AMD Radeon Pro 5700 vs NVIDIA GeForce RTX 2070 SUPER Comparison

Head-to-Head Benchmarks

The NVIDIA GeForce RTX 2070 SUPER wins all nine recorded head-to-head benchmark comparisons against the AMD Radeon Pro 5700. The margin is not uniform, however, and the spread between tests reveals where each architecture retains relative strength.

The largest single victory for the RTX 2070 SUPER comes in PassMark DirectX 11, where it scores 151 against 78 for the Radeon Pro 5700, a 93.6% advantage. That is nearly double the AMD card's output in a legacy API workload, and it sets the tone for the rest of the DirectX tests. PassMark DirectX 10 shows a similar pattern: 132 versus 71, a 85.9% lead. PassMark DirectX 9 favors the NVIDIA card by 55.9%, with scores of 223 and 143. The smallest DirectX gap is in PassMark DirectX 12, where the RTX 2070 SUPER scores 67 against 48, still a 39.6% lead, but the closest margin of any 3D test in the comparison.

Compute workloads also lean heavily toward NVIDIA. Geekbench OpenCL puts the RTX 2070 SUPER at 83358 versus 52787, a 57.9% advantage. Geekbench Vulkan is even more lopsided: 90637 versus 51289, a 76.7% lead. PassMark GPU Compute follows with 7557 versus 4961, a 52.3% margin. These three results indicate that the NVIDIA card's compute throughput is consistently more than half again as high as the AMD card's across different compute APIs.

The synthetic gaming-style benchmark, PassMark G3D, shows a 56.9% lead for NVIDIA: 18169 versus 11583. That margin aligns closely with the OpenCL and GPU Compute results, suggesting the performance gap is fairly consistent across general-purpose rendering workloads. The 2D test is the only place where the two cards are close. PassMark G2D scores 878 versus 805, a 9.1% lead for NVIDIA, which is the smallest delta in the entire head-to-head set.

The average benchmark score tells a similar story. The RTX 2070 SUPER averages 20282 across its recorded tests, while the Radeon Pro 5700 averages 18189, a difference of roughly 11.5%. The RTX 2070 SUPER sits at the 65th percentile among all GPUs in the database, while the Radeon Pro 5700 sits at the 62nd percentile. Both cards are in the middle of the pack, but the NVIDIA card is positioned slightly higher.

The nearest rivals in the database put these numbers in context. The RTX 2070 SUPER's closest competitors are the NVIDIA Quadro M4000M at 20480 (-1%), the NVIDIA GeForce RTX 3070 Mobile at 20534 (-1.2%), the Intel Arc B570 at 20556 (-1.3%), and the Intel Arc A750 at 20582 (-1.5%). The Radeon Pro 5700, by contrast, sits near the NVIDIA GeForce RTX 3060 Mobile at 18159 (+0.2%), the NVIDIA GeForce RTX 2060 SUPER at 18093 (+0.5%), the AMD Radeon RX 460 at 18373 (-1%), and the Intel Arc A770M at 18383 (-1.1%). Notably, the RTX 2060 SUPER, a card from the same generation as the RTX 2070 SUPER, essentially matches the Radeon Pro 5700's average score, while the RTX 2070 SUPER sits clearly above it.

The Verdict

The data is unambiguous: the NVIDIA GeForce RTX 2070 SUPER is the faster card in every recorded benchmark. Nine wins out of nine head-to-head tests is a clean sweep, and the margins are substantial in most cases. The smallest lead, 9.1% in PassMark G2D, is still a win, while the largest, 93.6% in PassMark DirectX 11, is a near-doubling of performance.

The Radeon Pro 5700 is not without merit, but its merits do not show up in raw performance numbers. It matches the RTX 2060 SUPER in average score (18189 versus 18093, a 0.5% difference), which places it in the same performance tier as a lower-tier NVIDIA card from the previous generation. The RTX 2070 SUPER, by comparison, sits roughly 11.5% higher in average score and clears the Radeon Pro 5700 by at least 39.6% in every 3D and compute test.

For users who prioritize peak frame rates, compute throughput, or API compatibility across DirectX versions, the RTX 2070 SUPER is the clear choice based on the recorded data. For users who need a card that fits into a 300 W system, the Radeon Pro 5700's lower power draw (130 W versus 215 W) and integrated form factor may be the deciding factor, but that is a system constraint, not a performance advantage. The database shows no benchmark where the AMD card wins.

FAQ

Q: Which card has the higher average benchmark score?

A: The NVIDIA GeForce RTX 2070 SUPER averages 20282 across its recorded tests, while the AMD Radeon Pro 5700 averages 18189.

Q: What is the largest performance gap between the two cards?

A: The largest gap is in PassMark DirectX 11, where the RTX 2070 SUPER scores 151 versus 78, a 93.6% advantage.

Q: Is there any test where the Radeon Pro 5700 comes close to the RTX 2070 SUPER?

A: The closest result is PassMark G2D, where the Radeon Pro 5700 scores 805 against 878, a 9.1% lead for NVIDIA. Every other test shows at least a 39.6% lead for the RTX 2070 SUPER.

Q: How do the two cards compare in compute workloads?

A: The RTX 2070 SUPER leads by 57.9% in Geekbench OpenCL (83358 versus 52787), 76.7% in Geekbench Vulkan (90637 versus 51289), and 52.3% in PassMark GPU Compute (7557 versus 4961).

Q: Where does each card rank among all GPUs in the database?

A: The RTX 2070 SUPER sits at the 65th percentile, and the Radeon 5700 Pro 5700 sits at the 62nd percentile.

Q: What is the Radeon Pro 5700's closest rival in average score?

A: The closest rival is the NVIDIA GeForce RTX 2070 SUPER's lower-tier sibling, the RTX 2060 SUPER, at 0.5% difference in average score (18189 versus 18093).

Specification Differences

The two cards differ in nearly every recorded specification field. The RTX 2070 SUPER uses a 12 nm process node from TSMC, while the Radeon Pro 5700 uses TSMC's 7 nm node. The transistor counts differ as well: 13,600 million for the RTX 2070 SUPER versus 10,300 million for the Radeon Pro 5700. Die sizes are 545 mm² and 251 mm² respectively, giving transistor densities of 25.0M per mm² for NVIDIA and 41.0M per mm² for AMD.

Clock speeds favor NVIDIA on both base and boost: 1605 MHz base and 1770 MHz boost for the RTX 2070 SUPER, compared to 1243 MHz base and 1350 MHz boost for the Radeon Pro 5700. Memory clocks also differ, with the RTX 2070 SUPER running at 1750 MHz (14 Gbps effective) and the Radeon Pro 5700 at 1500 MHz (12 Gbps effective). Memory bandwidth reflects this: 448.0 GB/s for NVIDIA versus 384.0 GB/s for AMD.

The compute configuration is larger on the NVIDIA card: 2560 shading units, 160 TMUs, 64 ROPs, 40 RT cores, and 320 tensor cores. The Radeon Pro 5700 has 2304 shading units, 144 TMUs, and 64 ROPs, with no RT cores or tensor cores listed. Pixel rate is 113.3 GPixel/s for NVIDIA versus 86.40 GPixel/s for AMD, and texture rate is 283.2 GTexel/s versus 194.4 GTexel/s. FP32 throughput is 9.062 TFLOPS versus 6.221 TFLOPS, and FP16 is 18.12 TFLOPS versus 12.44 TFLOPS, both at 2:1 ratio.

Power and physical specifications diverge sharply. The RTX 2070 SUPER is rated at 215 W with a suggested 550 W PSU, a dual-slot form factor, and 1x 6-pin plus 1x 8-pin power connectors. The Radeon Pro 5700 is rated at 130 W with a suggested 300 W PSU, an IGP form factor, and no power connectors. The RTX 2070 SUPER has display outputs (1x HDMI 2.0, 3x DisplayPort 1.4a, 1x USB Type-C), while the Radeon Pro 5700 has no outputs listed. The bus interface is PCIe 3.0 x16 for NVIDIA and PCIe 4.0 x16 for AMD. DirectX support is 12 Ultimate (12_2) for NVIDIA and 12 (12_1) for AMD. The RTX 2070 SUPER measures 267 mm by 116 mm by 35 mm, while the Radeon Pro 5700 has no dimensions listed.

Architecture Differences

The architectural split is fundamental. The RTX 2070 SUPER is built on NVIDIA's Turing architecture, using the TU104 chip, while the Radeon Pro 5700 uses AMD's RDNA 1.0 architecture on the Navi 10 chip. Turing is a 12 nm design from TSMC, and RDNA 1.0 is a 7 nm design from the same foundry. The smaller process node allows AMD to pack 41.0M transistors per mm² on a 251 mm² die, versus 25.0M per mm² on a 545 mm² die for NVIDIA. Despite the smaller die, the AMD card carries fewer total transistors: 10,300 million versus 13,600 million.

Turing includes dedicated ray tracing cores (40 on the RTX 2070 SUPER) and tensor cores (320), neither of which appear on the Radeon Pro 5700's specification sheet. This is a structural difference, not just a clock or core count difference. The NVIDIA card also supports DirectX 12 Ultimate (12_2), while the AMD card supports DirectX 12 (12_1), indicating a different level of feature support in the latest API tier. Both cards support OpenGL 4.6 and Vulkan 1.4, so those API levels match.

The memory systems are similar in size and type (8 GB GDDR6 on a 256-bit bus for both), but NVIDIA runs the memory at a higher effective speed, 14 Gbps versus 12 Gbps, yielding 448.0 GB/s versus 384.0 GB/s. The Radeon Pro 5700 is a Radeon Pro Mac (Navi Series) product, which is reflected in its integrated form factor, no display outputs, and no power connectors. It is designed to be embedded in a system rather than installed as a discrete card. The RTX 2070 SUPER is a conventional discrete card with a dual-slot cooler, external power connectors, and a full set of display outputs.

The release timeline also differs: the RTX 2070 SUPER launched on 2019-07-08, and the Radeon Pro 5700 on 2020-08-03. Both are end-of-life products. The RTX 2070 SUPER has a predecessor (GeForce 10) and successor (GeForce 30) in the database, while the Radeon Pro 5700 lists neither.

Where Each One Wins

The RTX 2070 SUPER wins in every recorded benchmark, so the use-case split is not about which card is faster in a given test. It is about what the test represents and what the user needs.

The RTX 2070 SUPER is the pick for DirectX workloads across all versions. Its 93.6% lead in DirectX 11 and 85.9% lead in DirectX 10 show particularly strong legacy API performance, and its 39.6% lead in DirectX 12, while smaller, is still decisive. For compute, the 57.9% OpenCL lead, 76.7% Vulkan lead, and 52.3% GPU Compute lead make it the clear choice for any application that offloads work to the GPU. For gaming-style rendering, the 56.9% PassMark G3D lead reinforces the same conclusion.

The Radeon Pro 5700 does not win any recorded benchmark, but it has structural advantages that matter in specific system contexts. Its 130 W TDP with no power connectors and IGP form factor means it can fit into systems that cannot accommodate a dual-slot, 215 W card with external power. Its 300 W suggested PSU is far below the RTX 2070 SUPER's 550 W requirement. For a user building or upgrading a Mac-based system with limited power delivery and no room for a discrete card, the Radeon Pro 5700 is the only one of the two that fits. The RTX 2070 SUPER also has no recorded Geekbench Metal score, while the Radeon Pro 5700 scores 60126 in that test, so on Apple platforms with Metal as the primary API, the AMD card has a data point the NVIDIA card lacks.

The RTX 2070 SUPER also offers a larger feature set for future-facing workloads: 40 RT cores, 320 tensor cores, DirectX 12 Ultimate support, and 18.12 TFLOPS FP16 throughput. The Radeon Pro 5700 lists no RT or tensor cores and caps at DirectX 12 (12_1). For users who expect to use ray tracing, tensor-based AI features, or the latest DirectX feature tier, the RTX 2070 SUPER is the only choice between these two. For users whose primary constraint is power and physical integration, the Radeon Pro 5700 is the only card that fits the requirement. The benchmark data does not change that, but it does show that the performance trade-off for that form factor is substantial.