AMD Radeon Pro W5700 vs NVIDIA Quadro RTX 8000 Comparison

The Verdict

The benchmark data presents a clear hierarchy: the NVIDIA Quadro RTX 8000 dominates the AMD Radeon Pro W5700 in nearly every recorded test. Across nine head-to-head benchmarks, the Quadro RTX 8000 wins seven, while the Radeon Pro W5700 takes only two. The average benchmark score for the Quadro RTX 8000 is 28,421, placing it in the 74th percentile of all GPUs, while the Radeon Pro W5700 averages 25,726, sitting in the 71st percentile. The overall score gap is 10.5% in favor of NVIDIA.

For users prioritizing raw compute, graphics rendering, or DirectX 11/12 workloads, the Quadro RTX 8000 is the statistically superior choice. Its wins include a 36.5% margin in Geekbench OpenCL, a 73.4% margin in Geekbench Vulkan, and a 53.8% margin in Passmark GPU Compute. The Radeon Pro W5700, however, wins in two legacy or 2D-oriented tests: Passmark DirectX 9 by 6.2% and Passmark G2D by 3.7%. This suggests the AMD card holds a niche for older API workloads and 2D acceleration, but for modern professional tasks, the NVIDIA card is the one to select.

The Radeon Pro W5700 also warrants consideration for systems where power draw and physical footprint matter, as it carries a 205 W TDP versus the Quadro’s 260 W, and it supports PCIe 4.0 x16 compared to PCIe 3.0 x16. Yet, the performance data consistently favors the Quadro RTX 8000, which also offers 48 GB of GDDR6 memory versus 8 GB, a sixfold capacity advantage that directly impacts large dataset handling.

Architecture Differences

The two cards come from different architectural lineages. The NVIDIA Quadro RTX 8000 uses the TU102 chip built on Turing architecture, fabricated by TSMC on a 12 nm process. It contains 18,600 million transistors on a 754 mm² die, yielding a transistor density of 24.7 million per mm². In contrast, the AMD Radeon Pro W5700 uses the Navi 10 chip with RDNA 1.0 architecture, also from TSMC but on a 7 nm process. It packs 10,300 million transistors into a 251 mm² die, achieving a transistor density of 41.0 million per mm². The AMD chip is denser and smaller, but the NVIDIA chip has nearly double the transistor count.

Core configurations diverge sharply. The Quadro RTX 8000 features 4,608 shading units, 288 texture mapping units, and 96 render output units. It also includes 72 RT cores and 576 tensor cores, enabling hardware-accelerated ray tracing and AI workloads. The Radeon Pro W5700 has 2,304 shading units, 144 TMUs, and 64 ROPs, with no dedicated RT or tensor cores. This structural difference explains the NVIDIA card’s massive lead in compute-oriented benchmarks.

Memory subsystems are also fundamentally different. The Quadro RTX 8000 has 48 GB of GDDR6 on a 384-bit bus, providing 672.0 GB/s of bandwidth. The Radeon Pro W5700 has 8 GB of GDDR6 on a 256-bit bus, delivering 448.0 GB/s. Clock speeds are comparable in base (1395 MHz vs 1400 MHz), but the AMD card boosts higher at 1880 MHz versus 1770 MHz. Despite the higher boost clock, the NVIDIA card’s superior core count and memory bandwidth dominate in practice.

Feature support shows differences as well. The Quadro RTX 8000 supports DirectX 12 Ultimate (12_2), while the Radeon Pro W5700 supports DirectX 12 (12_1). Both support OpenGL 4.6 and Vulkan 1.4. The AMD card uses PCIe 4.0 x16, doubling the bus bandwidth available to the NVIDIA card’s PCIe 3.0 x16. Display outputs differ: the NVIDIA card has 4x DisplayPort 1.4a plus 1x USB Type-C, while the AMD card has 5x mini-DisplayPort 1.4a plus 1x USB Type-C.

FAQ

Q: Which card has more memory, and what does that mean for workloads?

A: The NVIDIA Quadro RTX 8000 has 48 GB of GDDR6, while the AMD Radeon Pro W5700 has 8 GB. The NVIDIA card’s memory bandwidth is 672.0 GB/s versus 448.0 GB/s for AMD. This means the Quadro can hold larger datasets, such as complex 3D scenes or high-resolution textures, entirely in VRAM without spilling to system memory.

Q: Is the AMD Radeon Pro W5700 ever faster than the NVIDIA Quadro RTX 8000?

A: Yes, in two recorded benchmarks. The AMD card wins Passmark DirectX 9 with a score of 225 versus 211 for NVIDIA, a 6.2% advantage. It also wins Passmark G2D with 899 versus 866, a 3.7% lead. These are legacy API and 2D performance tests, respectively.

Q: How do the two cards compare in raw compute performance?

A: The NVIDIA Quadro RTX 8000 delivers 16.31 TFLOPS FP32 and 32.62 TFLOPS FP16 (2:1). The AMD Radeon Pro W5700 delivers 8.663 TFLOPS FP32 and 17.33 TFLOPS FP16 (2:1). The NVIDIA card is roughly 88% ahead in FP32 and 88% ahead in FP16 based on these specifications.

Q: What are the power requirements for each card?

A: The NVIDIA Quadro RTX 8000 has a 260 W TDP and a suggested PSU of 600 W. The AMD Radeon Pro W5700 has a 205 W TDP and a suggested PSU of 550 W. Both use dual-slot designs and require a 1x 6-pin plus 1x 8-pin power connector.

Q: Which card has better API support for modern games or applications?

A: The NVIDIA Quadro RTX 8000 supports DirectX 12 Ultimate (12_2), which includes features like hardware ray tracing via its 72 RT cores. The AMD Radeon Pro W5700 supports DirectX 12 (12_1) and lacks RT cores. Both cards support OpenGL 4.6 and Vulkan 1.4.

Q: Are these cards still in production?

A: No. Both are listed as end-of-life. The NVIDIA Quadro RTX 8000 was released in August 2018, and the AMD Radeon Pro W5700 was released in November 2019.

Specification Differences

The two cards differ across nearly every core specification. The NVIDIA Quadro RTX 8000 uses a 12 nm process and TU102 chip, while the AMD Radeon Pro W5700 uses a 7 nm process and Navi 10 chip. Transistor counts are 18,600 million versus 10,300 million, and die sizes are 754 mm² versus 251 mm². Transistor density is higher on AMD at 41.0M per mm² versus 24.7M per mm².

Shading units: 4,608 versus 2,304. TMUs: 288 versus 144. ROPs: 96 versus 64. The NVIDIA card has 72 RT cores and 576 tensor cores; the AMD card has none. Pixel rate is 169.9 GPixel/s versus 120.3 GPixel/s. Texture rate is 509.8 GTexel/s versus 270.7 GTexel/s. FP32 is 16.31 TFLOPS versus 8.663 TFLOPS. FP16 is 32.62 TFLOPS versus 17.33 TFLOPS.

Memory: 48 GB versus 8 GB GDDR6. Bus width: 384-bit versus 256-bit. Bandwidth: 672.0 GB/s versus 448.0 GB/s. Base clocks are nearly identical (1395 MHz versus 1400 MHz), but boost clocks differ (1770 MHz versus 1880 MHz). TDP: 260 W versus 205 W. Suggested PSU: 600 W versus 550 W. Bus interface: PCIe 3.0 x16 versus PCIe 4.0 x16. Display outputs: 4x DisplayPort plus 1x USB Type-C versus 5x mini-DisplayPort plus 1x USB Type-C. DirectX support: 12 Ultimate (12_2) versus 12 (12_1).

Head-to-Head Benchmarks

The Geekbench results show the largest NVIDIA advantage. In Geekbench OpenCL, the Quadro RTX 8000 scores 101,883 against 74,613 for AMD, a 36.5% lead. In Geekbench Vulkan, the gap widens to 73.4%, with scores of 122,637 versus 70,706. This indicates the NVIDIA card excels in both general compute and low-level graphics APIs.

Passmark results tell a similar story for modern DirectX versions. In DirectX 11, the Quadro RTX 8000 scores 188 versus 104, a 80.8% advantage, the largest margin in any test. DirectX 12 shows a 46.3% lead (79 versus 54), and DirectX 10 shows a 55.7% lead (137 versus 88). These results suggest the NVIDIA architecture is significantly more efficient at handling current-generation graphics APIs.

The Passmark G3D test, which measures overall 3D graphics performance, gives the Quadro RTX 8000 a score of 19,799 versus 14,520 for AMD, a 36.4% difference. Passmark GPU Compute shows a 53.8% lead for NVIDIA, with scores of 9,992 versus 6,495.

The AMD Radeon Pro W5700 claims two wins. In Passmark DirectX 9, it scores 225 versus 211, a 6.2% margin. In Passmark G2D, it scores 899 versus 866, a 3.7% margin. These wins are narrow compared to NVIDIA’s margins, which range from 36.4% to 80.8% in its seven victories.

Where Each One Wins

The NVIDIA Quadro RTX 8000 is the clear winner for compute-heavy professional workloads. Its 53.8% lead in Passmark GPU Compute, combined with 36.5% and 73.4% leads in Geekbench OpenCL and Vulkan, makes it the choice for rendering, simulation, machine learning inference, and any task that leverages FP32 or FP16 throughput. The presence of 576 tensor cores and 72 RT cores further supports AI and ray tracing workloads, though these are not directly benchmarked in the data. The 48 GB memory capacity is a decisive factor for large-scale data processing.

The AMD Radeon Pro W5700 wins in legacy DirectX 9 and 2D graphics tests. If a workflow relies on older software that uses DirectX 9, the AMD card’s 6.2% advantage in that specific test could matter. Its 3.7% lead in Passmark G2D indicates slightly better 2D acceleration, potentially useful for desktop compositing or CAD-like 2D views. Lower power draw (205 W versus 260 W) and PCIe 4.0 support are additional advantages for system builders prioritizing efficiency and modern bus bandwidth.

For all other scenarios, the data points decisively to the NVIDIA Quadro RTX 8000. It wins 7 of 9 benchmarks, with margins that far exceed AMD’s narrow victories. The average benchmark score difference of 10.5% and the percentile ranking (74th versus 71st) reinforce this conclusion. Users with modern DirectX 11/12 applications, Vulkan-based software, or general GPU compute should select the Quadro RTX 8000 without hesitation.