AMD Radeon PRO V620 vs AMD Radeon Pro W6800X Comparison

Head-to-Head Benchmarks

The benchmark data shows a narrow but consistent edge for the AMD Radeon PRO V620 over the AMD Radeon Pro W6800X, though the margin is far from decisive. In the only directly comparable benchmark, Geekbench OpenCL, the PRO V620 scores 128,580 against the W6800X’s 124,498, a 3.2% advantage. This is a modest win, but it establishes a pattern: the PRO V620 is the faster card in raw compute workloads, while the W6800X counters with broader API support and a higher overall percentile ranking.

Looking at the average benchmark scores, the gap widens. The W6800X posts an average score of 160,671, placing it in the 97th percentile of all GPUs. The PRO V620 averages 136,472, landing in the 96th percentile. This 17.7% difference in average scores is largely driven by the W6800X’s Geekbench Metal result of 196,844, a test the PRO V620 does not participate in. The PRO V620’s best result is its Geekbench Vulkan score of 144,364, which is 14.1% higher than its OpenCL score and demonstrates strong cross-API performance. The W6800X has no Vulkan benchmark in the data, so its Metal score stands as its sole high-water mark.

When placed against their nearest rivals, the two cards occupy different competitive tiers. The W6800X’s average score of 160,671 puts it 1.1% behind the NVIDIA A100 PCIe 40 GB (162,504), 2.6% behind the AMD Radeon PRO W7800 (164,894), and 2.8% behind the NVIDIA RTX A5500 (165,217). It is also 3.3% behind the NVIDIA RTX 4500 Ada Generation (166,094). These are tight margins, indicating the W6800X sits in a highly competitive performance band where a few percentage points separate it from several leading workstation cards. The PRO V620’s average of 136,472 places it a mere 0.5% ahead of the AMD Radeon Pro W6800X Duo (135,774), 0.8% ahead of the AMD Radeon PRO W6800 (135,396), and 0.9% ahead of both the NVIDIA A10M (135,230) and the NVIDIA RTX 4000 Ada Generation (135,218). This cluster is extremely dense, with the PRO V620 holding a razor-thin lead over four cards that are effectively performance equals.

Architecture Differences

Both cards share the same fundamental silicon: the Navi 21 chip built on TSMC’s 7 nm process, with 26,800 million transistors on a 520 mm² die, yielding a transistor density of 51.5M per mm². The architecture is RDNA 2.0 in both cases, and both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The differences emerge in how each card configures that silicon.

The Radeon PRO V620 is the more fully enabled part. It carries 4,608 shading units, 288 texture mapping units, and 128 render output units, with 72 ray tracing cores. The Radeon Pro W6800X is cut down in every compute resource: 3,840 shading units (a 16.7% reduction), 240 TMUs (a 16.7% reduction), 96 ROPs (a 25% reduction), and 60 ray tracing cores (a 16.7% reduction). This configuration difference directly explains the PRO V620’s higher peak throughput: 20.28 TFLOPS FP32 and 40.55 TFLOPS FP16 (2:1) versus the W6800X’s 16.03 TFLOPS FP32 and 32.06 TFLOPS FP16 (2:1). The PRO V620 is 26.5% ahead in FP32 and 26.5% ahead in FP16.

Clock speeds also favor the PRO V620. Its base clock is 1,825 MHz and boost clock is 2,200 MHz, while the W6800X runs at 1,800 MHz base and 2,087 MHz boost. The PRO V620’s boost clock is 5.4% higher than the W6800X’s. Memory is identical on paper: 32 GB of GDDR6 on a 256-bit bus, running at 2,000 MHz (16 Gbps effective), delivering 512.0 GB/s of bandwidth. The pixel rate and texture rate reflect the resource differences: the PRO V620 achieves 281.6 GPixel/s and 633.6 GTexel/s, while the W6800X manages 200.4 GPixel/s and 500.9 GTexel/s.

The power and physical profiles diverge sharply. The PRO V620 is a dual-slot card with a 300 W TDP, two 8-pin power connectors, and a suggested 700 W power supply. It uses a PCIe 4.0 x16 bus interface and has no display outputs, marking it as a compute-only accelerator. The W6800X is a quad-slot card with a 200 W TDP, an Apple MPX power connector, and a suggested 550 W power supply. It uses the Apple MPX bus interface and provides 1x HDMI 2.1 and 4x Thunderbolt outputs. Both cards share identical dimensions of 267 mm length and 120 mm height, but the W6800X is a quad-slot design while the PRO V620 is dual-slot and 50 mm wide.

FAQ

Q: Which card is faster in OpenCL compute workloads?

A: The AMD Radeon PRO V620 wins the Geekbench OpenCL test with a score of 128,580 against the W6800X’s 124,498, a 3.2% advantage.

Q: Does the W6800X have a benchmark advantage in any API?

A: Yes. The W6800X scores 196,844 in Geekbench Metal, a test the PRO V620 does not have a result for. This Metal score is the primary driver of the W6800X’s higher average benchmark score of 160,671 versus the PRO V620’s 136,472.

Q: What is the difference in shading units between the two cards?

A: The PRO V620 has 4,608 shading units, while the W6800X has 3,840, a difference of 768 units (16.7% fewer on the W6800X). The PRO V620 also has more TMUs, ROPs, and ray tracing cores.

Q: Do both cards use the same memory configuration?

A: Yes. Both have 32 GB of GDDR6 memory on a 256-bit bus with 512.0 GB/s bandwidth and 16 Gbps effective memory speed.

Q: Which card has a higher TDP and what are the power connector requirements?

A: The PRO V620 has a 300 W TDP and requires two 8-pin power connectors with a suggested 700 W power supply. The W6800X has a 200 W TDP, uses an Apple MPX power connector, and requires a suggested 550 W power supply.

Q: Can either card output video to displays?

A: The W6800X has display outputs (1x HDMI 2.1 and 4x Thunderbolt). The PRO V620 has no display outputs, making it a compute-only accelerator.

Specification Differences

| Specification | AMD Radeon Pro W6800X | AMD Radeon PRO V620 |

|---|---|---|

| Base Clock | 1800 MHz | 1825 MHz |

| Boost Clock | 2087 MHz | 2200 MHz |

| Shading Units | 3840 | 4608 |

| TMUs | 240 | 288 |

| ROPs | 96 | 128 |

| Ray Tracing Cores | 60 | 72 |

| Pixel Rate | 200.4 GPixel/s | 281.6 GPixel/s |

| Texture Rate | 500.9 GTexel/s | 633.6 GTexel/s |

| FP32 | 16.03 TFLOPS | 20.28 TFLOPS |

| FP16 | 32.06 TFLOPS (2:1) | 40.55 TFLOPS (2:1) |

| TDP | 200 W | 300 W |

| Slot Width | Quad-slot | Dual-slot |

| Power Connectors | Apple MPX | 2x 8-pin |

| Suggested PSU | 550 W | 700 W |

| Bus Interface | Apple MPX | PCIe 4.0 x16 |

| Display Outputs | 1x HDMI 2.1, 4x Thunderbolt | No outputs |

| Release Date | 2021-08-02 | 2021-11-03 |

| Predecessor | None | Radeon Pro Vega |

| Launch MSRP | 2,799 USD | None |

Where Each One Wins

The AMD Radeon PRO V620 wins on raw compute throughput. In every measured processing resource, it holds a clear edge: 26.5% more FP32 TFLOPS, 26.5% more FP16 TFLOPS, 25% more ROPs, and 20% more ray tracing cores. Its higher boost clock of 2,200 MHz versus 2,087 MHz contributes to a 40.5% higher pixel rate and a 26.5% higher texture rate. The OpenCL benchmark confirms this advantage in practice, with the PRO V620 scoring 3.2% higher. For workloads that rely on raw shader throughput, ray tracing, or compute density, the PRO V620 is the stronger choice.

The AMD Radeon Pro W6800X wins on versatility and integration. It offers display outputs (1x HDMI 2.1 and 4x Thunderbolt), making it a functional workstation card rather than a compute-only accelerator. Its Geekbench Metal score of 196,844 is the single highest benchmark result in this comparison, and its average benchmark score of 160,671 places it in the 97th percentile of all GPUs, one percentile higher than the PRO V620. The W6800X also consumes 33.3% less power (200 W versus 300 W) and requires a smaller power supply (550 W versus 700 W), making it easier to integrate into existing systems. Its Apple MPX bus interface and quad-slot design are tailored for Apple’s Mac Pro ecosystem, while the PRO V620’s PCIe 4.0 x16 interface and dual-slot form factor are more conventional for server or PC installations.

The benchmark gap in the nearest rival comparisons tells a similar story. The W6800X’s rivals are all higher-performing cards (NVIDIA A100, RTX A5500, RTX 4500 Ada), and it trails them by 1.1% to 3.3%. The PRO V620’s rivals are closely matched cards (W6800X Duo, PRO W6800, A10M, RTX 4000 Ada), and it leads them by 0.5% to 0.9%. This indicates the PRO V620 is a performance leader in its immediate tier, while the W6800X is a slightly trailing contender in a higher tier.

The Verdict

The data supports a clear but context-dependent verdict. If the priority is maximum compute performance in a dense, dual-slot server or PC form factor, the AMD Radeon PRO V620 is the better choice. It wins the OpenCL benchmark, has more shading units, TMUs, ROPs, and ray tracing cores, and delivers higher FP32 and FP16 throughput. Its 300 W TDP and 700 W PSU requirement are the cost of that performance, but for compute-only workloads, it is the more capable card.

If the priority is a workstation card with display output, lower power consumption, and broader API coverage, the AMD Radeon Pro W6800X is the better choice. Its Geekbench Metal score of 196,844 is the standout result in this comparison, and its 97th percentile ranking reflects a stronger overall position against all GPUs. The W6800X is also the only one of the two with any display outputs, making it suitable for interactive workloads, content creation, or any task requiring a visual output. Its 200 W TDP and 550 W PSU requirement make it more power-efficient, and its Apple MPX interface positions it specifically for Mac Pro systems.

The release timeline reinforces this split: the W6800X launched on 2021-08-02, three months before the PRO V620’s 2021-11-03 release. Both are now end-of-life products, but the PRO V620’s predecessor is the Radeon Pro Vega, while the W6800X has no listed predecessor. In terms of raw numbers, the PRO V620 is 26.5% ahead in FP32 and FP16, 3.2% ahead in OpenCL, and has a higher boost clock. The W6800X counters with a 44% higher Metal score, a 17.7% higher average benchmark score, and a one-percentile advantage in overall GPU ranking.

For compute-centric users who prioritize throughput and can live without display outputs, the PRO V620 is the clear winner. For users who need a functional workstation GPU with display connectivity, lower power draw, and a higher overall benchmark percentile, the W6800X is the more balanced option. The choice comes down to whether raw compute density or system integration matters more. The PRO V620 delivers more compute per slot; the W6800X delivers more utility per watt.