NVIDIA GeForce RTX 4090 D vs NVIDIA GeForce RTX 5090 Comparison

Head-to-Head Benchmarks

The recorded head-to-head data shows a clear, across-the-board performance advantage for the NVIDIA GeForce RTX 5090 in every shared benchmark. The most striking result is in the 3DMark Steel Nomad DX12 test, where the RTX 5090 scores 18,355 against the RTX 4090 D's 8,587. This represents a delta of -53.2% from the perspective of the older card, meaning the RTX 5090 is more than twice as fast in this specific workload. The margin is substantial and consistent with the generational leap in raw compute resources.

In the Geekbench Vulkan test, the RTX 5090 again takes a commanding lead, scoring 376,728 versus 246,941 for the RTX 4090 D, a delta of -34.5%. The OpenCL result follows the same pattern, with the RTX 5090 at 334,370 and the RTX 4090 D at 278,621, a delta of -16.7%. While the OpenCL gap is the smallest of the three, it still represents a significant and unambiguous win for the newer architecture. In total, the RTX 5090 wins all three head-to-head tests, leaving the RTX 4090 D with zero recorded victories in this comparison.

The average benchmark scores in the database further contextualize this gap, though they draw from different test sets. The RTX 4090 D holds an average score of 178,050, placing it in the 98th percentile of all GPUs. Its nearest rivals include the NVIDIA RTX PRO 5000 Blackwell at 182,109 (-2.2%), the NVIDIA A100 SXM4 80 GB at 183,725 (-3.1%), and the NVIDIA RTX 5000 Ada Generation at 184,664 (-3.6%). The RTX 5090's average score of 79,842 comes from a broader set of tests, including several Passmark entries, and places it in the 92nd percentile. Its nearest rivals are markedly different, with the NVIDIA Tesla P100 PCIe 16 GB at 79,605 (0.3% delta), the AMD Radeon RX 6850M XT at 78,940 (1.1%), and the AMD Radeon Pro Vega 64X at 80,959 (-1.4%). The divergent average scores and rival sets reflect the different workloads captured in each card's benchmark suite.

Where Each One Wins

Based on the head-to-head results, the RTX 5090 wins in every recorded category. In DX12 gaming workloads, represented by 3DMark Steel Nomad, the RTX 5090's advantage is decisive, nearly doubling the RTX 4090 D's score. This suggests a clear edge for the newer card in modern DirectX 12 titles, particularly those that stress geometry, ray tracing, and memory bandwidth.

For compute-oriented tasks, the Geekbench OpenCL and Vulkan results both favor the RTX 5090. The Vulkan gap is especially large at -34.5%, indicating that the RTX 5090's architecture scales better with Vulkan's low-level API overhead. The OpenCL margin is narrower but still solid, with the RTX 5090 outpacing the RTX 4090 D by a meaningful 16.7%. Users running OpenCL-based scientific or professional workloads will see a measurable improvement, while Vulkan-based applications will benefit even more.

The RTX 4090 D does not win any category in this head-to-head comparison. However, its own benchmark profile shows it remains a very capable card in absolute terms. Its 3DMark Steel Nomad score of 8,587 and Geekbench OpenCL score of 278,621 are strong numbers, and its 98th percentile rank places it among the top tier of all GPUs in the database. The issue is relative: the RTX 5090 simply outperforms it in every shared test, with no workload where the older card pulls ahead.

Architecture Differences

The two cards are built on different architectures and chips. The RTX 4090 D uses the AD102 chip based on Ada Lovelace, while the RTX 5090 uses the GB202 chip based on Blackwell 2.0. Both are manufactured on a 5 nm process at TSMC, but the silicon itself differs significantly in scale. The RTX 5090 packs 92,200 million transistors on a 750 mm² die, compared to 76,300 million transistors on a 609 mm² die for the RTX 4090 D. Interestingly, the transistor density is slightly higher on the older chip: 125.3M per mm² versus 122.9M per mm² for the RTX 5090, a reflection of the larger, less dense design of the newer GPU.

The RTX 5090's compute resources are substantially larger. It features 21,760 shading units, 680 texture mapping units, 170 ray tracing cores, and 680 tensor cores. The RTX 4090 D, by contrast, has 14,592 shading units, 456 TMUs, 114 RT cores, and 456 tensor cores. Both cards share the same 176 ROPs, so pixel throughput is comparable, but the RTX 5090's higher shading unit and TMU counts drive its texture rate to 1,636.8 GTexel/s versus 1,149.1 GTexel/s for the RTX 4090 D. The FP32 compute rating is 104.8 TFLOPS for the RTX 5090 and 73.54 TFLOPS for the RTX 4090 D, a 42% advantage for the newer card.

Memory is another major differentiator. The RTX 5090 comes with 32 GB of GDDR7 on a 512-bit bus, yielding 1.79 TB/s of bandwidth. The RTX 4090 D has 24 GB of GDDR6X on a 384-bit bus, with 1.01 TB/s of bandwidth. The newer card also runs its memory at a higher effective speed of 28 Gbps versus 21 Gbps. Clock speeds are slightly lower on the RTX 5090, with a base of 2017 MHz and boost of 2407 MHz, compared to 2280 MHz base and 2520 MHz boost on the RTX 4090 D. The RTX 5090 compensates with its larger core count and faster memory.

The cards also differ in power and physical design. The RTX 5090 has a TDP of 575 W and a recommended PSU of 950 W, while the RTX 4090 D is rated at 425 W with a recommended 800 W PSU. The RTX 5090 is a dual-slot card with a width of 40 mm, while the RTX 4090 D is triple-slot with a width of 61 mm. Both use a single 16-pin power connector, and both have the same length (304 mm) and height (137 mm). The RTX 5090 uses PCIe 5.0 x16, while the RTX 4090 D uses PCIe 4.0 x16. Display outputs differ as well: the RTX 5090 offers 1x HDMI 2.1b and 3x DisplayPort 2.1b, while the RTX 4090 D has 1x HDMI 2.1 and 3x DisplayPort 1.4a. Both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

FAQ

Q: Which card is faster in 3DMark Steel Nomad DX12?

A: The NVIDIA GeForce RTX 5090 is significantly faster, scoring 18,355 compared to the RTX 4090 D's 8,587, a delta of -53.2% from the older card's perspective.

Q: How does memory capacity differ between the two cards?

A: The RTX 5090 has 32 GB of GDDR7 memory on a 512-bit bus, while the RTX 4090 D has 24 GB of GDDR6X on a 384-bit bus. The RTX 5090's bandwidth is 1.79 TB/s versus 1.01 TB/s for the RTX 4090 D.

Q: What is the transistor count difference?

A: The RTX 5090 contains 92,200 million transistors on a 750 mm² die, while the RTX 4090 D contains 76,300 million transistors on a 609 mm² die. Both are built on a 5 nm TSMC process.

Q: Which card has more ray tracing cores?

A: The RTX 5090 has 170 RT cores, while the RTX 4090 D has 114 RT cores. The RTX 5090 also has more tensor cores: 680 versus 456.

Q: What are the power requirements?

A: The RTX 5090 has a TDP of 575 W and requires a recommended 950 W PSU. The RTX 4090 D has a TDP of 425 W and recommends an 800 W PSU. Both use a single 16-pin power connector.

Q: Do both cards support the same APIs?

A: Yes, both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. However, the RTX 5090 uses PCIe 5.0 x16, while the RTX 4090 D uses PCIe 4.0 x16.

Specification Differences

| Specification | NVIDIA GeForce RTX 4090 D | NVIDIA GeForce RTX 5090 |

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

| Architecture | Ada Lovelace | Blackwell 2.0 |

| Chip | AD102 | GB202 |

| 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 | 2017 MHz |

| Boost Clock | 2520 MHz | 2407 MHz |

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

| Memory Size | 24 GB | 32 GB |

| Memory Type | GDDR6X | GDDR7 |

| Memory Bus Width | 384 bit | 512 bit |

| Memory Bandwidth | 1.01 TB/s | 1.79 TB/s |

| Shading Units | 14,592 | 21,760 |

| TMUs | 456 | 680 |

| RT Cores | 114 | 170 |

| Tensor Cores | 456 | 680 |

| Pixel Rate | 443.5 GPixel/s | 423.6 GPixel/s |

| Texture Rate | 1,149.1 GTexel/s | 1,636.8 GTexel/s |

| FP32 Performance | 73.54 TFLOPS | 104.8 TFLOPS |

| FP16 Performance | 73.54 TFLOPS (1:1) | 104.8 TFLOPS (1:1) |

| TDP | 425 W | 575 W |

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

| Suggested PSU | 800 W | 950 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 |

| Width | 61 mm (2.4 inches) | 40 mm (1.6 inches) |

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

| Release Date | 2023-12-27 | 2025-01-29 |

| Launch MSRP | 1,599 USD | 1,999 USD |

The Verdict

The benchmark data is unambiguous: the NVIDIA GeForce RTX 5090 is the superior card in every shared test. Its 3DMark Steel Nomad score is more than double that of the RTX 4090 D, and its Geekbench Vulkan and OpenCL results are 34.5% and 16.7% higher, respectively. Users seeking maximum performance in DX12 gaming or Vulkan compute workloads should choose the RTX 5090 without hesitation. Its larger memory pool (32 GB versus 24 GB) and higher bandwidth (1.79 TB/s versus 1.01 TB/s) also make it the better choice for memory-intensive applications like high-resolution texture loading or large dataset processing.

The RTX 4090 D, despite losing all head-to-head tests, remains a strong card in its own right. Its 98th percentile ranking and average score of 178,050 place it well above the vast majority of GPUs in the database. It is a viable option for users who already own one or find it at a lower price point in the used market, but the data shows no workload where it beats the RTX 5090. Its lower TDP of 425 W and triple-slot width mean it may fit in systems with less robust power supplies, but the RTX 5090's dual-slot design and higher performance make it the more compelling new purchase.

The RTX 5090's launch MSRP is 1,999 USD, while the RTX 4090 D launched at 1,599 USD. The price difference is notable, but the performance gap is larger in relative terms, especially in the most demanding workloads. For anyone building a new high-end system, the RTX 5090 is the clear choice based on the recorded measurements.