AMD Radeon RX 6800 XT vs NVIDIA GeForce RTX 5090 Mobile Comparison

The AMD Radeon RX 6800 XT and NVIDIA GeForce RTX 5090 Mobile represent two very different approaches to high-end graphics, separated by nearly half a decade of silicon evolution. The data shows a clear overall victor in the RTX 5090 Mobile, which wins all ten head-to-head benchmark comparisons, but the context of each card’s design—a 300 W desktop behemoth versus a 95 W mobile IGP—makes the performance gap more nuanced than a simple scoreboard. This analysis breaks down where each card excels, the architectural gulf between them, and what the benchmark numbers actually mean for a prospective buyer.

Where Each One Wins

The RTX 5090 Mobile wins every single benchmark in the comparison, but the margin of victory varies wildly by workload. Its largest advantage comes in modern DirectX 12 scenarios, where it leads by 37.2% in 3DMark Steel Nomad. This is the headline result: in a current-generation gaming workload, the laptop chip is decisively faster than the desktop card. The NVIDIA part also dominates Vulkan performance, posting a 31% higher score in Geekbench Vulkan, and it holds a 27.5% lead in Passmark DirectX 12. These are the metrics that matter most for contemporary AAA gaming and forward-looking titles.

The RX 6800 XT’s best showing is in compute-heavy legacy workloads. In Passmark GPU Compute, the desktop card trails by only 1.2%, and in Passmark DirectX 11 the deficit shrinks to 1.9%. These are the areas where the older AMD card’s raw shader throughput and 512.0 GB/s of bandwidth keep it competitive. The gap in Passmark G2D is similarly narrow at 2.6%, indicating that 2D desktop acceleration is effectively a wash. However, these are not the workloads that define a high-end GPU purchase. The RTX 5090 Mobile’s clean sweep, combined with its crushing leads in modern APIs, means it wins the only categories that most gamers and creators will care about.

Architecture Differences

The architectural divide here is generational and fundamental. The RX 6800 XT uses the Navi 21 chip on RDNA 2.0, built on a 7 nm TSMC process with 26,800 million transistors spread across a 520 mm² die. The RTX 5090 Mobile uses the GB203 chip on Blackwell 2.0, fabricated on a 5 nm process with 45,600 million transistors packed into a 378 mm² die. The transistor density tells the story: the NVIDIA chip packs 120.6M transistors per mm² versus 51.5M for AMD, a 2.3x density advantage that explains how it fits more hardware into a smaller footprint.

The memory subsystems diverge sharply. AMD pairs 16 GB of GDDR6 on a 256-bit bus for 512.0 GB/s of bandwidth. NVIDIA counters with 24 GB of GDDR7 on the same 256-bit bus, delivering 896.0 GB/s. That is a 75% bandwidth advantage for the mobile part, which is critical for 4K textures and compute workloads. The RTX 5090 Mobile also has a vastly different compute configuration: 10496 shading units, 328 TMUs, and 328 tensor cores, versus the RX 6800 XT’s 4608 shading units, 288 TMUs, and no tensor cores. The RT core counts are closer (82 vs 72), but NVIDIA’s tensor core array gives it a massive edge in AI-accelerated features.

Clock speeds and power envelopes flip the script. The RX 6800 XT runs at a 1825 MHz base and 2250 MHz boost, with a 300 W TDP. The RTX 5090 Mobile operates at a much lower 990 MHz base and 1515 MHz boost, but its TDP is just 95 W. This is the key practical difference: the desktop card needs a 700 W suggested PSU and dual 8-pin connectors, while the mobile part is an IGP with no power connectors. The NVIDIA chip achieves its performance at less than one-third the power draw, a signal of the efficiency gains of the newer node and architecture.

Head-to-Head Benchmarks

The 3DMark Steel Nomad DX12 test is the most lopsided result. The RX 6800 XT scores 3689, while the RTX 5090 Mobile hits 5871, a 37.2% delta. This is the single largest gap in the entire comparison and signals that the mobile part is the better choice for future-proof DirectX 12 gaming. The delta is so large that it cannot be explained by driver maturity or test variance; it reflects a genuine architectural superiority in modern rendering paths.

Geekbench results show a similar but less extreme pattern. In Vulkan, the RTX 5090 Mobile scores 198405 versus 136875 for AMD, a 31% advantage. In OpenCL, the gap narrows to 15.1% (201834 vs 171304). These compute-oriented tests highlight the NVIDIA part’s strength in parallel workloads, though the OpenCL result shows that the RX 6800 XT’s raw FP32 throughput (20.74 TFLOPS) still keeps it respectable.

Passmark results paint a more complex picture. The DirectX 12 test shows a 27.5% lead for NVIDIA (138 vs 100), reinforcing the modern API trend. DirectX 11 is nearly tied at 269 vs 264 (1.9% delta), suggesting that older games and engines run similarly on both. DirectX 9 and DirectX 10 show NVIDIA leading by 18.2% and 15.3% respectively, which is counterintuitive for a newer architecture but consistent across the board. The G3D score, a general gaming metric, has NVIDIA ahead by 16.8% (30034 vs 24980), while G2D is nearly identical at 1057 vs 1030. Compute is the closest result: 13401 vs 13245, a mere 1.2% delta, meaning the RX 6800 XT remains a viable option for pure compute tasks.

FAQ

Q: Is the RTX 5090 Mobile faster in every benchmark?

A: Yes. The data shows the NVIDIA GeForce RTX 5090 Mobile wins all 10 head-to-head comparisons, with deltas ranging from 1.2% in Passmark GPU Compute to 37.2% in 3DMark Steel Nomad DX12.

Q: How do their memory configurations compare?

A: The RX 6800 XT has 16 GB of GDDR6 on a 256-bit bus with 512.0 GB/s bandwidth. The RTX 5090 Mobile has 24 GB of GDDR7 on a 256-bit bus with 896.0 GB/s bandwidth, a 75% bandwidth advantage.

Q: Which card has better compute performance?

A: The RTX 5090 Mobile wins in Geekbench OpenCL (201834 vs 171304, a 15.1% lead) and Passmark GPU Compute (13401 vs 13245, a 1.2% lead), but the RX 6800 XT is much closer in the latter test.

Q: What is the power consumption difference?

A: The RX 6800 XT has a 300 W TDP and requires a 700 W suggested PSU with dual 8-pin connectors. The RTX 5090 Mobile has a 95 W TDP and is an IGP with no power connectors.

Q: Are they in the same performance percentile?

A: The RX 6800 XT sits at the 85th percentile of all GPUs, while the RTX 5090 Mobile is at the 84th percentile, despite the NVIDIA part winning all head-to-head tests.

Q: Does the RX 6800 XT have any tensor cores?

A: No. The RX 6800 XT has no tensor cores, while the RTX 5090 Mobile has 328 tensor cores, giving it a significant advantage in AI-accelerated workloads.

Specification Differences

| Specification | AMD Radeon RX 6800 XT | NVIDIA GeForce RTX 5090 Mobile |

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

| Architecture | RDNA 2.0 | Blackwell 2.0 |

| Process Node | 7 nm | 5 nm |

| Transistors | 26,800 million | 45,600 million |

| Die Size | 520 mm² | 378 mm² |

| Base Clock | 1825 MHz | 990 MHz |

| Boost Clock | 2250 MHz | 1515 MHz |

| Memory Size | 16 GB | 24 GB |

| Memory Type | GDDR6 | GDDR7 |

| Memory Bandwidth | 512.0 GB/s | 896.0 GB/s |

| Shading Units | 4608 | 10496 |

| TMUs | 288 | 328 |

| ROPs | 128 | 112 |

| RT Cores | 72 | 82 |

| Tensor Cores | None | 328 |

| FP32 Performance | 20.74 TFLOPS | 31.80 TFLOPS |

| FP16 Performance | 41.47 TFLOPS (2:1) | 31.80 TFLOPS (1:1) |

| TDP | 300 W | 95 W |

| Slot Width | Dual-slot | IGP |

| Power Connectors | 2x 8-pin | None |

| Suggested PSU | 700 W | None |

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

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

| Release Date | 2020-10-27 | 2025-03-26 |

The Verdict

The data is unambiguous: the NVIDIA GeForce RTX 5090 Mobile is the faster GPU in every measured workload. Its 37.2% lead in 3DMark Steel Nomad and 31% lead in Geekbench Vulkan make it the clear choice for modern gaming and compute-heavy applications. The 24 GB of GDDR7 memory at 896.0 GB/s provides headroom that the RX 6800 XT cannot match, and the 328 tensor cores enable AI features that the AMD card lacks entirely.

However, the verdict is not purely about raw speed. The RX 6800 XT remains a capable desktop card, particularly in legacy APIs where its deficit shrinks to under 2%. Its 85th percentile ranking versus the RTX 5090 Mobile’s 84th suggests that in the broader GPU landscape, they are closer than the head-to-head results imply. The RX 6800 XT also offers a dual-slot form factor with standard display outputs (1x HDMI 2.1, 2x DisplayPort 1.4a, 1x USB Type-C), whereas the RTX 5090 Mobile is an IGP with display outputs that are portable device dependent.

For a desktop builder with an existing ATX system, the RX 6800 XT is a proven, end-of-life part that still delivers solid performance in older titles and compute tasks. For anyone building a new system or prioritizing modern DirectX 12 and Vulkan performance, the RTX 5090 Mobile’s superior scores, lower power draw, and active production status make it the rational pick. The 95 W TDP versus 300 W TDP is a decisive practical advantage that cannot be overstated. The RTX 5090 Mobile is the better GPU by every measurable benchmark, and the RX 6800 XT’s narrow compute wins do not offset its sweeping losses elsewhere.