS for the desktop card. Interestingly, the RTX 2070 has more tensor cores (288 vs 160), but the RTX 3070 Mobile has more RT cores (40 vs 36). The RTX 2070’s FP16 performance is listed as 14.93 TFLOPS (2:1), while the RTX 3070 Mobile achieves 15.97 TFLOPS (1:1).

Both cards share 8 GB of GDDR6 memory on a 256-bit bus, giving them identical 448.0 GB/s bandwidth. Their memory clocks are also the same at 1750 MHz (14 Gbps effective). However, the RTX 3070 Mobile operates with a lower base clock of 1110 MHz and boost clock of 1560 MHz, while the RTX 2070 runs at 1410 MHz base and 1620 MHz boost. Despite lower clocks, the mobile chip’s massive shading unit count more than compensates.

Power and interface differences are stark. The RTX 3070 Mobile is rated at 115 W TDP with no power connectors, while the RTX 2070 draws 175 W and requires a 1x 8-pin connector, with a suggested PSU of 450 W. The mobile chip uses PCIe 4.0 x16, while the desktop card is limited to PCIe 3.0 x16. The RTX 2070 also includes display outputs (1x DVI, 1x HDMI 2.0, 2x DisplayPort 1.4a, 1x USB Type-C) and is a dual-slot card sized 229 mm long, whereas the RTX 3070 Mobile’s outputs are described as "Portable Device Dependent."

Where Each One Wins

Based on the benchmark data, the RTX 3070 Mobile is the clear winner for modern and compute-intensive workloads. Its 51.7% lead in 3DMark Steel Nomad DX12 makes it the better choice for contemporary DirectX 12 gaming and ray-traced titles. The 16.2% OpenCL advantage and 6.5% compute lead further cement its position for GPU-accelerated tasks like rendering, machine learning inference, and scientific computing. Vulkan performance is also in its favor, albeit by a smaller 5.1% margin.

The RTX 2070, despite being older and slower overall, wins in specific legacy scenarios. Its 24.2% lead in DirectX 9 suggests it is better suited for older games or applications that rely on DX9 rendering paths. The 21.7% win in 2D tests (passmark_g2d) indicates superior performance in desktop compositing, image processing, and other 2D-accelerated tasks. The 4.9% win in passmark_g3d is more puzzling, but it could point to certain OpenGL-based 3D workloads where Turing’s driver optimizations or clock speeds give it an edge.

For raw gaming performance, the RTX 3070 Mobile’s wins in DX11 (+7.0%) and DX12 (+6.7%) are more relevant than the RTX 2070’s DX9 win, as most modern titles use these newer APIs. The RTX 2070’s higher boost clock (1620 MHz vs 1560 MHz) may explain its win in the older passmark_g3d test, where single-threaded clock speed matters more than raw shader count.

Specification Differences

The key specification differences between the two cards are as follows:

• Process Node: RTX 3070 Mobile is 8 nm (Samsung); RTX 2070 is 12 nm (TSMC).
• Transistors: 17,400 million vs 10,800 million.
• Die Size: 392 mm² vs 445 mm².
• Transistor Density: 44.4M / mm² vs 24.3M / mm².
• Base Clock: 1110 MHz vs 1410 MHz.
• Boost Clock: 1560 MHz vs 1620 MHz.
• Shading Units: 5120 vs 2304.
• TMUs: 160 vs 144.
• ROPs: 80 vs 64.
• RT Cores: 40 vs 36.
• Tensor Cores: 160 vs 288.
• Pixel Rate: 124.8 GPixel/s vs 103.7 GPixel/s.
• Texture Rate: 249.6 GTexel/s vs 233.3 GTexel/s.
• FP32 Performance: 15.97 TFLOPS vs 7.465 TFLOPS.
• FP16 Performance: 15.97 TFLOPS (1:1) vs 14.93 TFLOPS (2:1).
• TDP: 115 W vs 175 W.
• Power Connectors: None vs 1x 8-pin.
• Suggested PSU: Not listed vs 450 W.
• Bus Interface: PCIe 4.0 x16 vs PCIe 3.0 x16.
• Slot Width: Not listed vs Dual-slot.
• Dimensions: Not listed vs 229 mm x 113 mm x 35 mm.
• Display Outputs: Portable Device Dependent vs 1x DVI, 1x HDMI 2.0, 2x DisplayPort 1.4a, 1x USB Type-C.
• Release Date: 2021-01-11 vs 2018-10-16.
• Predecessor: GeForce 20 Mobile vs GeForce 10.
• Successor: None vs GeForce 30.

Memory size, type, bus width, bandwidth, and memory clock are identical. Both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

FAQ

Q: Which GPU has a higher average benchmark score?

A: The NVIDIA GeForce RTX 3070 Mobile has an average benchmark score of 20534, which is higher than the RTX 2070’s 18789.

Q: How much faster is the RTX 3070 Mobile in DirectX 12?

A: In 3dmark_3dmark_steel_nomad_dx12, the RTX 3070 Mobile scores 2380 versus 1569, a 51.7% advantage. In passmark_directx_12, it leads 64 to 60, a 6.7% margin.

Q: In which tests does the RTX 2070 win?

A: The RTX 2070 wins passmark_directx_9 (211 vs 160, +24.2%), passmark_g2d (819 vs 641, +21.7%), and passmark_g3d (16094 vs 15309, +4.9%).

Q: What are the transistor counts for each chip?

A: The RTX 3070 Mobile’s GA104 chip contains 17,400 million transistors, while the RTX 2070’s TU106 chip contains 10,800 million transistors.

Q: Do both cards have the same memory configuration?

A: Yes, both have 8 GB of GDDR6 memory on a 256-bit bus, with 448.0 GB/s bandwidth and a memory clock of 1750 MHz (14 Gbps effective).

Q: What is the TDP difference between the two?

A: The RTX 3070 Mobile has a TDP of 115 W with no power connectors, while the RTX 2070 has a TDP of 175 W and requires a 1x 8-pin connector.

The Verdict

The data makes a compelling case for the NVIDIA GeForce RTX 3070 Mobile as the superior GPU for modern workloads. Its 51.7% lead in 3DMark Steel Nomad DX12 and 16.2% lead in OpenCL are decisive, and its wins in DirectX 11, DirectX 12, and compute benchmarks show broad applicability. The mobile chip achieves this while consuming 60 W less power (115 W vs 175 W), a remarkable efficiency gain enabled by the 8 nm Samsung process versus the 12 nm TSMC node. For anyone building a system for current-generation gaming, ray tracing, or GPU compute, the RTX 3070 Mobile is the clear choice.

The RTX 2070, however, is not without merit. Its wins in DirectX 9 and 2D benchmarks indicate that it handles legacy software better, and its higher boost clock (1620 MHz vs 1560 MHz) may benefit certain clock-sensitive workloads. The passmark_g3d victory (16094 vs 15309) shows it can still hold its own in some 3D tasks. On top of that, the RTX 2070 is a desktop card with standard display outputs (DVI, HDMI, DisplayPort, USB Type-C) and a dual-slot form factor, making it easier to integrate into a traditional desktop build. The RTX 3070 Mobile’s "Portable Device Dependent" outputs mean it is tied to a laptop or proprietary mobile platform.

The verdict depends on the use case. For a new purchase focused on longevity and modern performance, the RTX 3070 Mobile wins outright. For a system that must run older DirectX 9 titles or prioritize 2D acceleration, the RTX 2070’s specific strengths should not be dismissed. The percentile rankings (65th vs 63rd) and average scores (20534 vs 18789) both favor the RTX 3070 Mobile, but the RTX 2070 remains a capable option for niche legacy workloads.