NVIDIA GeForce RTX 4090 D vs NVIDIA GeForce RTX 5090 SE Comparison

Head-to-Head Benchmarks

The recorded data for the NVIDIA GeForce RTX 4090 D shows a substantial performance lead over the RTX 5090 SE in the available benchmark suite. In 3DMark Steel Nomad DX12, the RTX 4090 D scores 8587 points, while the RTX 5090 SE has no recorded score in this test. The Geekbench OpenCL result follows the same pattern, with the RTX 4090 D reaching 278621 points against no recorded result for the RTX 5090 SE. The Geekbench Vulkan test shows the RTX 4090 D at 246941 points, again with no counterpart from the RTX 5090 SE.

Because the RTX 5090 SE has zero benchmark entries in the database, the head-to-head comparison relies entirely on the RTX 4090 D's measured results. The RTX 4090 D's average benchmark score stands at 178050 points, placing it in the 98th percentile among all GPUs tracked. The RTX 5090 SE holds a 50th percentile ranking with an average score of zero, indicating that no validated performance data has been recorded for this card yet.

The nearest rivals for the RTX 4090 D provide context for its standing. The NVIDIA RTX PRO 5000 Blackwell averages 182109 points, which is 2.2% higher than the RTX 4090 D. The NVIDIA A100 SXM4 80 GB scores 183725 points, 3.1% above. The NVIDIA RTX 5000 Ada Generation manages 184664 points, a 3.6% advantage. The NVIDIA A100 SXM4 40 GB completes the group at 187147 points, 4.9% ahead. These deltas show the RTX 4090 D trailing its nearest competitors by a narrow margin, between 2.2% and 4.9%.

Where Each One Wins

The RTX 4090 D wins every category where a measured benchmark exists. For 3DMark Steel Nomad DX12, a test that stresses DirectX 12 rendering workloads, the RTX 4090 D delivers a score of 8587. In compute-oriented tasks, Geekbench OpenCL and Vulkan results of 278621 and 246941 respectively confirm strong general-purpose and graphics API performance. The RTX 5090 SE has no recorded wins in any test, as the database contains no benchmark scores for this model.

The RTX 4090 D's advantage in raw throughput is supported by its specifications. It reaches 73.54 TFLOPS for both FP32 and FP16 operations, with a 1:1 ratio between the two. Pixel fill rate measures 443.5 GPixel/s and texture fill rate reaches 1,149.1 GTexel/s. These figures indicate a card optimized for high-frequency rendering and compute. The RTX 5090 SE, by contrast, lists 66.94 TFLOPS for FP32 and FP16, with a pixel rate of 380.3 GPixel/s and a texture rate of 1,045.9 GTexel/s. The RTX 4090 D is ahead in every throughput metric recorded.

For memory bandwidth, the RTX 4090 D uses GDDR6X across a 384-bit bus, achieving 1.01 TB/s. The RTX 5090 SE employs GDDR7 on the same 384-bit bus, reaching 1.34 TB/s. This 0.33 TB/s bandwidth advantage belongs to the RTX 5090 SE, though no benchmark data confirms its real-world impact. The RTX 5090 SE also operates with a higher TDP of 500 W compared to 425 W for the RTX 4090 D, suggesting greater power draw but no measured performance payoff in the database.

Architecture Differences

The two cards come from different NVIDIA generations. The RTX 4090 D uses the AD102 chip based on Ada Lovelace architecture, while the RTX 5090 SE uses the GB202 chip based on Blackwell 2.0 architecture. Both are manufactured on a 5 nm process at TSMC, but the transistor counts differ significantly. The RTX 4090 D packs 76,300 million transistors on a 609 mm² die, giving a density of 125.3 million transistors per mm². The RTX 5090 SE integrates 92,200 million transistors on a larger 750 mm² die, with a slightly lower density of 122.9 million per mm².

Core configuration also diverges. The RTX 4090 D has 14592 shading units, 456 texture mapping units, and 176 raster output units. It includes 114 ray tracing cores and 456 tensor cores. The RTX 5090 SE reduces these counts: 14080 shading units, 440 TMUs, 160 ROPs, 110 ray tracing cores, and 440 tensor cores. Despite the architectural update, the RTX 5090 SE carries fewer cores across every category.

Clock speeds favor the RTX 4090 D. Its base clock is 2280 MHz with a boost of 2520 MHz. The RTX 5090 SE runs at a 1740 MHz base and 2377 MHz boost. Memory clocks show the reverse: the RTX 4090 D operates at 1313 MHz with 21 Gbps effective, while the RTX 5090 SE runs at 1750 MHz with 28 Gbps effective. The higher memory clock on the RTX 5090 SE contributes to its bandwidth advantage.

Connectivity and physical design differ as well. The RTX 4090 D uses PCIe 4.0 x16, while the RTX 5090 SE moves to PCIe 5.0 x16. Display outputs vary: the RTX 4090 D offers 1x HDMI 2.1 and 3x DisplayPort 1.4a, while the RTX 5090 SE provides 1x HDMI 2.1b and 3x DisplayPort 2.1b. The RTX 5090 SE is a dual-slot card measuring 267 mm in length, 111 mm in height, and 40 mm in width. The RTX 4090 D is triple-slot, measuring 304 mm by 137 mm by 61 mm. Both use a single 16-pin power connector, but the suggested PSU differs: 800 W for the RTX 4090 D and 900 W for the RTX 5090 SE.

FAQ

Q: Which card has a higher average benchmark score?

A: The RTX 4090 D has an average benchmark score of 178050 points. The RTX 5090 SE has an average score of 0, as no benchmarks are recorded for it.

Q: How does the RTX 4090 D compare to its nearest rivals?

A: The RTX 4090 D trails the RTX PRO 5000 Blackwell by 2.2%, the A100 SXM4 80 GB by 3.1%, the RTX 5000 Ada Generation by 3.6%, and the A100 SXM4 40 GB by 4.9%.

Q: What memory types do the two cards use?

A: The RTX 4090 D uses 24 GB of GDDR6X, while the RTX 5090 SE uses 24 GB of GDDR7. Both have a 384-bit bus width.

Q: Which card has more shading units?

A: The RTX 4090 D has 14592 shading units, compared to 14080 on the RTX 5090 SE.

Q: What is the TDP difference between the two?

A: The RTX 4090 D has a TDP of 425 W, while the RTX 5090 SE has a TDP of 500 W.

Q: Are the DirectX and Vulkan versions the same?

A: Yes, both cards support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

The Verdict

The data points to the RTX 4090 D as the stronger performer in every measurable category. Its 73.54 TFLOPS FP32 output exceeds the RTX 5090 SE's 66.94 TFLOPS by 6.6 TFLOPS. Pixel rate of 443.5 GPixel/s versus 380.3 GPixel/s gives the RTX 4090 D a 63.2 GPixel/s advantage. Texture rate of 1,149.1 GTexel/s versus 1,045.9 GTexel/s adds a 103.2 GTexel/s lead. These are not marginal differences; they represent consistent advantages across throughput metrics.

The RTX 5090 SE does hold two specification wins. Its 1.34 TB/s memory bandwidth exceeds the RTX 4090 D's 1.01 TB/s, and its PCIe 5.0 interface is a generation newer. The GDDR7 memory and Blackwell 2.0 architecture signal newer technology, but without benchmark scores, the database cannot confirm whether these translate into actual performance gains. The RTX 5090 SE's 50th percentile ranking and zero average score leave its capabilities unverified.

For users selecting based on recorded data, the RTX 4090 D is the clear choice. It delivers higher compute throughput, faster pixel and texture processing, and a proven benchmark record in the 98th percentile. The RTX 5090 SE, while featuring newer memory and a smaller dual-slot footprint, has yet to demonstrate any measured performance. Its 500 W TDP and 900 W suggested PSU also impose higher system requirements without corresponding benchmark confirmation.

The RTX 4090 D's release date of December 27, 2023, places it earlier in the product cycle, but its production status is listed as end-of-life. The RTX 5090 SE is active and newer, with a release date of December 31, 2025. For immediate, verified performance, the RTX 4090 D stands alone. For future potential with newer architecture, the RTX 5090 SE remains an unproven option in the database.

Specification Differences

| Specification | RTX 4090 D | RTX 5090 SE |

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

| Chip | AD102 | GB202 |

| Architecture | Ada Lovelace | Blackwell 2.0 |

| Process Node | 5 nm | 5 nm |

| Transistors | 76,300 million | 92,200 million |

| Die Size | 609 mm² | 750 mm² |

| Transistor Density | 125.3M / mm² | 122.9M / mm² |

| Base Clock | 2280 MHz | 1740 MHz |

| Boost Clock | 2520 MHz | 2377 MHz |

| Memory Clock | 1313 MHz, 21 Gbps effective | 1750 MHz, 28 Gbps effective |

| Memory Type | GDDR6X | GDDR7 |

| Memory Size | 24 GB | 24 GB |

| Memory Bus | 384 bit | 384 bit |

| Memory Bandwidth | 1.01 TB/s | 1.34 TB/s |

| Shading Units | 14592 | 14080 |

| TMUs | 456 | 440 |

| ROPs | 176 | 160 |

| Ray Tracing Cores | 114 | 110 |

| Tensor Cores | 456 | 440 |

| Pixel Rate | 443.5 GPixel/s | 380.3 GPixel/s |

| Texture Rate | 1,149.1 GTexel/s | 1,045.9 GTexel/s |

| FP32 | 73.54 TFLOPS | 66.94 TFLOPS |

| FP16 | 73.54 TFLOPS (1:1) | 66.94 TFLOPS (1:1) |

| TDP | 425 W | 500 W |

| Slot Width | Triple-slot | Dual-slot |

| Power Connector | 1x 16-pin | 1x 16-pin |

| Suggested PSU | 800 W | 900 W |

| Bus Interface | PCIe 4.0 x16 | PCIe 5.0 x16 |

| Display Outputs | 1x HDMI 2.1, 3x DisplayPort 1.4a | 1x HDMI 2.1b, 3x DisplayPort 2.1b |

| DirectX | 12 Ultimate (12_2) | 12 Ultimate (12_2) |

| OpenGL | 4.6 | 4.6 |

| Vulkan | 1.4 | 1.4 |

| Length | 304 mm | 267 mm |

| Height | 137 mm | 111 mm |

| Width | 61 mm | 40 mm |

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

| Release Date | 2023-12-27 | 2025-12-31 |

| Predecessor | GeForce 30 | GeForce 40 |

| Successor | GeForce 50 | GeForce 60 |

| Launch MSRP | 1,599 USD | 1,499 USD |

| Percentile vs All GPUs | 98 | 50 |

| Average Benchmark Score | 178050 | 0 |