NVIDIA GeForce RTX 4090 Mobile vs NVIDIA GeForce RTX 5090 Mobile Comparison

The benchmark data is unambiguous: the NVIDIA GeForce RTX 5090 Mobile wins every single head-to-head comparison against the RTX 4090 Mobile, taking all 9 tests with deltas ranging from a modest 2.7% to a commanding 29%. Despite the RTX 4090 Mobile's higher boost clock and slightly higher raw FP32 throughput, the newer Blackwell architecture, faster GDDR7 memory, and larger 24 GB frame buffer give the RTX 5090 Mobile a sweeping victory across compute, DirectX, and OpenCL workloads. The average benchmark score of 45,152 for the RTX 5090 Mobile versus 43,667 for the RTX 4090 Mobile places both in the 84th percentile of all GPUs, but the former does so with a decisive 3.4% aggregate advantage.

Where Each One Wins

The RTX 5090 Mobile wins everywhere, but the margin varies by workload type, revealing where its architecture pays off most. The largest victory comes in PassMark DirectX 12, where the RTX 5090 Mobile scores 138 against 107 for the RTX 4090 Mobile — a 29% lead that signals a major generational uplift in modern API efficiency. This is not a marginal improvement; it is a leap that suggests the Blackwell 2.0 architecture handles low-level graphics workloads substantially better than Ada Lovelace.

Compute-oriented tests also favor the RTX 5090 Mobile strongly. In Geekbench Vulkan, it scores 198,405 versus 170,774, a 16.2% advantage, while Geekbench OpenCL shows an 11.6% lead (201,834 vs 180,831). These are the tests that stress raw throughput and memory bandwidth, and the RTX 5090 Mobile's 896.0 GB/s GDDR7 bandwidth versus 576.0 GB/s GDDR6 appears to be the decisive factor. The RTX 5090 Mobile also wins PassMark GPU Compute by 8.5% (13,401 vs 12,347), confirming that the advantage extends beyond gaming into general-purpose compute.

The RTX 4090 Mobile's only potential claim is in legacy DirectX 9 and DirectX 10 scenarios, but even there it loses. The RTX 5090 Mobile leads DirectX 9 by 4.5% (324 vs 310) and DirectX 10 by 5.8% (183 vs 173). The closest contest is DirectX 11, where the RTX 5090 Mobile wins by just 2.7% (269 vs 262), suggesting that older APIs narrow the gap because neither architecture fully exploits its modern features under those paths. Still, the pattern is clear: the RTX 5090 Mobile is the faster GPU in every measured scenario, with the largest deltas in the most demanding, modern workloads.

Architecture Differences

The two GPUs represent distinct generations built on different architectures. The RTX 5090 Mobile uses the GB203 chip on Blackwell 2.0, while the RTX 4090 Mobile uses the AD103 chip on Ada Lovelace. Both are fabricated on a 5 nm process at TSMC, and their transistor counts are nearly identical — 45,600 million for the RTX 5090 Mobile versus 45,900 million for the RTX 4090 Mobile — with die sizes of 378 mm² and 379 mm² respectively. The transistor density is also close: 120.6M / mm² for the newer part versus 121.1M / mm² for the older one. The architectural differences manifest in the core configuration and memory subsystem rather than the silicon itself.

The RTX 5090 Mobile packs 10,496 shading units, 328 TMUs, 82 RT cores, and 328 tensor cores, compared to the RTX 4090 Mobile's 9,728 shading units, 304 TMUs, 76 RT cores, and 304 tensor cores. That is a 7.9% increase in shading units and an 8.2% increase in ray tracing cores. Both have 112 ROPs. The memory configuration diverges sharply: the RTX 5090 Mobile has 24 GB of GDDR7 on a 256-bit bus delivering 896.0 GB/s, while the RTX 4090 Mobile has 16 GB of GDDR6 on the same 256-bit bus at 576.0 GB/s. Memory bandwidth is 55.6% higher on the newer card, a massive advantage for texture-heavy and compute-bound scenes.

Clock speeds tell a different story. The RTX 4090 Mobile has a higher base clock (1335 MHz vs 990 MHz) and a higher boost clock (1695 MHz vs 1515 MHz). This partially explains why the RTX 4090 Mobile retains a slight edge in theoretical FP32 throughput: 32.98 TFLOPS versus 31.80 TFLOPS. The RTX 5090 Mobile compensates with more cores and faster memory, but its lower clocks and lower TDP of 95 W versus 120 W mean it achieves its wins through efficiency and bandwidth rather than raw frequency. The RTX 5090 Mobile also uses PCIe 5.0 x16 versus PCIe 4.0 x16 on the RTX 4090 Mobile, though both are integrated into portable devices with no external power connectors.

Head-to-Head Benchmarks

The most striking result is the PassMark DirectX 12 test, where the RTX 5090 Mobile scores 138 versus 107 for the RTX 4090 Mobile, a 29% delta. This is the single largest margin in the entire comparison and a strong indicator that Blackwell's optimizations for modern graphics APIs are substantial. DirectX 11, by contrast, shows only a 2.7% gap (269 vs 262), suggesting that legacy API overhead diminishes the architectural advantages.

Geekbench results reinforce the compute narrative. In Vulkan, the RTX 5090 Mobile leads by 16.2% (198,405 vs 170,774), and in OpenCL by 11.6% (201,834 vs 180,831). These are the highest absolute scores in the dataset, and the deltas indicate that the RTX 5090 Mobile's memory bandwidth and core count translate directly into compute performance. PassMark GPU Compute shows a smaller but still clear 8.5% lead (13,401 vs 12,347).

The PassMark G3D test, which represents overall 3D gaming performance, gives the RTX 5090 Mobile a 10.4% win (30,034 vs 27,212). That is the headline gaming metric, and it is consistent with the DirectX 12 result in showing a solid generational improvement. The 2D test (PassMark G2D) also favors the RTX 5090 Mobile at 7.4% (1,057 vs 984). Legacy API tests show the smallest margins: DirectX 9 at 4.5% (324 vs 310) and DirectX 10 at 5.8% (183 vs 173). Across all nine benchmarks, the RTX 5090 Mobile wins every one, with an average delta of 10.7% across the suite.

FAQ

Q: Which GPU has more memory bandwidth?

A: The NVIDIA GeForce RTX 5090 Mobile has 896.0 GB/s of bandwidth from 24 GB of GDDR7 on a 256-bit bus, versus the RTX 4090 Mobile's 576.0 GB/s from 16 GB of GDDR6 on the same bus width.

Q: Does the RTX 4090 Mobile have a higher clock speed?

A: Yes, the RTX 4090 Mobile has a base clock of 1335 MHz and a boost clock of 1695 MHz, while the RTX 5090 Mobile operates at 990 MHz base and 1515 MHz boost.

Q: How much faster is the RTX 5090 Mobile in DirectX 12?

A: The RTX 5090 Mobile scores 138 in PassMark DirectX 12 versus 107 for the RTX 4090 Mobile, a 29% advantage.

Q: Which GPU has more ray tracing cores?

A: The RTX 5090 Mobile has 82 RT cores, compared to 76 RT cores on the RTX 4090 Mobile.

Q: What is the average benchmark score difference?

A: The RTX 5090 Mobile averages 45,152 across all benchmarks, while the RTX 4090 Mobile averages 43,667, giving the newer card a 3.4% higher aggregate score.

Q: Are both GPUs in the same performance percentile?

A: Yes, both GPUs sit in the 84th percentile of all GPUs, indicating they are closely matched overall despite the RTX 5090 Mobile's consistent wins in individual tests.

The Verdict

The data points to a single conclusion: the RTX 5090 Mobile is the superior GPU across every benchmark tested. It wins all nine head-to-head comparisons, with the largest margins in modern DirectX 12 (29%) and Vulkan (16.2%) workloads, and the smallest in legacy DirectX 11 (2.7%). The RTX 4090 Mobile's higher clock speeds and slightly higher FP32 throughput (32.98 TFLOPS vs 31.80 TFLOPS) do not translate into real-world wins; the RTX 5090 Mobile's additional shading units, RT cores, and dramatically higher memory bandwidth dominate where it matters.

For gaming, the RTX 5090 Mobile's 10.4% lead in PassMark G3D and 29% lead in DirectX 12 make it the clear choice for modern titles that leverage current APIs. For compute workloads, the 11.6% OpenCL and 8.5% GPU Compute leads reinforce its versatility. The RTX 5090 Mobile also does this at a lower TDP of 95 W versus 120 W, meaning it achieves better performance with less power draw. The only reason to consider the RTX 4090 Mobile is if 16 GB of memory is sufficient and the older platform is already owned, but from a pure performance standpoint, the RTX 5090 Mobile is the definitive winner.

Specification Differences

| Specification | NVIDIA GeForce RTX 5090 Mobile | NVIDIA GeForce RTX 4090 Mobile |

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

| Architecture | Blackwell 2.0 | Ada Lovelace |

| Chip | GB203 | AD103 |

| Process Node | 5 nm | 5 nm |

| Transistors | 45,600 million | 45,900 million |

| Die Size | 378 mm² | 379 mm² |

| Base Clock | 990 MHz | 1335 MHz |

| Boost Clock | 1515 MHz | 1695 MHz |

| Memory Size | 24 GB | 16 GB |

| Memory Type | GDDR7 | GDDR6 |

| Memory Bandwidth | 896.0 GB/s | 576.0 GB/s |

| Shading Units | 10496 | 9728 |

| TMUs | 328 | 304 |

| RT Cores | 82 | 76 |

| Tensor Cores | 328 | 304 |

| FP32 | 31.80 TFLOPS | 32.98 TFLOPS |

| TDP | 95 W | 120 W |

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

| Release Date | 2025-03-26 | 2023-01-02 |