AMD Radeon RX 9070 XT vs NVIDIA P106-090 Comparison

The AMD Radeon RX 9070 XT and the NVIDIA P106-090 occupy radically different positions in the GPU landscape, yet their average benchmark scores place them within 0.5% of each other. The RX 9070 XT posts an average benchmark score of 13543, while the P106-090 scores 13470, a delta of just 0.5%. This near-parity in aggregate metrics, however, masks a massive divergence in individual workload performance, architectural capabilities, and intended use cases, which the head-to-head data clarifies.

Head-to-Head Benchmarks

The most striking result is in the 3DMark Steel Nomad DX12 test, where the AMD Radeon RX 9070 XT achieves a score of 7260 compared to the NVIDIA P106-090’s 509. This represents a 1326.3% advantage for the RX 9070 XT, a margin that dwarfs any other comparison in this dataset. This single test underscores the RX 9070 XT’s dominance in modern, API-heavy rendering workloads, where its architecture is optimized for contemporary game engines. The P106-090’s score of 509 is not merely lower; it indicates that the card is effectively out of its depth in this particular DirectX 12 scenario.

The Vulkan benchmark tells a similar story, though with a smaller gap. The RX 9070 XT scores 65879 in Geekbench Vulkan, while the P106-090 manages 18596. This yields a 254.3% win for the AMD card. While not as extreme as the Steel Nomad result, this still demonstrates a substantial lead in compute-oriented graphics tasks that leverage Vulkan’s low-level overhead. The RX 9070 XT’s ability to nearly triple the P106-090’s output in this test points to a fundamental difference in raw shading power and memory bandwidth.

The only benchmark where the NVIDIA P106-090 wins is Geekbench OpenCL. Here, the P106-090 scores 21304, exceeding the RX 9070 XT’s 17428 by 18.2%. This is a peculiar result, given the RX 9070 XT’s massive hardware advantage. The data shows that the P106-090, despite its Pascal-era architecture and lower specifications, is able to outperform the newer AMD card in this specific compute API. This suggests that the P106-090’s driver stack or compute scheduling for OpenCL is particularly efficient, or that the RX 9070 XT’s OpenCL implementation is not a priority. Regardless of the reason, this is a clear, albeit isolated, victory for the NVIDIA product.

Overall, the AMD Radeon RX 9070 XT wins 2 out of 3 head-to-head benchmarks. The wins in DX12 and Vulkan are decisive, while the loss in OpenCL is substantial. The average benchmark scores, which are nearly identical, obscure these dramatic swings, reinforcing that a single aggregate number is insufficient to understand the performance profile of either card.

Architecture Differences

The architectural chasm between these two GPUs is vast. The AMD Radeon RX 9070 XT is built on the RDNA 4.0 architecture using TSMC’s 4 nm process node, packing 53,900 million transistors onto a 357 mm² die. This yields a transistor density of 151.0M per mm². In contrast, the NVIDIA P106-090 is a Pascal-generation part from the "Mining GPUs" lineup, fabricated on a 16 nm process with just 4,400 million transistors on a 200 mm² die, resulting in a density of 22.0M per mm². The process node difference alone—4 nm versus 16 nm—explains much of the disparity in efficiency and transistor budget.

The core configurations are equally divergent. The RX 9070 XT features 4096 shading units, 256 TMUs, and 128 ROPs, alongside 64 dedicated ray tracing cores. The P106-090 has only 768 shading units, 48 TMUs, and 48 ROPs, with no ray tracing cores present. This means the AMD card offers over five times the shading units and more than double the ROPs, directly translating to its higher pixel rate of 380.2 GPixel/s and texture rate of 760.3 GTexel/s, compared to the P106-090’s 73.49 GPixel/s and 73.49 GTexel/s. The FP32 compute is similarly lopsided: 48.66 TFLOPS for the RX 9070 XT versus 2.352 TFLOPS for the P106-090.

Memory subsystems further separate the two. The RX 9070 XT uses 16 GB of GDDR6 on a 256-bit bus, delivering 644.6 GB/s of bandwidth. The P106-090 is limited to 3 GB of GDDR5 on a 192-bit bus, providing 192.2 GB/s. This is a 16 GB versus 3 GB capacity difference and a bandwidth gap of over 450 GB/s, which heavily impacts high-resolution texture loading and compute workloads. Interface support also differs: the RX 9070 XT supports PCIe 5.0 x16 and features display outputs (1x HDMI 2.1b, 3x DisplayPort 2.1a), while the P106-090 is limited to PCIe 1.0 x1 and has no display outputs at all, reflecting its mining-oriented design.

Where Each One Wins

Based strictly on the benchmark data, the AMD Radeon RX 9070 XT is the clear winner for any modern graphics workload that relies on DirectX 12 or Vulkan. The 1326.3% lead in 3DMark Steel Nomad and the 254.3% lead in Geekbench Vulkan indicate that this card is designed for current-generation gaming, high-fidelity rendering, and compute tasks that utilize these APIs. Its 64 ray tracing cores and 16 GB of VRAM make it suited for ray-traced titles and large asset packs, though the benchmark scores themselves do not explicitly test those features. The data supports its use in any scenario where raw graphical output is the primary metric.

The NVIDIA P106-090’s single win in Geekbench OpenCL (18.2% ahead) suggests it may have a niche in legacy or specialized compute applications that rely heavily on OpenCL. The card’s higher OpenCL score of 21304 versus 17428 is the only data point where it leads, so any advantage it holds is confined to that specific API. Its 3 GB of memory and lack of display outputs preclude it from being a general-purpose gaming or workstation card. The P106-090 is best viewed as a compute-only accelerator for tasks that are optimized for OpenCL and do not require video output or modern graphics API features.

The average benchmark scores reinforce this split. The RX 9070 XT’s percentile of 55 and the P106-090’s percentile of 54 are nearly identical, placing both in the same tier of overall performance according to the database. However, this percentile ranking is heavily influenced by the inclusion of the P106-090’s strong OpenCL result and the RX 9070 XT’s weaker one. The head-to-head results show that in every other tested workload, the RX 9070 XT is overwhelmingly superior.

The Verdict

From the data, the choice between these two cards is not a matter of preference but of workload compatibility. The AMD Radeon RX 9070 XT should be selected by anyone needing a fully featured, high-performance graphics card for DirectX 12 and Vulkan applications. Its 2-to-1 win record in head-to-head tests, combined with wins of 1326.3% and 254.3% in the two most demanding modern benchmarks, demonstrates an overwhelming advantage in contemporary rendering. Its active production status, 16 GB of VRAM, and display outputs make it a viable, long-term solution for gaming and professional visualization.

The NVIDIA P106-090 is appropriate only for a narrow set of circumstances. Its 18.2% win in OpenCL is its sole redeeming benchmark result. Given its end-of-life production status, 3 GB of memory, and lack of any display outputs, it is not suitable for a primary workstation or gaming PC. The data suggests it could serve as a secondary compute device for OpenCL-specific tasks, but even then, its 0.5% lower average score compared to the RX 9070 XT means it offers no aggregate performance benefit. For any user who requires a general-purpose GPU, the RX 9070 XT is the only rational choice based on these benchmarks.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The AMD Radeon RX 9070 XT has an average benchmark score of 13543, which is 0.5% higher than the NVIDIA P106-090’s score of 13470.

Q: How large is the performance gap in the 3DMark Steel Nomad test?

A: The RX 9070 XT scores 7260, while the P106-090 scores 509. This gives the AMD card a 1326.3% advantage in that specific DirectX 12 benchmark.

Q: Is there any benchmark where the NVIDIA P106-090 outperforms the AMD card?

A: Yes, in Geekbench OpenCL, the P106-090 scores 21304, which is 18.2% higher than the RX 9070 XT’s score of 17428.

Q: What is the difference in memory capacity?

A: The AMD Radeon RX 9070 XT has 16 GB of GDDR6 memory, while the NVIDIA P106-090 has 3 GB of GDDR5 memory.

Q: Do both cards support the same modern APIs?

A: No. The RX 9070 XT supports DirectX 12 Ultimate (12_2) and Vulkan 1.4, while the P106-090 supports DirectX 12 (12_1) and Vulkan 1.4. Both support OpenGL 4.6.

Q: Which card has a higher transistor count?

A: The AMD Radeon RX 9070 XT has 53,900 million transistors, compared to the NVIDIA P106-090’s 4,400 million transistors.

Specification Differences

| Specification | AMD Radeon RX 9070 XT | NVIDIA P106-090 |

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

| Architecture | RDNA 4.0 | Pascal |

| Process Node | 4 nm | 16 nm |

| Transistors | 53,900 million | 4,400 million |

| Die Size | 357 mm² | 200 mm² |

| Base Clock | 1660 MHz | 1354 MHz |

| Boost Clock | 2970 MHz | 1531 MHz |

| Memory Size | 16 GB | 3 GB |

| Memory Type | GDDR6 | GDDR5 |

| Memory Bus Width | 256 bit | 192 bit |

| Memory Bandwidth | 644.6 GB/s | 192.2 GB/s |

| Shading Units | 4096 | 768 |

| TMUs | 256 | 48 |

| ROPs | 128 | 48 |

| RT Cores | 64 | None |

| FP32 Performance | 48.66 TFLOPS | 2.352 TFLOPS |

| Pixel Rate | 380.2 GPixel/s | 73.49 GPixel/s |

| Texture Rate | 760.3 GTexel/s | 73.49 GTexel/s |

| TDP | 304 W | 75 W |

| Power Connectors | 2x 8-pin | 1x 6-pin |

| Suggested PSU | 700 W | 250 W |

| Bus Interface | PCIe 5.0 x16 | PCIe 1.0 x1 |

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

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

| Release Date | 2025-03-05 | 2017-07-30 |