NVIDIA GeForce RTX 5080 SUPER vs NVIDIA GeForce RTX 5090 Comparison

Where Each One Wins

The benchmark data splits these two GeForce 50-series cards cleanly along performance tiers. The NVIDIA GeForce RTX 5080 SUPER does not win a single recorded benchmark against the RTX 5090. The RTX 5090 claims the only head-to-head victory, and its wider benchmark portfolio shows dominance across every category where both cards have recorded results.

The RTX 5080 SUPER sits in the 19th percentile of all GPUs in the database, with an average benchmark score of 3075 from a single 3DMark Steel Nomad DX12 run. Its nearest rivals in the database include the NVIDIA Quadro P1000 (2.8% higher average score), the Intel Arc Pro B60 (3.4% higher), and the NVIDIA GeForce 820A (3.1% lower). These are not flagship-level competitors; the 5080 SUPER is positioned far below the top of the stack despite its "SUPER" designation.

The RTX 5090, by contrast, lands in the 92nd percentile of all GPUs, with an average benchmark score of 79,842 across ten separate tests. Its nearest rivals are high-end workstation and gaming parts: the NVIDIA Tesla P100 PCIe 16 GB (0.3% higher), the Tesla P100 PCIe 12 GB (0.6% higher), the AMD Radeon RX 6850M XT (1.1% higher), and the AMD Radeon Pro Vega 64X (1.4% lower). The 5090's nearest competitors are within 1.4% of its average, indicating it sits at the apex of the current database distribution.

In practical terms, the RTX 5090 wins every scenario where raw compute throughput matters. The 5080 SUPER's only recorded benchmark result is the same 3DMark Steel Nomad DX12 test where the 5090 scores 18,355, a 496.9% advantage. The data shows no use case where the 5080 SUPER overtakes the 5090; the gap is consistent and large across all recorded metrics.

FAQ

Q: How much faster is the RTX 5090 than the RTX 5080 SUPER in the shared benchmark?

A: In the 3DMark Steel Nomad DX12 test, the RTX 5090 scores 18,355 versus the 5080 SUPER's 3,075. The delta is -83.2% from the 5090's perspective, meaning the 5080 SUPER delivers only 16.8% of the 5090's score.

Q: What is the average benchmark score for each card?

A: The RTX 5080 SUPER averages 3,075 across its single recorded benchmark. The RTX 5090 averages 79,842 across ten tests, including 3DMark, Geekbench OpenCL, Geekbench Vulkan, and multiple Passmark subtests.

Q: Which card has better memory bandwidth?

A: The RTX 5090 has 1.79 TB/s of bandwidth from a 512-bit bus with 32 GB of GDDR7 memory. The RTX 5080 SUPER has 1.02 TB/s from a 256-bit bus with 24 GB of GDDR7 memory.

Q: Are the two cards from the same architecture generation?

A: Yes, both use the Blackwell 2.0 architecture on TSMC's 5 nm process. They differ in chip: the 5080 SUPER uses GB203, while the 5090 uses GB202.

Q: What are the launch MSRPs?

A: The RTX 5080 SUPER has a launch MSRP of 999 USD. The RTX 5090 has a launch MSRP of 1,999 USD.

Q: Which card has more shading units?

A: The RTX 5090 has 21,760 shading units, while the RTX 5080 SUPER has 10,752. The 5090 also doubles the TMUs (680 vs 336), ROPs (176 vs 112), RT cores (170 vs 84), and tensor cores (680 vs 336).

Head-to-Head Benchmarks

The database records exactly one direct head-to-head benchmark between these two cards: 3DMark Steel Nomad DX12. The RTX 5090 posts 18,355 points, while the RTX 5080 SUPER scores 3,075. That is a 15,280-point gap, equivalent to the 5090 delivering roughly six times the performance of the 5080 SUPER in this workload. The delta percentage of -83.2% means the 5080 SUPER trails by 83.2% of the 5090's score.

Beyond the single shared test, the 5090's other benchmarks provide context for its overall capability. In Geekbench OpenCL, it scores 334,370. In Geekbench Vulkan, it scores 376,728. Passmark results show 39,650 in G3D, 26,756 in GPU Compute, 1,413 in G2D, and lower scores in legacy DirectX tests: 395 in DirectX 9, 341 in DirectX 11, 226 in DirectX 10, and 185 in DirectX 12.

The 5080 SUPER has no recorded results in any of those additional tests, so its only measurable point of comparison remains the single 3DMark run. The data indicates the 5090's advantage is not marginal but categorical. The nearest rivals to the 5080 SUPER in the database (Quadro P1000, Arc Pro B60, GeForce 820A, GTX 860M) cluster within 3.6% of its score, while the 5090's rivals (Tesla P100 variants, RX 6850M XT, Pro Vega 64X) cluster within 1.4% of a much higher average. This places the two cards in entirely different performance strata.

Specification Differences

The two cards diverge on nearly every countable specification. Core configuration: the RTX 5080 SUPER has 10,752 shading units, 336 TMUs, 112 ROPs, 84 RT cores, and 336 tensor cores. The RTX 5090 has 21,760 shading units, 680 TMUs, 176 ROPs, 170 RT cores, and 680 tensor cores. The 5090 doubles or exceeds the 5080 SUPER in every category, with shading units and tensor cores exactly doubled.

Memory: the 5080 SUPER uses 24 GB of GDDR7 on a 256-bit bus with 1.02 TB/s bandwidth. The 5090 uses 32 GB of GDDR7 on a 512-bit bus with 1.79 TB/s bandwidth. Clock speeds favor the smaller card: the 5080 SUPER boosts to 2617 MHz from a 2295 MHz base, while the 5090 boosts to 2407 MHz from a 2017 MHz base. Memory clocks also differ, with the 5080 SUPER at 2000 MHz (32 Gbps effective) versus the 5090 at 1750 MHz (28 Gbps effective).

Compute rates: the 5080 SUPER delivers 56.28 TFLOPS FP32 and FP16 (1:1), with a pixel rate of 293.1 GPixel/s and texture rate of 879.3 GTexel/s. The 5090 delivers 104.8 TFLOPS FP32 and FP16 (1:1), with 423.6 GPixel/s and 1,636.8 GTexel/s. The 5090's FP32 throughput is 86.2% higher, its pixel rate 44.5% higher, and its texture rate 86.1% higher.

Power and physical characteristics: the 5080 SUPER has a TDP of 415 W with one 16-pin connector, while the 5090 has a TDP of 575 W with one 16-pin connector and a suggested PSU of 950 W. Both are dual-slot cards with identical dimensions: 304 mm length, 137 mm height, and 40 mm width. Both use PCIe 5.0 x16 and offer 1x HDMI 2.1b plus 3x DisplayPort 2.1b outputs.

Release timing differs: the 5090 launched on 2025-01-29 with a predecessor in GeForce 40 and a successor in GeForce 60. The 5080 SUPER launched on 2025-12-31 with no recorded predecessor or successor.

Architecture Differences

Both cards use NVIDIA's Blackwell 2.0 architecture fabricated by TSMC on a 5 nm process, so the fundamental instruction set and feature support are identical. Both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The difference lies in the physical implementation.

The RTX 5090 uses the GB202 chip, which contains 92,200 million transistors on a 750 mm² die, yielding a transistor density of 122.9 million per mm². The RTX 5080 SUPER uses the GB203 chip, with 45,600 million transistors on a 378 mm² die, giving 120.6 million per mm². The GB202 has more than double the transistor count and nearly double the die area, but the density is only slightly higher (122.9 vs 120.6 million per mm²), indicating the 5090 scales primarily through additional silicon rather than denser packing.

The RT core and tensor core counts scale with the shading unit counts: the 5090 has 170 RT cores and 680 tensor cores versus 84 RT cores and 336 tensor cores on the 5080 SUPER. This doubling of specialized hardware means ray tracing and AI workloads see the same proportional improvement as raw rasterization. The 1:1 FP16 to FP32 ratio on both cards indicates no dedicated half-precision path, and both deliver equal throughput in both precisions.

The 5090's higher TDP (575 W vs 415 W) reflects the larger chip and higher memory bandwidth. Both cards use a single 16-pin power connector, so the 5090's additional power draw is delivered through the same interface. The 5090's suggested PSU of 950 W is recorded in the database; the 5080 SUPER has no suggested PSU listed.

The memory subsystem represents the most significant architectural difference beyond the chip itself. The 5090's 512-bit bus with 32 GB GDDR7 and 1.79 TB/s bandwidth is a 100% wider bus and 75.5% higher bandwidth than the 5080 SUPER's 256-bit bus, 24 GB, and 1.02 TB/s. Despite the 5090's lower memory clock (1750 MHz vs 2000 MHz), the wider bus more than compensates, delivering substantially higher effective throughput. The 5080 SUPER's higher memory clock may help in latency-sensitive tasks, but the bandwidth deficit is decisive in throughput-bound workloads.

The production status for both cards is Active, and both are in the GeForce 50-series generation. The 5090 has recorded successor and predecessor relationships (GeForce 60 and GeForce 40, respectively), while the 5080 SUPER has neither, suggesting it may be a late-cycle addition to the 50-series lineup.