AMD Radeon Pro W6600X vs NVIDIA GeForce RTX 5090 D Comparison

Head-to-Head Benchmarks

The benchmark data for these two GPUs does not contain any overlapping test results, making a direct apples-to-apples comparison impossible. The AMD Radeon Pro W6600X has a single recorded benchmark score, while the NVIDIA GeForce RTX 5090 D has ten scores across different tests. Consequently, the head-to-head table is empty, and neither card registers a win in a shared benchmark category.

What can be compared is the overall standing of each product within the database. The Radeon Pro W6600X achieves an average benchmark score of 107342, placing it in the 94th percentile of all GPUs tracked. Its nearest rival, the AMD Radeon Pro Vega II Duo, scores 106750, putting the W6600X 0.6% ahead. The W6600X also leads the NVIDIA Quadro RTX 6000 by 5.4%, with that rival scoring 101872. However, the W6600X trails two other AMD workstation cards: the Radeon Pro Vega II scores 109617 (2.1% higher), and the Radeon PRO W7900 scores 110725 (3.1% higher). These figures show the W6600X sits in a competitive mid-to-high tier for its generation, but it is not the strongest AMD professional card.

The RTX 5090 D presents a different profile. Its average benchmark score across all listed tests is 77712, placing it in the 92nd percentile. Its nearest rival is the AMD Radeon RX 6650M XT, which scores 76904 — the RTX 5090 D leads by 1.1%. The RTX 5090 D also edges out the AMD Radeon RX 6850M XT (78940, a 1.6% deficit) and the NVIDIA Tesla P100 PCIe 12 GB (79396, a 2.1% deficit). The Tesla P100 PCIe 16 GB is 2.4% ahead at 79605. Notably, the RTX 5090 D’s percentile rank (92nd) is slightly lower than the W6600X’s (94th), despite the RTX 5090 D being a far more powerful card in absolute terms. This discrepancy arises because the RTX 5090 D’s average is dragged down by low scores in legacy DirectX 9, 10, and 11 tests, where it scores 434, 231, and 371 respectively.

Looking at the RTX 5090 D’s individual results, the strongest showing is in Geekbench Vulkan, where it scores 376915. Its Geekbench OpenCL score is 310674, and its Passmark G3D score is 44065. In compute-heavy workloads, the Passmark GPU Compute score is 28396. The weakest results appear in older DirectX tests, with DirectX 12 scoring 219 and DirectX 10 scoring 231. These numbers indicate that the RTX 5090 D excels in modern API and compute scenarios but suffers in legacy rasterization tests.

The W6600X, by contrast, has only a Geekbench Metal score of 107342. This test is specific to Apple’s Metal API, reflecting the card’s intended use in Mac Pro systems. Without a comparable Metal score for the RTX 5090 D, cross-referencing the two cards requires relying on their percentile ranks and nearest-rival deltas rather than direct head-to-head results.

Architecture Differences

The architectural gap between these two GPUs is substantial, reflecting their different release eras and target markets. The AMD Radeon Pro W6600X uses the Navi 23 chip, built on RDNA 2.0 architecture, and manufactured on a 7 nm process at TSMC. It contains 11,060 million transistors on a 237 mm² die, yielding a transistor density of 46.7 million per mm². The NVIDIA GeForce RTX 5090 D uses the GB202 chip, built on Blackwell 2.0 architecture, and manufactured on a 5 nm process, also at TSMC. It packs 92,200 million transistors on a 750 mm² die, with a transistor density of 122.9 million per mm². The RTX 5090 D has roughly 8.3 times more transistors on a die about 3.2 times larger, and its density is 2.6 times higher.

Memory subsystems differ sharply. The W6600X has 8 GB of GDDR6 memory on a 128-bit bus, delivering 256.0 GB/s of bandwidth. The RTX 5090 D has 32 GB of GDDR7 memory on a 512-bit bus, delivering 1.79 TB/s. That is four times the capacity, four times the bus width, and seven times the bandwidth. Memory clocks also differ: the W6600X runs at 2000 MHz (16 Gbps effective), while the RTX 5090 D runs at 1750 MHz (28 Gbps effective). The higher effective data rate on the RTX 5090 D compensates for its lower base clock through the wider bus.

Compute resources are vastly different. The W6600X has 2048 shading units, 128 texture mapping units, and 64 raster output units. It also has 32 ray tracing cores. The RTX 5090 D has 21760 shading units, 680 TMUs, and 176 ROPs, along with 170 ray tracing cores and 680 tensor cores. The RTX 5090 D’s shading unit count is over 10 times higher, and its TMU count is over 5 times higher. The RTX 5090 D’s tensor cores, absent from the W6600X, enable AI-accelerated workloads that the AMD card cannot perform via dedicated hardware.

Clock speeds are similar. The W6600X has a base clock of 2068 MHz and a boost clock of 2479 MHz. The RTX 5090 D has a base clock of 2017 MHz and a boost clock of 2407 MHz. The AMD card is slightly faster in both metrics, but the difference is small (2.5% base, 3.0% boost). However, the sheer scale of the RTX 5090 D’s compute units dwarfs the clock advantage.

Pixel and texture rates reflect the resource disparity. The W6600X outputs 158.7 GPixel/s and 317.3 GTexel/s. The RTX 5090 D outputs 423.6 GPixel/s and 1,636.8 GTexel/s. The RTX 5090 D is 2.7 times faster in pixel throughput and 5.2 times faster in texture throughput. FP32 performance is 10.15 TFLOPS for the W6600X versus 104.8 TFLOPS for the RTX 5090 D — a 10.3-fold difference. FP16 performance is 20.31 TFLOPS (2:1 ratio) for the W6600X, while the RTX 5090 D achieves 104.8 TFLOPS with a 1:1 ratio, meaning it does not double FP16 throughput but still matches its FP32 rate.

Power and interface differences are equally stark. The W6600X has a TDP of 120 W and a suggested PSU of 300 W, while the RTX 5090 D has a TDP of 575 W and a suggested PSU of 950 W. The W6600X uses an Apple MPX bus interface with no display outputs, while the RTX 5090 D uses PCIe 5.0 x16 and has 1x HDMI 2.1b and 3x DisplayPort 2.1b outputs. The RTX 5090 D is a dual-slot card measuring 304 mm in length, 137 mm in height, and 48 mm in width, with a 1x 16-pin power connector. The W6600X is also dual-slot but has no listed dimensions or power connector details.

The Verdict

The data clearly separates these two GPUs into different categories. The AMD Radeon Pro W6600X is an end-of-life workstation card designed for Apple MPX systems, with no display outputs and a Geekbench Metal score of 107342. It ranks in the 94th percentile, slightly ahead of the Radeon Pro Vega II Duo and Quadro RTX 6000, but behind the Radeon Pro Vega II and PRO W7900. Its 120 W TDP and 300 W suggested PSU make it a low-power option suited to Mac Pro environments where the host system handles display output.

The NVIDIA GeForce RTX 5090 D is an active, current-generation product in the GeForce 50-series, built on Blackwell 2.0. Its average score of 77712 across multiple tests places it in the 92nd percentile, but that average is heavily influenced by weak legacy DirectX scores. In modern workloads like Geekbench Vulkan (376915) and OpenCL (310674), it is exceptionally strong. Its nearest rivals are mobile Radeon parts (RX 6650M XT and RX 6850M XT) and older Tesla P100 accelerators, which is a reflection of the mixed benchmark suite rather than the card’s true capability.

For a user selecting a GPU for an Apple Mac Pro, the W6600X is the only viable choice between these two, given its Apple MPX interface and lack of display outputs. It also consumes far less power, with a 575 W TDP difference. However, for any workload requiring modern API support, massive memory bandwidth, or AI acceleration, the RTX 5090 D is overwhelmingly superior in raw throughput, with 10.3 times the FP32 performance and 7 times the memory bandwidth.

The 94th versus 92nd percentile ranking is misleading. The W6600X’s single Metal score benefits from a focused test, while the RTX 5090 D’s average is penalized by legacy DirectX results that are not relevant to its target use cases. In practice, the RTX 5090 D is the higher-performance card by any meaningful metric, but it is not a replacement for the W6600X in Apple-specific systems.

Specification Differences

| Specification | AMD Radeon Pro W6600X | NVIDIA GeForce RTX 5090 D |

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

| Architecture | RDNA 2.0 | Blackwell 2.0 |

| Process Node | 7 nm | 5 nm |

| Transistors | 11,060 million | 92,200 million |

| Die Size | 237 mm² | 750 mm² |

| Transistor Density | 46.7M / mm² | 122.9M / mm² |

| Base Clock | 2068 MHz | 2017 MHz |

| Boost Clock | 2479 MHz | 2407 MHz |

| Memory Size | 8 GB | 32 GB |

| Memory Type | GDDR6 | GDDR7 |

| Memory Bus Width | 128 bit | 512 bit |

| Memory Bandwidth | 256.0 GB/s | 1.79 TB/s |

| Memory Clock | 2000 MHz (16 Gbps effective) | 1750 MHz (28 Gbps effective) |

| Shading Units | 2048 | 21760 |

| TMUs | 128 | 680 |

| ROPs | 64 | 176 |

| RT Cores | 32 | 170 |

| Tensor Cores | N/A | 680 |

| Pixel Rate | 158.7 GPixel/s | 423.6 GPixel/s |

| Texture Rate | 317.3 GTexel/s | 1,636.8 GTexel/s |

| FP32 | 10.15 TFLOPS | 104.8 TFLOPS |

| FP16 | 20.31 TFLOPS (2:1) | 104.8 TFLOPS (1:1) |

| TDP | 120 W | 575 W |

| Suggested PSU | 300 W | 950 W |

| Bus Interface | Apple MPX | PCIe 5.0 x16 |

| Display Outputs | No outputs | 1x HDMI 2.1b, 3x DisplayPort 2.1b |

| Power Connectors | N/A | 1x 16-pin |

| Dimensions | N/A | 304 mm x 137 mm x 48 mm |

| Production Status | End-of-life | Active |

| Release Date | 2021-08-02 | 2025-01-29 |

| Launch MSRP | 699 USD | 2,299 USD |

FAQ

Q: Which GPU has the higher percentile ranking?

A: The AMD Radeon Pro W6600X ranks in the 94th percentile of all GPUs, while the NVIDIA GeForce RTX 5090 D ranks in the 92nd percentile.

Q: What is the memory capacity difference?

A: The RTX 5090 D has 32 GB of GDDR7 memory, which is four times the 8 GB of GDDR6 memory in the W6600X.

Q: How do the FP32 performance figures compare?

A: The RTX 5090 D delivers 104.8 TFLOPS of FP32 performance, which is 10.3 times the 10.15 TFLOPS of the W6600X.

Q: Does the W6600X have tensor cores?

A: No, the W6600X has no tensor cores. The RTX 5090 D has 680 tensor cores.

Q: What is the TDP of each card?

A: The W6600X has a TDP of 120 W with a suggested PSU of 300 W. The RTX 5090 D has a TDP of 575 W with a suggested PSU of 950 W.

Q: What display outputs does the W6600X offer?

A: The W6600X has no display outputs. The RTX 5090 D has 1x HDMI 2.1b and 3x DisplayPort 2.1b outputs.