NVIDIA GeForce RTX 4080 SUPER vs NVIDIA GeForce RTX 5080 Comparison

The RTX 5080 and RTX 4080 SUPER are both 16 GB cards aimed at high-end PC builds, but the benchmark data draws a clear line between them. The RTX 5080 wins 9 out of 10 head-to-head tests, with an average benchmark score of 56083 compared to the RTX 4080 SUPER’s 54209. The RTX 4080 SUPER retains one narrow victory in a Vulkan compute test, but that is the exception. The data shows a generational step forward in raw performance, especially in modern DX12 workloads, while the older card remains a competent alternative for specific tasks.

Where Each One Wins

The RTX 5080 dominates the head-to-head results across nearly every category. Its largest win comes in the 3DMark Steel Nomad DX12 test, where it scores 8637 against the RTX 4080 SUPER’s 6600, a 30.9% advantage. That is the single biggest performance gap in the comparison and signals that the newer card is far stronger in next-generation game workloads. The RTX 5080 also leads in PassMark’s DirectX 12 test by 12.7% (151 vs 134) and in DirectX 10 by 7.8% (208 vs 193), confirming a consistent edge across API levels. In compute-heavy tasks, the RTX 5080’s PassMark GPU Compute score of 21789 is 9.9% higher than the 19822 of the RTX 4080 SUPER, and its Geekbench OpenCL score of 235901 is 7.7% ahead of 219065. Even in 2D performance, the RTX 5080 wins by 11.4% in PassMark G2D (1415 vs 1270), showing gains beyond just 3D rendering.

The RTX 4080 SUPER’s single win is in Geekbench Vulkan, scoring 260075 versus the RTX 5080’s 255450, a 1.8% margin. That is a meaningful result for users who rely on Vulkan-based compute or certain game engines, but it is isolated. The RTX 4080 SUPER also posts a respectable PassMark G3D score of 34245 (6.8% behind the RTX 5080’s 36565), so it is not a weak card—it simply loses in every other benchmark category. The RTX 5080 is the default choice for users prioritizing 3D gaming and general compute, while the RTX 4080 SUPER has a narrow niche in Vulkan-specific workloads where its architecture still holds a slight edge.

Architecture Differences

The two cards are built on different NVIDIA architectures, and those differences explain the benchmark gaps. The RTX 5080 uses the GB203 chip with Blackwell 2.0 architecture, while the RTX 4080 SUPER uses the AD103 chip with Ada Lovelace. Both are manufactured on a 5 nm process at TSMC, and their transistor counts are nearly identical—45,600 million for the RTX 5080 and 45,900 million for the RTX 4080 SUPER. Die size is also close: 378 mm² for the newer card versus 379 mm² for the older one. The RTX 5080 has a slightly lower transistor density at 120.6M / mm² compared to 121.1M / mm², but that does not translate into a performance deficit.

The key architectural differences lie in the compute units. The RTX 5080 packs 10752 shading units, 336 TMUs, and 112 ROPs, while the RTX 4080 SUPER has 10240 shading units, 320 TMUs, and the same 112 ROPs. The RTX 5080 also has more ray tracing cores (84 vs 80) and more tensor cores (336 vs 320). Those extra cores give the RTX 5080 a theoretical FP32 performance of 56.28 TFLOPS, which is 7.8% higher than the RTX 4080 SUPER’s 52.22 TFLOPS. Pixel rate and texture rate follow suit: the RTX 5080 hits 293.1 GPixel/s and 879.3 GTexel/s, versus 285.6 GPixel/s and 816.0 GTexel/s for the RTX 4080 SUPER. The RTX 5080 also runs at a higher boost clock of 2617 MHz compared to 2550 MHz on the older card, with the same 2295 MHz base clock.

Memory is another major split. The RTX 5080 uses GDDR7 memory at 1875 MHz (30 Gbps effective) with a 256-bit bus, delivering 960.0 GB/s of bandwidth. The RTX 4080 SUPER uses GDDR6X at 1438 MHz (23 Gbps effective) with the same 256-bit bus, but only 736.3 GB/s. That 30.4% bandwidth advantage for the RTX 5080 is a key reason it pulls ahead in memory-intensive DX12 tests. Both cards have 16 GB of memory, so capacity is not a differentiator. The RTX 5080 also supports PCIe 5.0 x16, while the RTX 4080 SUPER is limited to PCIe 4.0 x16, which matters for future platform compatibility. Display outputs differ too: the RTX 5080 has 1x HDMI 2.1b and 3x DisplayPort 2.1b, while the RTX 4080 SUPER offers 1x HDMI 2.1 and 3x DisplayPort 1.4a.

Head-to-Head Benchmarks

The 3DMark Steel Nomad DX12 result is the headline. The RTX 5080 scores 8637, which is 30.9% higher than the RTX 4080 SUPER’s 6600. That is not a marginal improvement—it is a generational jump that will be visible in demanding titles at high resolutions. In PassMark’s DirectX 12 test, the RTX 5080’s 151 beats 134 by 12.7%, reinforcing the DX12 advantage. DirectX 11 shows a similar pattern: the RTX 5080 scores 324 versus 301, a 7.6% lead. Even in older APIs, the RTX 5080 wins—DirectX 10 scores 208 to 193 (7.8% up) and DirectX 9 scores 389 to 381 (2.1% up). Those older API wins are smaller but confirm that the newer architecture does not sacrifice backward compatibility.

Compute benchmarks paint a consistent picture. In Geekbench OpenCL, the RTX 5080’s 235901 is 7.7% ahead of the RTX 4080 SUPER’s 219065. PassMark GPU Compute shows a 9.9% lead (21789 vs 19822). The RTX 5080 also wins PassMark G3D with 36565 against 34245, a 6.8% margin. The one loss for the RTX 5080 is Geekbench Vulkan, where it scores 255450 versus 260075 for the RTX 4080 SUPER, a 1.8% deficit. That is a narrow loss, and in practice it means Vulkan-based applications may run slightly better on the older card, but the overall compute picture strongly favors the RTX 5080. The RTX 5080 also leads in PassMark G2D (1415 vs 1270, 11.4% up), showing that even 2D desktop workloads benefit from the newer architecture.

Specification Differences

Several specifications differ between the two cards. The RTX 5080 has a boost clock of 2617 MHz versus 2550 MHz on the RTX 4080 SUPER, while base clocks are identical at 2295 MHz. Memory type changes from GDDR6X to GDDR7, and memory clock jumps from 1438 MHz (23 Gbps) to 1875 MHz (30 Gbps). Bandwidth rises from 736.3 GB/s to 960.0 GB/s. Shading units increase from 10240 to 10752, TMUs from 320 to 336, and ROPs remain at 112. Ray tracing cores go from 80 to 84, and tensor cores from 320 to 336. The RTX 5080 has a higher TDP of 360 W versus 320 W, and it suggests a 750 W PSU instead of 700 W. It is also a dual-slot card versus the triple-slot RTX 4080 SUPER, and it is slightly shorter (304 mm vs 310 mm) and narrower (40 mm vs 61 mm). The RTX 5080 uses PCIe 5.0 x16, while the RTX 4080 SUPER uses PCIe 4.0 x16. Display outputs differ as noted, with the RTX 5080 supporting DisplayPort 2.1b. The RTX 5080 is listed as Active production, while the RTX 4080 SUPER is End-of-life. Both have a launch MSRP of 999 USD.

FAQ

Q: Which card is faster in 3DMark Steel Nomad DX12?

A: The RTX 5080 scores 8637, which is 30.9% higher than the RTX 4080 SUPER’s 6600.

Q: Does the RTX 4080 SUPER win any benchmark?

A: Yes, it wins Geekbench Vulkan with a score of 260075, which is 1.8% higher than the RTX 5080’s 255450.

Q: How much more memory bandwidth does the RTX 5080 have?

A: The RTX 5080 provides 960.0 GB/s with GDDR7 memory, while the RTX 4080 SUPER provides 736.3 GB/s with GDDR6X.

Q: What is the difference in shading units?

A: The RTX 5080 has 10752 shading units, while the RTX 4080 SUPER has 10240, a difference of 512 units.

Q: Which card has a higher boost clock?

A: The RTX 5080 boosts to 2617 MHz, compared to 2550 MHz on the RTX 4080 SUPER.

Q: Are both cards the same price?

A: Both have a launch MSRP of 999 USD.

The Verdict

The data points to the RTX 5080 for almost any use case. It wins 9 of 10 head-to-head tests, and its 30.9% lead in 3DMark Steel Nomad DX12 is decisive for modern gaming. The extra 512 shading units, 84 ray tracing cores, and higher memory bandwidth of 960.0 GB/s give it a clear edge in compute and memory-intensive workloads. The RTX 5080 also has a higher boost clock (2617 MHz vs 2550 MHz) and a newer PCIe 5.0 interface, making it more future-proof. Its dual-slot design and shorter length (304 mm vs 310 mm) are practical benefits for case compatibility.

The RTX 4080 SUPER is the choice only if Vulkan performance is your priority. Its 260075 Geekbench Vulkan score is 1.8% higher, so users running Vulkan-based compute or specific engines may see a slight benefit. It also runs cooler in terms of TDP (320 W vs 360 W) and suggests a smaller PSU (700 W vs 750 W), which could matter in constrained builds. But its end-of-life production status and lower scores everywhere else make it a niche pick. For a new build, the RTX 5080 is the stronger investment, with a 3.5% higher average benchmark score (56083 vs 54209) and a percentile rank of 87 versus 86. The RTX 4080 SUPER remains a capable card, but the benchmark data does not support choosing it over the RTX 5080 except in very specific Vulkan scenarios.