NVIDIA GeForce RTX 4090 Mobile vs NVIDIA H20 NVL16 Comparison

Head-to-Head Benchmarks

The benchmark comparison between the NVIDIA GeForce RTX 4090 Mobile and the NVIDIA H20 NVL16 is notably one-sided, but not in the direction that raw hardware specifications might suggest. The database records no direct head-to-head benchmark results between these two GPUs, and the H20 NVL16 carries no recorded benchmark scores whatsoever. Its average benchmark score sits at 0, and its percentile ranking against all GPUs is 50. In contrast, the RTX 4090 Mobile holds a rich set of nine recorded benchmark results, an average benchmark score of 43,667, and a percentile ranking of 84.

The RTX 4090 Mobile delivers a Geekbench OpenCL score of 180,831 and a Geekbench Vulkan score of 170,774. These two results represent the strongest recorded outputs for the mobile GPU, and they indicate substantial compute throughput in cross-platform API environments. The Passmark suite adds further granularity. The mobile part scores 27,212 in Passmark G3D, which is the primary gaming and graphics workload metric, and 12,347 in Passmark GPU Compute. Legacy DirectX tests show 310 in Passmark DirectX 9, 262 in Passmark DirectX 11, 173 in Passmark DirectX 10, and 107 in Passmark DirectX 12. The 2D graphics score is 984.

Because the H20 NVL16 has no benchmark entries, the database cannot produce a direct percentage delta between the two parts. The only comparable data points come from the RTX 4090 Mobile's nearest rivals, which bracket its average score. The Quadro M6000 sits 0.8% behind with an average score of 43,301, the GeForce RTX 5050 Mobile sits 0.9% behind with 43,268, the Quadro M6000 24 GB sits 0.9% behind with 43,262, and the RTX A6000 sits 0.9% ahead with 44,075. These figures place the RTX 4090 Mobile in a tight performance cluster where a single percentage point separates the nearest competitors. The H20 NVL16 has no nearest rivals listed, so no such contextual comparison exists for it.

The practical interpretation is straightforward. The RTX 4090 Mobile's measured results place it in the upper quartile of all GPUs tracked in the database, with an 84th percentile ranking. The H20 NVL16's 50th percentile ranking with zero recorded scores means the database treats it as an untested part, not a poorly performing one. The absence of benchmark data for the H20 NVL16 does not indicate failure, it indicates that no standardized workload results have been recorded. The RTX 4090 Mobile, by contrast, has been exercised across multiple API generations and compute workloads, producing a comprehensive performance profile.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA GeForce RTX 4090 Mobile records an average benchmark score of 43,667. The NVIDIA H20 NVL16 records an average benchmark score of 0, as no benchmark results have been logged for it.

Q: How does the RTX 4090 Mobile compare to its closest rivals?

A: The RTX 4090 Mobile's average score of 43,667 places it 0.8% ahead of the Quadro M6000 (43,301), 0.9% ahead of the GeForce RTX 5050 Mobile (43,268), 0.9% ahead of the Quadro M6000 24 GB (43,262), and 0.9% behind the RTX A6000 (44,075).

Q: What does the H20 NVL16's percentile ranking indicate?

A: The H20 NVL16 holds a 50th percentile ranking against all GPUs in the database. This is a median placement, but it is based on no recorded benchmark scores, so the ranking reflects its default position rather than measured performance.

Q: What is the strongest recorded benchmark result for the RTX 4090 Mobile?

A: The Geekbench OpenCL score of 180,831 is the highest recorded benchmark result for the RTX 4090 Mobile, followed by the Geekbench Vulkan score of 170,774.

Q: Does the H20 NVL16 support DirectX, OpenGL, or Vulkan?

A: The H20 NVL16 lists DirectX as N/A, OpenGL as N/A, and Vulkan as N/A. The RTX 4090 Mobile supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

Q: Which GPU has the higher Passmark G3D score?

A: Only the RTX 4090 Mobile has a Passmark G3D score, which is 27,212. The H20 NVL16 has no Passmark scores recorded in the database.

The Verdict

The data presents a clear division of roles. The RTX 4090 Mobile is a measured performer, with nine benchmark scores spanning Geekbench OpenCL, Geekbench Vulkan, and the Passmark suite. Its 84th percentile ranking against all GPUs confirms that its average score of 43,667 places it well above the median. The nearest rival data reinforces this, as the RTX 4090 Mobile trades places within a single percentage point of the Quadro M6000, Quadro M6000 24 GB, GeForce RTX 5050 Mobile, and RTX A6000. This is a tightly contested performance tier, and the RTX 4090 Mobile holds its ground.

The H20 NVL16 is an unmeasured part. It has no benchmark scores, no rivals, and a 50th percentile default ranking. The database cannot say how it performs relative to the RTX 4090 Mobile because no standardized workload results exist for it. What the recorded data does show is that the H20 NVL16 is a fundamentally different class of hardware. It carries 80,000 million transistors, a 96 GB HBM3 memory pool, a 4.03 TB/s memory bandwidth, and a 400 W TDP rating. It uses no display outputs and supports no consumer graphics APIs. It is a server accelerator engineered for compute tasks, not for the benchmark suites that populate consumer GPU databases.

For users who rely on the recorded measurements, the RTX 4090 Mobile is the only part with verified performance data. Its scores across DirectX 9, 10, 11, and 12, plus OpenCL and Vulkan, make it a versatile mobile graphics solution with strong compute credentials. The H20 NVL16, meanwhile, cannot be evaluated on the same terms because the database contains no test results for it. Its specification sheet indicates a high-end server compute device, but the absence of benchmark data means any performance claim would not be supported by the recorded facts.

The practical guidance from the data is simple. Choose the RTX 4090 Mobile when measured graphics and compute performance are required, as its scores are documented and its percentile ranking is strong. Choose the H20 NVL16 when the workload demands the hardware characteristics listed in its specification: 96 GB of HBM3 memory, 4.03 TB/s bandwidth, 79.07 TFLOPS of FP16 compute, and a PCIe 5.0 x16 interface. The database cannot rank the H20 NVL16 against the RTX 4090 Mobile in raw performance, but it can confirm that the two parts occupy entirely different segments of the GPU market.

Specification Differences

The two GPUs differ across nearly every major specification category. The RTX 4090 Mobile uses an AD103 chip built on the Ada Lovelace architecture, while the H20 NVL16 uses a GH100 chip built on the Hopper architecture. The RTX 4090 Mobile belongs to the GeForce 40 Mobile generation, and the H20 NVL16 belongs to the Server Hopper (Hxx) generation.

Transistor counts diverge sharply. The RTX 4090 Mobile packs 45,900 million transistors on a 379 mm² die, yielding a transistor density of 121.1 million per mm². The H20 NVL16 packs 80,000 million transistors on an 814 mm² die, yielding a density of 98.3 million per mm². Both use a 5 nm process from TSMC, but the H20 NVL16's die is more than twice the size.

Clock speeds favor the H20 NVL16 in raw terms. The RTX 4090 Mobile runs at a base clock of 1335 MHz and a boost clock of 1695 MHz. The H20 NVL16 runs at a base clock of 1830 MHz and a boost clock of 1980 MHz. Memory clocks also differ, with the RTX 4090 Mobile using 2250 MHz memory at 18 Gbps effective, while the H20 NVL16 uses 1313 MHz memory at 5.3 Gbps effective.

Memory configurations are dramatically different. The RTX 4090 Mobile has 16 GB of GDDR6 on a 256-bit bus, providing 576.0 GB/s of bandwidth. The H20 NVL16 has 96 GB of HBM3 on a 6144-bit bus, providing 4.03 TB/s of bandwidth. That is roughly seven times the memory capacity and seven times the bandwidth, though the HBM3 memory type and bus width are entirely distinct from the GDDR6 implementation.

Compute unit counts are similar in shading units and texture units but differ in rasterization. The RTX 4090 Mobile has 9,728 shading units, 304 texture mapping units, and 112 raster output units. The H20 NVL16 has 9,984 shading units, 312 texture mapping units, and only 24 raster output units. The RTX 4090 Mobile also carries 76 ray tracing cores and 304 tensor cores, while the H20 NVL16 lists no ray tracing cores and 312 tensor cores.

Pixel and texture rates reflect these differences. The RTX 4090 Mobile achieves 189.8 GPixel/s and 515.3 GTexel/s. The H20 NVL16 achieves 47.52 GPixel/s and 617.8 GTexel/s. FP32 compute favors the H20 NVL16 at 39.54 TFLOPS versus 32.98 TFLOPS, and FP16 compute favors it more substantially at 79.07 TFLOPS (2:1 ratio) versus 32.98 TFLOPS (1:1 ratio).

Power and physical specifications also diverge. The RTX 4090 Mobile is rated at 120 W TDP, uses an IGP slot width, has no power connectors, and requires no suggested PSU. The H20 NVL16 is rated at 400 W TDP, uses an SXM Module slot width, and specifies an 800 W suggested PSU. The bus interface differs as well, with the RTX 4090 Mobile using PCIe 4.0 x16 and the H20 NVL16 using PCIe 5.0 x16. Display outputs are listed as portable device dependent for the RTX 4090 Mobile, while the H20 NVL16 has no outputs.

Release dates separate the two by more than two and a half years. The RTX 4090 Mobile was released on 2023-01-02, and the H20 NVL16 was released on 2025-09-01. Both are marked as active production status.

Architecture Differences

The architectural split between Ada Lovelace and Hopper represents two different design philosophies. The RTX 4090 Mobile's Ada Lovelace architecture is built around the AD103 chip, a design optimized for consumer graphics with full DirectX 12 Ultimate support, OpenGL 4.6, and Vulkan 1.4. Its 76 ray tracing cores and 304 tensor cores support hardware-accelerated ray tracing and AI workloads, and its FP16 compute runs at a 1:1 ratio with FP32, meaning both precision levels deliver 32.98 TFLOPS.

The H20 NVL16's Hopper architecture, built on the GH100 chip, is designed for server compute. It has no ray tracing cores listed, no display outputs, and no consumer graphics API support. Its FP16 compute runs at a 2:1 ratio with FP32, delivering 79.07 TFLOPS of FP16 versus 39.54 TFLOPS of FP32. This ratio indicates a deliberate focus on mixed-precision and tensor-heavy workloads, where the 312 tensor cores can be fully utilized. The 2:1 FP16 ratio is a signature of compute-oriented accelerators, whereas the 1:1 ratio on the RTX 4090 Mobile reflects a balanced approach for graphics and general compute.

The memory architecture reinforces this divide. The H20 NVL16's HBM3 memory with a 6144-bit bus is a server-class configuration designed for massive bandwidth, 4.03 TB/s, and large capacity, 96 GB. The RTX 4090 Mobile's GDDR6 memory on a 256-bit bus delivers 576.0 GB/s, which is ample for gaming and mobile graphics but far below the server part's throughput. The H20 NVL16's 24 raster output units, versus 112 on the RTX 4090 Mobile, confirm that rasterization is not its primary function. Its pixel rate of 47.52 GPixel/s is a fraction of the RTX 4090 Mobile's 189.8 GPixel/s, while its texture rate of 617.8 GTexel/s actually exceeds the mobile part's 515.3 GTexel/s, indicating a focus on compute-heavy shader work rather than traditional graphics rendering.

The process node is identical, 5 nm TSMC, but the die sizes and transistor densities tell different stories. The H20 NVL16 uses 80,000 million transistors across 814 mm², a massive die that prioritizes compute throughput over efficiency. The RTX 4090 Mobile uses 45,900 million transistors across 379 mm², a smaller die that balances performance with the thermal and power constraints of a mobile form factor. The transistor density of 121.1M per mm² on the RTX 4090 Mobile is higher than the H20 NVL16's 98.3M per mm², suggesting the Ada Lovelace chip packs logic more tightly, while the Hopper chip spreads out across a larger area.

The bus interface and power architecture also differ. The H20 NVL16 uses PCIe 5.0 x16, double the bandwidth of the RTX 4090 Mobile's PCIe 4.0 x16. Its 400 W TDP and 800 W suggested PSU place it firmly in server rack territory, whereas the RTX 4090 Mobile's 120 W TDP, IGP slot width, and lack of power connectors suit it for laptops and compact devices. The H20 NVL16's predecessor is listed as Server Ada, and its successor is Server Blackwell, while the RTX 4090 Mobile's predecessor is GeForce 30 Mobile and its successor is GeForce 50 Mobile. These lineage entries confirm that each part evolves along a separate track, one for mobile consumer graphics and one for server compute.