GPU Comparison

NVIDIA’s RTX 2000 Ada Generation and the older GeForce RTX 2070 occupy the same performance percentile (63rd), but they achieve that status through very different means. The RTX 2000 Ada Generation is a modern, power-sipping workstation card built on a 5 nm process, while the RTX 2070 is a 12 nm gaming-class part from 2018. The benchmark data shows the RTX 2000 Ada Generation wins 8 of 10 head-to-head tests, yet the RTX 2070 still holds specific advantages in legacy DirectX 10 workloads and OpenCL compute. This analysis breaks down where each card dominates, the architectural reasons behind those results, and which user should choose which.

Head-to-Head Benchmarks

The most decisive win for the RTX 2000 Ada Generation comes in the Passmark G2D test, where it scores 1072 against the RTX 2070’s 819, a 30.9% lead. This is a massive gap for a 2D workload, reflecting the newer card’s superior memory subsystem efficiency and display output handling. In GPU compute, the RTX 2000 Ada Generation also flexes its muscles, scoring 7834 versus 6411, a 22.2% advantage. That compute lead is significant for any task involving GPGPU acceleration, from rendering to simulation.

The 3DMark Steel Nomad DX12 test shows the RTX 2000 Ada Generation ahead by 12.6%, with scores of 1767 against 1569. This is a modern DirectX 12 workload, and the newer architecture’s advantage is clear. The Passmark DirectX 12 test shows an even larger margin for the RTX 2000 Ada Generation: 71 versus 60, an 18.3% win. These DX12 results suggest the Ada card is better optimized for current-generation graphics APIs.

In older DirectX 11 workloads, the RTX 2000 Ada Generation wins by 7% (138 vs 129), and in DirectX 9 it edges out the RTX 2070 by 2.4% (216 vs 211). The Vulkan test also goes to the RTX 2000 Ada Generation, but by a narrow 1% margin (83360 vs 82521). The Passmark G3D overall score favors the RTX 2000 Ada Generation at 16927 versus 16094, a 5.2% difference.

The RTX 2070’s two wins are notable. The first is in Geekbench OpenCL, where it scores 79966 against 78074, a 2.4% lead. The second is in Passmark DirectX 10, where it wins decisively by 26.1% (111 vs 82). This DirectX 10 result is the single largest margin in either direction across all tests, indicating that the RTX 2070’s older Turing design retains a significant advantage in legacy API paths. The Geekbench OpenCL result is interesting because it contradicts the Passmark GPU Compute result, showing that compute performance is highly workload-dependent.

Where Each One Wins

The RTX 2000 Ada Generation is the clear winner for modern API workloads. It dominates DirectX 12, Vulkan, and DirectX 11 tests, making it the better choice for current-generation games and professional applications that use these APIs. Its 30.9% lead in 2D performance also makes it superior for desktop compositing, multi-monitor setups, and any application that stresses the display output pipeline. The 22.2% compute advantage in Passmark GPU Compute suggests it handles general-purpose compute tasks more efficiently.

The RTX 2070 wins in two specific niches. First, it is significantly better in DirectX 10, with a 26.1% margin. If you run legacy applications or games that rely on DirectX 10, the RTX 2070 is the better performer. Second, it wins in Geekbench OpenCL by 2.4%, which could matter for specific OpenCL-accelerated workloads that align with its architecture. The RTX 2070’s higher texture rate (233.3 GTexel/s) and pixel rate (103.7 GPixel/s) also suggest it may handle certain fill-rate-bound tasks better, though the benchmark data shows the RTX 2000 Ada Generation winning the overall G3D score.

For a professional user, the RTX 2000 Ada Generation’s wins are more relevant. Its DirectX 12, Vulkan, and compute advantages align with modern CAD, rendering, and AI workloads. For a gamer playing older titles, the RTX 2070’s DirectX 10 dominance could be a deciding factor.

Architecture Differences

The two cards are built on entirely different architectures. The RTX 2000 Ada Generation uses the AD107 chip on the Ada Lovelace architecture, fabricated on a 5 nm process at TSMC. The RTX 2070 uses the TU106 chip on the Turing architecture, fabricated on a 12 nm process. This is a massive generational leap in manufacturing technology. The Ada chip packs 18,900 million transistors into a 159 mm² die, giving it a transistor density of 118.9M / mm². The Turing chip has 10,800 million transistors on a 445 mm² die, with a density of just 24.3M / mm².

This density difference explains the power and efficiency gap. The RTX 2000 Ada Generation has a TDP of 70 W, while the RTX 2070 draws 175 W. The Ada card also has a higher base clock of 1620 MHz versus 1410 MHz, and a higher boost clock of 2130 MHz versus 1620 MHz. Despite the higher clocks, the Ada card uses less than half the power, thanks to the 5 nm process.

The memory architecture also differs fundamentally. The RTX 2000 Ada Generation has 16 GB of GDDR6 on a 128-bit bus, delivering 256.0 GB/s of bandwidth. The RTX 2070 has 8 GB of GDDR6 on a 256-bit bus, delivering 448.0 GB/s. The RTX 2070 has significantly more bandwidth, but the Ada card has double the capacity. The Ada card’s memory runs at 2000 MHz (16 Gbps effective), while the RTX 2070’s runs at 1750 MHz (14 Gbps effective).

In terms of compute units, the Ada card has 2816 shading units, 88 TMUs, and 48 ROPs. The RTX 2070 has 2304 shading units, 144 TMUs, and 64 ROPs. The Ada card has more shading units but fewer TMUs and ROPs. The Ada card has 22 RT cores and 88 tensor cores, while the RTX 2070 has 36 RT cores and 288 tensor cores. The RTX 2070 has more RT and tensor cores, but the Ada architecture’s newer design allows for higher FP32 throughput: 12.00 TFLOPS versus 7.465 TFLOPS. The Ada card also achieves 12.00 TFLOPS FP16 (1:1), while the RTX 2070 achieves 14.93 TFLOPS FP16 (2:1).

Specification Differences

The key specification differences are as follows:

• Process Node: 5 nm (Ada) vs 12 nm (Turing)
• Transistors: 18,900 million vs 10,800 million
• Die Size: 159 mm² vs 445 mm²
• Base Clock: 1620 MHz vs 1410 MHz
• Boost Clock: 2130 MHz vs 1620 MHz
• Memory Size: 16 GB vs 8 GB
• Memory Bus Width: 128 bit vs 256 bit
• Memory Bandwidth: 256.0 GB/s vs 448.0 GB/s
• Shading Units: 2816 vs 2304
• TMUs: 88 vs 144
• ROPs: 48 vs 64
• RT Cores: 22 vs 36
• Tensor Cores: 88 vs 288
• FP32 Performance: 12.00 TFLOPS vs 7.465 TFLOPS
• FP16 Performance: 12.00 TFLOPS (1:1) vs 14.93 TFLOPS (2:1)
• TDP: 70 W vs 175 W
• Power Connectors: None vs 1x 8-pin
• Suggested PSU: 250 W vs 450 W
• Bus Interface: PCIe 4.0 x8 vs PCIe 3.0 x16
• Display Outputs: 4x mini-DisplayPort 1.4a vs 1x DVI, 1x HDMI 2.0, 2x DisplayPort 1.4a, 1x USB Type-C
• Dimensions: 168 mm length vs 229 mm length
• Production Status: Active vs End-of-life

FAQ

Q: Which card is faster in DirectX 12?

A: The RTX 2000 Ada Generation wins both DX12 tests. It scores 1767 in 3DMark Steel Nomad DX12 (12.6% ahead) and 71 in Passmark DirectX 12 (18.3% ahead of the RTX 2070’s 60).

Q: Does the RTX 2070 have any clear advantages?

A: Yes. It wins Passmark DirectX 10 by 26.1% (111 vs 82) and Geekbench OpenCL by 2.4% (79966 vs 78074). It also has higher memory bandwidth (448.0 GB/s vs 256.0 GB/s).

Q: Which card is more power-efficient?

A: The RTX 2000 Ada Generation has a TDP of 70 W, while the RTX 2070 has a TDP of 175 W. The Ada card also requires no power connectors, while the RTX 2070 needs a single 8-pin connector.

Q: How do they compare in overall average benchmark scores?

A: The RTX 2000 Ada Generation has an average benchmark score of 18954, slightly ahead of the RTX 2070’s 18789. Both cards sit at the 63rd percentile of all GPUs.

Q: Which card has more memory?

A: The RTX 2000 Ada Generation has 16 GB of GDDR6, which is double the RTX 2070’s 8 GB. However, the RTX 2070 has a wider 256-bit bus and higher bandwidth (448.0 GB/s vs 256.0 GB/s).

Q: Are these cards the same physical size?

A: No. The RTX 2000 Ada Generation is 168 mm long and 69 mm high, while the RTX 2070 is 229 mm long, 113 mm high, and 35 mm wide. The Ada card is significantly smaller.

The Verdict

The data paints a clear picture: the RTX 2000 Ada Generation is the superior card for modern workloads. It wins 8 of 10 head-to-head benchmarks, including all DirectX 12, Vulkan, and DirectX 11 tests, plus the overall Passmark G3D score. Its 30.9% lead in 2D performance and 22.2% lead in GPU compute make it a better workstation tool. The 70 W TDP and lack of power connectors mean it can fit into systems with a 250 W PSU, whereas the RTX 2070 requires a 450 W PSU and an 8-pin connector.

The RTX 2070’s only compelling reason to choose it is legacy DirectX 10 performance, where it wins by 26.1%. If you run software that depends on DirectX 10, the RTX 2070 is the better choice. Its higher memory bandwidth (448 GB/s) and higher texture rate (233.3 GTexel/s) may also help in specific fill-rate-bound scenarios, though the overall G3D score still favors the Ada card.

For professionals, the RTX 2000 Ada Generation is the obvious pick. It offers double the VRAM (16 GB vs 8 GB), higher FP32 throughput (12.00 TFLOPS vs 7.465 TFLOPS), and a smaller physical footprint. Its active production status means long-term availability, while the RTX 2070 is end-of-life. For gamers playing modern titles, the Ada card’s DX12 and Vulkan advantages make it the stronger performer, despite the RTX 2070’s higher raw bandwidth. The RTX 2000 Ada Generation is the more future-proof, efficient, and capable card.