NVIDIA GeForce RTX 5070 Ti Mobile vs NVIDIA P104-100 Comparison

The NVIDIA GeForce RTX 5070 Ti Mobile and the NVIDIA P104-100 represent two vastly different eras of GPU design, one a modern Blackwell mobile powerhouse and the other a Pascal-era mining card. The benchmark data shows a decisive generational gap, with the RTX 5070 Ti Mobile leading by massive margins in compute workloads, yet the comparison highlights fundamental architectural shifts in efficiency, memory, and feature support that go beyond raw performance.

FAQ

Q: How do the two GPUs compare in raw compute performance?

A: The RTX 5070 Ti Mobile dominates. In Geekbench OpenCL, it scores 143,870 versus 52,368 for the P104-100, a 174.7% advantage. In Geekbench Vulkan, the lead expands to 208.2%, with scores of 139,213 versus 45,165.

Q: Which GPU has a higher percentile ranking among all GPUs?

A: The RTX 5070 Ti Mobile sits in the 80th percentile, while the P104-100 ranks in the 77th percentile. Despite the huge performance gap in the head-to-head tests, both cards are positioned relatively high in the overall distribution.

Q: What is the difference in memory capacity and type?

A: The RTX 5070 Ti Mobile comes with 12 GB of GDDR7 memory on a 192-bit bus, delivering 672.0 GB/s bandwidth. The P104-100 has 4 GB of GDDR5X on a 256-bit bus, yielding 320.3 GB/s.

Q: How do the architectures differ in terms of manufacturing process?

A: The RTX 5070 Ti Mobile uses the Blackwell 2.0 architecture on a 5 nm TSMC process, packing 31,100 million transistors into a 263 mm² die. The P104-100 uses the older Pascal architecture on a 16 nm TSMC process, with 7,200 million transistors on a larger 314 mm² die.

Q: Do both cards support the same modern graphics APIs?

A: No. The RTX 5070 Ti Mobile supports DirectX 12 Ultimate (12_2) and Vulkan 1.4, while the P104-100 is limited to DirectX 12 (12_1) and Vulkan 1.4. Both support OpenGL 4.6.

Q: What are the power and physical characteristics of each card?

A: The RTX 5070 Ti Mobile has a 60 W TDP, is an IGP (integrated) form factor, and requires no power connectors. The P104-100 is a dual-slot card, 267 mm long, with a single 8-pin power connector and a suggested PSU of 200 W.

Architecture Differences

The RTX 5070 Ti Mobile is built on the Blackwell 2.0 architecture, a significant leap from the Pascal architecture of the P104-100. The manufacturing process highlights the generational shift: the newer card uses a 5 nm TSMC node, whereas the older card uses 16 nm TSMC. This is reflected in transistor density, with the RTX 5070 Ti Mobile achieving 118.3M transistors per mm² versus just 22.9M for the P104-100, despite the P104-100 having a larger physical die (314 mm² vs 263 mm²).

The compute core counts differ dramatically. The RTX 5070 Ti Mobile features 5,888 shading units, 184 TMUs, and 80 ROPs. It also includes 46 RT cores and 184 tensor cores, enabling hardware-accelerated ray tracing and AI workloads. The P104-100 has 1,920 shading units, 120 TMUs, and 64 ROPs, with no dedicated RT or tensor cores. The FP32 throughput reflects this: 17.04 TFLOPS for the RTX 5070 Ti Mobile versus 6.655 TFLOPS for the P104-100. The FP16 ratio is also starkly different, with the RTX 5070 Ti Mobile achieving 17.04 TFLOPS (1:1 with FP32), while the P104-100 manages only 104.0 GFLOPS (1:64 ratio), making it poorly suited for compute tasks that benefit from FP16.

Memory architecture is another differentiator. The RTX 5070 Ti Mobile uses 12 GB of GDDR7 with a 192-bit bus, while the P104-100 uses 4 GB of GDDR5X on a 256-bit bus. The newer card's bandwidth of 672.0 GB/s is more than double the P104-100's 320.3 GB/s. The bus interface also differs: PCIe 5.0 x16 for the RTX 5070 Ti Mobile versus PCIe 1.0 x4 for the P104-100, which severely limits data transfer rates for the older card. The RTX 5070 Ti Mobile also supports DirectX 12 Ultimate, while the P104-100 tops out at DirectX 12 (12_1).

Head-to-Head Benchmarks

The two shared benchmark results show a complete sweep for the RTX 5070 Ti Mobile, with the data indicating a generational chasm in compute performance. In Geekbench OpenCL, the RTX 5070 Ti Mobile scores 143,870 against the P104-100's 52,368. This represents a 174.7% delta, meaning the newer card is nearly 2.75 times faster in this general-purpose compute test. The gap is even more pronounced in Geekbench Vulkan, where the RTX 5070 Ti Mobile posts 139,213 versus 45,165 for the P104-100, a 208.2% difference. This suggests that the RTX 5070 Ti Mobile scales particularly well in graphics-oriented compute workloads, likely due to its modern architecture and superior memory bandwidth.

The average benchmark scores reinforce this narrative. The RTX 5070 Ti Mobile has an average score of 35,435 across its benchmark suite, while the P104-100 averages 32,982 based on its available tests. Interestingly, the nearest rivals for each card highlight their relative positioning. The RTX 5070 Ti Mobile is closest to the NVIDIA Quadro GV100 (35,520, -0.2% delta) and AMD Radeon Pro Duo (35,860, -1.2%), indicating it sits in a performance tier well above the P104-100. The P104-100, in contrast, is bracketed by entry-level cards like the NVIDIA T600 Mobile (32,849, 0.4% delta) and GeForce RTX 3050 Mobile (33,170, -0.6%).

The wins tally is decisive: the RTX 5070 Ti Mobile wins 2 out of 2 head-to-head tests, with the P104-100 failing to secure a single victory. The scale of these margins is notable — even in the less-favorable Vulkan test for the older card, the RTX 5070 Ti Mobile more than triples its score. This data suggests that for any modern compute or graphics workload, the RTX 5070 Ti Mobile is in a different performance class entirely.

The Verdict

The data clearly favors the NVIDIA GeForce RTX 5070 Ti Mobile for virtually any application. Its 174.7% lead in OpenCL and 208.2% lead in Vulkan over the P104-100 are not incremental improvements; they represent a complete architectural overhaul. The RTX 5070 Ti Mobile offers 12 GB of GDDR7 versus 4 GB of GDDR5X, a 60 W TDP versus a dual-slot card requiring a 200 W PSU, and supports modern DirectX 12 Ultimate features. Anyone considering a GPU for gaming, content creation, or AI workloads should choose the RTX 5070 Ti Mobile without hesitation.

The NVIDIA P104-100, however, is not without its niche. As a mining-specific card with no display outputs, it was never designed for gaming or desktop use. Its Pascal architecture lacks RT and tensor cores, making it obsolete for modern ray-traced titles or compute tasks that leverage FP16. Its only potential advantage lies in its 256-bit memory bus, which offers a wider path than the 192-bit bus of the RTX 5070 Ti Mobile, but the 4 GB capacity and lower bandwidth render this moot for practical use. The P104-100 is end-of-life, while the RTX 5070 Ti Mobile is actively produced.

The verdict is straightforward: the RTX 5070 Ti Mobile is the superior GPU in every measurable way. The P104-100 should only be considered by those with a specific legacy mining workload that cannot be migrated — and even then, its 77th percentile ranking and sub-33,000 average score suggest its utility is limited. The RTX 5070 Ti Mobile's 80th percentile ranking and 35,435 average score place it in a higher tier, and its feature set ensures it will remain relevant far longer.

Specification Differences

| Specification | NVIDIA GeForce RTX 5070 Ti Mobile | NVIDIA P104-100 |

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

| Architecture | Blackwell 2.0 | Pascal |

| Process Node | 5 nm (TSMC) | 16 nm (TSMC) |

| Transistors | 31,100 million | 7,200 million |

| Die Size | 263 mm² | 314 mm² |

| Transistor Density | 118.3M / mm² | 22.9M / mm² |

| Shading Units | 5,888 | 1,920 |

| TMUs | 184 | 120 |

| ROPs | 80 | 64 |

| RT Cores | 46 | null |

| Tensor Cores | 184 | null |

| FP32 Performance | 17.04 TFLOPS | 6.655 TFLOPS |

| FP16 Performance | 17.04 TFLOPS (1:1) | 104.0 GFLOPS (1:64) |

| Memory Size | 12 GB | 4 GB |

| Memory Type | GDDR7 | GDDR5X |

| Memory Bus Width | 192 bit | 256 bit |

| Memory Bandwidth | 672.0 GB/s | 320.3 GB/s |

| Base Clock | 847 MHz | 1607 MHz |

| Boost Clock | 1447 MHz | 1733 MHz |

| Memory Clock | 1750 MHz (28 Gbps effective) | 1251 MHz (10 Gbps effective) |

| TDP | 60 W | null |

| Slot Width | IGP | Dual-slot |

| Power Connectors | None | 1x 8-pin |

| Suggested PSU | null | 200 W |

| Bus Interface | PCIe 5.0 x16 | PCIe 1.0 x4 |

| Display Outputs | Portable Device Dependent | No outputs |

| DirectX Support | 12 Ultimate (12_2) | 12 (12_1) |

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

| Release Date | 2025-02-28 | 2017-12-11 |