AMD Radeon Pro 5700 vs NVIDIA GeForce RTX 2070 Comparison

The NVIDIA GeForce RTX 2070 is the clear performance leader in this comparison, winning all nine shared benchmark tests against the AMD Radeon Pro 5700. However, the Radeon Pro 5700 is not without merit, offering a more efficient architecture, a smaller die, and a lower power draw that makes it a compelling choice for specific, thermally constrained professional environments. This analysis breaks down the data to show exactly where each card excels and for whom each is better suited.

Head-to-Head Benchmarks

The RTX 2070's dominance is comprehensive, but the margin of victory varies significantly across different workloads. The largest gap appears in API-specific legacy and compute tests. In the Passmark DirectX 11 benchmark, the RTX 2070 scores 129 against the Radeon Pro 5700's 78, a substantial 65.4% advantage. Similarly, in Passmark DirectX 10, the RTX 2070's score of 111 is 56.3% higher than the Radeon's 71. These results suggest the NVIDIA card maintains superior performance in older DirectX titles and applications, likely due to driver optimizations and a more mature feature set for those APIs.

Compute performance tells a similar story. In Geekbench Vulkan, the RTX 2070 scores 82521, which is 60.9% higher than the Radeon Pro 5700's 51289. The Geekbench OpenCL test also heavily favors the RTX 2070, with a score of 79966 versus 52787, a 51.5% lead. This indicates that for general-purpose GPU computing tasks like rendering or data processing, the NVIDIA card delivers significantly more raw throughput.

The gap narrows in more modern DirectX 12 workloads. In Passmark DirectX 12, the RTX 2070 scores 60, which is only 25% higher than the Radeon's 48. While still a win for NVIDIA, this smaller delta suggests that the AMD RDNA architecture closes the performance gap when leveraging the lower-level hardware access of DirectX 12. The same trend is visible in the Passmark DirectX 9 test, where the RTX 2070's 211 points gives it a 47.6% lead over the 143 points of the Radeon.

Even in 2D and rasterization-focused benchmarks, NVIDIA maintains a lead, though it is much smaller. The Passmark G2D test shows a near-tie, with the RTX 2070 scoring 819 versus 805, a marginal 1.7% advantage. However, in the more demanding Passmark G3D test, which assesses overall 3D graphics performance, the RTX 2070's score of 16094 is 38.9% higher than the Radeon Pro 5700's 11583. Finally, in Passmark GPU Compute, the RTX 2070 leads with 6411 points, a 29.2% advantage over the Radeon's 4961. This confirms that while the AMD card is competitive in some sub-tests, the NVIDIA card provides a decisive overall performance advantage.

The Verdict

Based on the benchmark data, the choice is clear for users prioritizing raw performance: the NVIDIA GeForce RTX 2070 is the superior option. It wins every single head-to-head benchmark in the data, often by significant margins. Its average benchmark score of 18789 places it in the 63rd percentile of all GPUs, compared to the Radeon Pro 5700's 18189 and 62nd percentile. For gamers, 3D artists, or anyone running compute-intensive applications where frame rates and render times are paramount, the RTX 2070 is the undeniable winner, offering a 38.9% lead in the overall G3D test and a 51.5% lead in OpenCL compute.

The AMD Radeon Pro 5700 is the better choice for a very specific user: one operating in a power- and space-constrained system, such as a Mac Pro. Its 130 W TDP is significantly lower than the RTX 2070's 175 W, and it requires no power connectors, being an integrated GPU (IGP) with no display outputs. It also uses a more advanced 7 nm process node compared to the RTX 2070's 12 nm, resulting in a much denser and more power-efficient chip. If a user needs a discrete-class GPU for a system with a 300 W power supply suggested and no ability to add a card, the Radeon Pro 5700 is the only viable option. The data shows it is less performant, but its physical and power characteristics make it the default choice for that niche.

FAQ

Q: Which card is faster in Vulkan applications?

A: The NVIDIA GeForce RTX 2070 is significantly faster, scoring 82521 in the Geekbench Vulkan test compared to the AMD Radeon Pro 5700's 51289, a 60.9% advantage.

Q: Does the AMD Radeon Pro 5700 win any benchmark?

A: No. In the nine head-to-head benchmarks where both cards have scores, the NVIDIA GeForce RTX 2070 wins all of them, with the AMD Radeon Pro 5700 winning zero.

Q: Which card has a higher average benchmark score?

A: The NVIDIA GeForce RTX 2070 has a higher average benchmark score of 18789, compared to the AMD Radeon Pro 5700's 18189. This places the RTX 2070 in the 63rd percentile of all GPUs, one point higher than the Radeon Pro 5700's 62nd percentile.

Q: How do the cards compare in terms of power consumption?

A: The AMD Radeon Pro 5700 has a significantly lower TDP of 130 W and requires no power connectors, while the NVIDIA GeForce RTX 2070 has a TDP of 175 W and requires a single 8-pin power connector.

Q: What is the performance difference in the DirectX 12 benchmark?

A: The NVIDIA GeForce RTX 2070 scores 60 in the Passmark DirectX 12 test, which is 25% higher than the AMD Radeon Pro 5700's score of 48. This is the smallest performance gap between the two cards in any test.

Q: Which card is better for compute workloads?

A: The NVIDIA GeForce RTX 2070 is better for compute workloads. It leads by 51.5% in Geekbench OpenCL and by 29.2% in Passmark GPU Compute.

Specification Differences

The core specifications reveal two very different design philosophies. The NVIDIA GeForce RTX 2070 is built around the TU106 chip on a 12 nm process, while the AMD Radeon Pro 5700 uses the Navi 10 chip on a 7 nm node. Both cards feature 8 GB of GDDR6 memory on a 256-bit bus, but the RTX 2070's memory runs at a higher effective speed of 14 Gbps, giving it a bandwidth of 448.0 GB/s versus the Radeon Pro 5700's 384.0 GB/s from its 12 Gbps memory.

The RTX 2070 has higher clock speeds, with a base of 1410 MHz and a boost of 1620 MHz, compared to the Radeon Pro 5700's 1243 MHz base and 1350 MHz boost. This, combined with the memory bandwidth advantage, leads to higher theoretical fill rates for the NVIDIA card: 103.7 GPixel/s and 233.3 GTexel/s versus the AMD's 86.40 GPixel/s and 194.4 GTexel/s. The RTX 2070 also delivers higher compute performance, with 7.465 TFLOPS FP32 versus 6.221 TFLOPS for the Radeon. Despite this, the Radeon Pro 5700 is far more power-efficient, with a 130 W TDP and a suggested PSU of 300 W, compared to the RTX 2070's 175 W TDP and 450 W suggested PSU. The form factors also differ drastically: the RTX 2070 is a dual-slot, add-in card with a length of 229 mm, while the Radeon Pro 5700 is an IGP with no dimensions listed and no display outputs.

Architecture Differences

The architecture differences are fundamental. The RTX 2070 is based on NVIDIA's Turing architecture, which introduces dedicated hardware for real-time ray tracing and AI. It features 36 RT Cores and 288 Tensor Cores, which are absent from the AMD Radeon Pro 5700's RDNA 1.0 architecture. This gives the RTX 2070 a distinct feature advantage, supporting DirectX 12 Ultimate (12_2), which includes features like hardware-accelerated ray tracing. The Radeon Pro 5700 only supports DirectX 12 (12_1).

While both cards have the same number of shading units (2304), TMUs (144), and ROPs (64), the underlying architectures are optimized differently. The Turing architecture in the RTX 2070 is designed to handle complex, mixed workloads, leveraging its Tensor Cores for AI-powered features like DLSS. The RDNA 1.0 architecture in the Radeon Pro 5700 focuses on raw efficiency and performance-per-watt, which is evident in its 41.0M / mm² transistor density on a 251 mm² die, versus the RTX 2070's 24.3M / mm² on a much larger 445 mm² die. The Radeon Pro 5700 also supports PCIe 4.0, while the RTX 2070 is limited to PCIe 3.0. The data shows that the RTX 2070's specialized hardware provides a massive performance advantage in the tested benchmarks, but the Radeon Pro 5700's more efficient design is a significant engineering achievement in its own right.

Where Each One Wins

NVIDIA GeForce RTX 2070:

• Raw Performance: Wins every benchmark, with leads ranging from 25% (DirectX 12) to 65.4% (DirectX 11).
• Compute & AI: Dominates in OpenCL (51.5% lead) and Vulkan (60.9% lead) due to its dedicated Tensor and RT cores.
• Gaming: The 38.9% lead in the Passmark G3D test makes it the clear choice for high-fidelity gaming.
• Legacy API Support: Significantly better performance in DirectX 9 (47.6% lead) and DirectX 10 (56.3% lead).

AMD Radeon Pro 5700:

• Power Efficiency: Its 130 W TDP is substantially lower than the RTX 2070's 175 W, requiring only a 300 W power supply.
• Integration: As an IGP with no power connectors or display outputs, it is the only option for systems where a discrete card is physically impossible.
• Modern Process Node: Built on the more advanced 7 nm process, it offers a higher transistor density (41.0M / mm²) and a much smaller die (251 mm²) than the RTX 2070.
• Close Contests: It comes closest to the RTX 2070 in the DirectX 12 benchmark, reducing the deficit to just 25%, and in the G2D test, where it trails by only 1.7%.