NVIDIA GeForce RTX 5090 Mobile vs NVIDIA P102-100 Comparison

NVIDIA’s P102-100 is a relic of the cryptocurrency mining boom, while the GeForce RTX 5090 Mobile is a current-generation laptop flagship. The database shows a stark generational divide: across the two shared benchmark tests, the RTX 5090 Mobile wins both, with an average benchmark score of 45,152 compared to the P102-100’s 58,528. That average is heavily weighted by the P102-100’s strong OpenCL and Vulkan results, but the head-to-head data tells a different story. The P102-100 does hold a higher percentile rank among all GPUs (88th vs. 84th), which reflects its specialized compute-oriented design, yet the RTX 5090 Mobile’s raw performance in modern workloads is categorically superior. This comparison is not about incremental gains; it is about two entirely different eras of GPU engineering.

Where Each One Wins

The P102-100 is a mining-specific card with no display outputs, and its benchmark profile reflects that narrow purpose. In Geekbench OpenCL, it scores 49,602, and in Geekbench Vulkan, it reaches 67,454. These are respectable numbers for a 2018 product, placing it within 0.2% of the AMD Radeon RX 6950 XT and 0.5% of the Intel Arc A570M in average score. Its strength lies in raw compute throughput for its era: 10.77 TFLOPS FP32, 440.3 GB/s of memory bandwidth, and a 320-bit bus with 5 GB of GDDR5X. For tasks that rely on massive parallel integer or floating-point operations without modern graphics features, the P102-100 can still hold its own. It is also an end-of-life product, so its wins are historical, not current.

The RTX 5090 Mobile dominates every shared benchmark by a wide margin. Its Geekbench OpenCL score of 201,834 is 306.8% higher than the P102-100’s, and its Vulkan score of 198,405 is 194.1% higher. In the database’s broader test suite, it also records a Passmark G3D score of 30,034 and a Passmark GPU Compute score of 13,401, which are not available for the P102-100. The RTX 5090 Mobile wins on architectural features too: it has 10,496 shading units, 82 RT cores, and 328 tensor cores, enabling hardware ray tracing and AI acceleration that the Pascal-based P102-100 lacks entirely. Its 31.80 TFLOPS FP32 is nearly three times the P102-100’s output, and its FP16 performance is 31.80 TFLOPS (1:1) versus the P102-100’s paltry 168.3 GFLOPS (1:64). The RTX 5090 Mobile is the clear winner for any modern workload, from gaming to content creation to machine learning inference.

The Verdict

The data is unambiguous: the NVIDIA GeForce RTX 5090 Mobile is the superior GPU in every measurable category that matters today. It wins 2 out of 2 head-to-head benchmarks, with deltas of -75.4% and -66% (meaning the P102-100 trails by those percentages). The RTX 5090 Mobile’s average benchmark score of 45,152 is lower than the P102-100’s 58,528, but that is a statistical artifact: the P102-100 only has two scores, both in older APIs, while the RTX 5090 Mobile’s average includes demanding modern tests like 3DMark Steel Nomad DX12 (5,871) and multiple Passmark DX10/11/12 runs. When comparing apples to apples, the RTX 5090 Mobile is 4.1 times faster in OpenCL and 2.9 times faster in Vulkan.

Who should pick the P102-100? Only someone maintaining legacy mining rigs or specialized compute clusters that cannot use modern drivers or APIs. It has no display outputs, so it is useless for gaming or desktop use. Its 250 W TDP and dual-slot design with 2x 8-pin connectors are also outdated compared to the RTX 5090 Mobile’s 95 W TDP and IGP form factor. The P102-100’s nearest rivals, like the AMD Radeon PRO V710 (-0.2%) and AMD Radeon RX 6950 XT (0.2%), show it is competitive with mid-range 2020-era cards, but that is faint praise against a 2025 flagship.

Who should pick the RTX 5090 Mobile? Anyone needing portable high-performance computing. It delivers desktop-class performance in a laptop package, with 24 GB of GDDR7 memory, 896.0 GB/s bandwidth, and a PCIe 5.0 x16 interface. Its nearest rivals include the NVIDIA GeForce RTX 4070 Ti (0.8% delta), which is a desktop card, and the NVIDIA RTX 5880 Ada Generation (-1.8%), indicating it punches well above its mobile designation. The RTX 5090 Mobile is active in production, supports DirectX 12 Ultimate, and has full ray tracing and tensor core support. For gaming, 3D rendering, or AI workloads, the choice is obvious.

Head-to-Head Benchmarks

The two shared benchmarks are Geekbench OpenCL and Geekbench Vulkan. In OpenCL, the RTX 5090 Mobile scores 201,834 against the P102-100’s 49,602, a delta of -75.4% (the P102-100 is 75.4% slower). This is the largest gap in the comparison, driven by the RTX 5090 Mobile’s 10,496 shading units versus 3,200, and its 31.80 TFLOPS FP32 versus 10.77 TFLOPS. The P102-100’s 1:64 FP16 ratio also cripples it in compute tasks that use half-precision, while the RTX 5090 Mobile’s 1:1 ratio makes it a compute monster.

In Vulkan, the RTX 5090 Mobile scores 198,405 versus the P102-100’s 67,454, a delta of -66%. This gap is slightly smaller because Vulkan is a lower-level API that can better utilize the P102-100’s raw rasterization throughput. The P102-100’s 80 ROPs and 200 TMUs are not negligible, but the RTX 5090 Mobile counters with 112 ROPs and 328 TMUs, plus modern geometry processing. The RTX 5090 Mobile also benefits from GDDR7 memory at 896.0 GB/s, double the P102-100’s 440.3 GB/s bandwidth, which reduces bottlenecks in memory-intensive Vulkan workloads.

Beyond these two, the database records no other shared tests, but the RTX 5090 Mobile’s additional benchmarks paint a fuller picture. Its 3DMark Steel Nomad DX12 score of 5,871 is a modern rasterization test, and its Passmark G3D score of 30,034 places it in the upper echelon of mobile GPUs. The P102-100 has no comparable data, so any claims of its competitiveness in modern DirectX 12 titles are unsupported. The P102-100’s only wins are historical: it was released on February 11, 2018, while the RTX 5090 Mobile launched on March 26, 2025.

FAQ

Q: Which GPU has a higher average benchmark score?

A: The NVIDIA P102-100 has a higher average benchmark score of 58,528, compared to the RTX 5090 Mobile’s 45,152. However, this is because the P102-100 only has two benchmark entries, both in older APIs, while the RTX 5090 Mobile’s average includes many modern, more demanding tests.

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

A: The RTX 5090 Mobile scores 201,834 in Geekbench OpenCL, which is 306.8% higher than the P102-100’s 49,602. The delta percentage recorded is -75.4%, meaning the P102-100 trails by that amount.

Q: Does the P102-100 support ray tracing?

A: No. The P102-100 has no RT cores or tensor cores listed in the database. The RTX 5090 Mobile has 82 RT cores and 328 tensor cores, enabling hardware-accelerated ray tracing and AI features.

Q: What is the memory configuration difference?

A: The P102-100 has 5 GB of GDDR5X on a 320-bit bus with 440.3 GB/s bandwidth. The RTX 5090 Mobile has 24 GB of GDDR7 on a 256-bit bus with 896.0 GB/s bandwidth, which is more than double the bandwidth despite a narrower bus.

Q: Which GPU is more power-efficient?

A: The RTX 5090 Mobile has a TDP of 95 W, while the P102-100 has a TDP of 250 W. The RTX 5090 Mobile delivers significantly higher performance at less than half the power draw.

Q: Can the P102-100 be used for gaming?

A: No. The P102-100 has no display outputs listed in the database, making it impossible to connect a monitor. It was designed as a mining GPU.

Architecture Differences

The two GPUs are separated by seven years of architectural evolution. The P102-100 uses the GP102 chip on TSMC’s 16 nm process, with 11,800 million transistors on a 471 mm² die, yielding a transistor density of 25.1 million per mm². It is based on the Pascal architecture, which supports DirectX 12 (12_1) but lacks dedicated hardware for ray tracing or tensor operations. Its FP16 performance is a token 168.3 GFLOPS at a 1:64 ratio, meaning it is effectively a single-precision-only compute device. The P102-100 also uses a PCIe 1.0 x4 interface, a severe bottleneck for data transfer, and has no display outputs. Its 250 W TDP and dual-slot cooler with 2x 8-pin power connectors are typical of 2018-era high-end cards, but its bus interface is oddly antiquated.

The RTX 5090 Mobile is built on the GB203 chip using TSMC’s 5 nm process, with 45,600 million transistors on a 378 mm² die, a density of 120.6 million per mm². This is nearly five times the transistor density of the P102-100. The Blackwell 2.0 architecture brings 82 RT cores and 328 tensor cores, enabling real-time ray tracing and AI-accelerated features like DLSS. Its FP16 performance is 31.80 TFLOPS at a 1:1 ratio, matching its FP32 output, which is critical for modern compute workloads. The RTX 5090 Mobile uses a PCIe 5.0 x16 interface, a massive upgrade over the P102-100’s PCIe 1.0 x4, and its 24 GB of GDDR7 memory operates at 28 Gbps effective, delivering 896.0 GB/s. It is an IGP (integrated graphics package) with no power connectors, drawing only 95 W, and its display outputs are listed as "Portable Device Dependent," reflecting its laptop-focused design. The RTX 5090 Mobile also supports DirectX 12 Ultimate (12_2), which includes features like mesh shaders and variable rate shading, while the P102-100 is limited to DirectX 12 (12_1). In every architectural metric, the RTX 5090 Mobile is not just newer, it is a fundamentally more advanced design.