NVIDIA GeForce RTX 2070 SUPER vs NVIDIA GeForce RTX 3080 Comparison

Head-to-Head Benchmarks

The benchmark data paints a remarkably lopsided picture. Across the ten recorded head-to-head tests, the NVIDIA GeForce RTX 3080 claims nine outright victories, while the RTX 2070 SUPER manages just a single win. The margins, however, tell a more nuanced story than the raw win count suggests.

The most dramatic separation occurs in the 3DMark Steel Nomad DX12 test. Here, the RTX 3080 scores 4407 against the RTX 2070 SUPER's 1651, a delta of 166.9%. That is not a marginal improvement; it is a generational leap expressed in a single benchmark. The RTX 3080 more than doubles the performance of its predecessor in this particular workload, which stresses modern DirectX 12 rendering paths.

Compute-heavy workloads also show substantial divergence. In Geekbench OpenCL, the RTX 3080 records 152423 points versus 83358 for the RTX 2070 SUPER, a 82.9% advantage. The PassMark GPU Compute test similarly favors the newer card, with scores of 14397 and 7557 respectively, representing a 90.5% delta. These figures suggest the Ampere architecture's raw compute throughput scales far beyond what the Turing design can muster.

DirectX 12 performance in PassMark shows a 49.3% lead for the RTX 3080 (100 versus 67). The DirectX 11 test shows a 37.1% gap (207 versus 151). Even the legacy DirectX 9 and DirectX 10 workloads, which often narrow differences between architectures, show the RTX 3080 ahead by 15.7% (258 versus 223) and 28.8% (170 versus 132) respectively.

The aggregate PassMark G3D score, which represents overall 3D graphics performance, lands at 25086 for the RTX 3080 against 18169 for the RTX 2070 SUPER, a 38.1% delta. The 2D graphics score shows a 20% difference (1054 versus 878), a smaller but still consistent gap.

The single exception is Geekbench Vulkan. Here, the RTX 2070 SUPER scores 90637, while the RTX 3080 manages only 33620. The delta is -62.9%, meaning the older card outperforms the newer one by a wide margin. This anomaly stands out sharply against the otherwise uniform pattern. It suggests that Vulkan driver optimization or specific API dispatch behavior on the Turing architecture yields an unusual advantage in this particular test, one that does not translate to other workloads.

Where Each One Wins

The RTX 3080 dominates across nearly every measured category. Its wins span compute, legacy DirectX, modern DirectX, 2D, and 3D workloads. The data indicates a card that excels in both general-purpose GPU compute and traditional rasterization. The PassMark DirectX 9, 10, 11, and 12 results all favor the RTX 3080, which implies consistent architectural efficiency across API generations rather than optimization for a single path.

The RTX 2070 SUPER's sole victory in Geekbench Vulkan is difficult to categorize as a use-case win. It is a single benchmark result, not a pattern. For users running Vulkan-based applications, the data does show a measurable advantage for the older card, but this must be weighed against the overwhelming deficits elsewhere. The RTX 3080's Vulkan score of 33620 is lower than its OpenCL score of 152423, which is itself unusual. The RTX 2070 SUPER, by contrast, scores higher in Vulkan (90637) than in OpenCL (83358). This asymmetry hints at different API efficiency profiles between the two architectures.

For gaming workloads, the 3DMark Steel Nomad result is the most relevant modern indicator. The 166.9% lead for the RTX 3080 suggests that current-generation DirectX 12 titles will see substantial performance gains. The PassMark G3D score, which aggregates multiple DirectX tests, reinforces this with a 38.1% overall advantage. For compute-oriented tasks such as rendering, machine learning inference, or scientific simulations, the OpenCL and GPU Compute results point decisively toward the RTX 3080.

Architecture Differences

The two cards represent entirely different architectural generations. The RTX 3080 uses the GA102 chip built on Ampere architecture, fabricated on an 8 nm process at Samsung. The RTX 2070 SUPER uses the TU104 chip based on Turing architecture, manufactured on a 12 nm process at TSMC. The process node difference alone explains part of the performance gap: the newer card packs 28,300 million transistors into a 628 mm² die, while the older card contains 13,600 million transistors on a 545 mm² die. Transistor density tells the story clearly, at 45.1M per mm² for the RTX 3080 versus 25.0M per mm² for the RTX 2070 SUPER.

The shading unit count diverges dramatically. The RTX 3080 has 8704 shading units, 272 texture mapping units, and 96 raster output units. The RTX 2070 SUPER has 2560 shading units, 160 TMUs, and 64 ROPs. This is not a modest increment; it is a fundamental scaling difference in the core compute pipeline.

Ray tracing hardware also differs. The RTX 3080 carries 68 RT cores, while the RTX 2070 SUPER has 40. Tensor cores show a different pattern: the RTX 3080 has 272, but the RTX 2070 SUPER has 320. This is one of the few areas where the older card has more hardware units, though the newer card's tensor cores operate within a more efficient architecture.

The FP32 and FP16 performance profiles diverge sharply. The RTX 3080 delivers 29.77 TFLOPS in both FP32 and FP16, a 1:1 ratio. The RTX 2070 SUPER delivers 9.062 TFLOPS in FP32 and 18.12 TFLOPS in FP16, a 2:1 ratio. This indicates that the Ampere architecture treats FP32 and FP16 as equally capable paths, while Turing halves its FP16 throughput relative to FP32.

Memory technology differs as well. The RTX 3080 uses 10 GB of GDDR6X on a 320-bit bus, achieving 760.3 GB/s bandwidth. The RTX 2070 SUPER uses 8 GB of GDDR6 on a 256-bit bus, achieving 448.0 GB/s. The memory clock is listed at 1188 MHz (19 Gbps effective) for the RTX 3080 and 1750 MHz (14 Gbps effective) for the RTX 2070 SUPER.

Specification Differences

The two cards differ across nearly every specification field. Clock speeds show the RTX 2070 SUPER with higher base and boost clocks: 1605 MHz and 1770 MHz respectively, versus 1440 MHz and 1710 MHz for the RTX 3080. Despite lower clocks, the RTX 3080 achieves higher throughput due to its larger execution resources.

Power requirements diverge significantly. The RTX 3080 has a TDP of 320 W and a suggested PSU rating of 700 W. The RTX 2070 SUPER has a TDP of 215 W and a suggested PSU of 550 W. The power connector configuration also differs: the RTX 3080 uses a single 12-pin connector, while the RTX 2070 SUPER uses one 6-pin plus one 8-pin.

The bus interface advances from PCIe 3.0 x16 on the RTX 2070 SUPER to PCIe 4.0 x16 on the RTX 3080. Physical dimensions shift as well: the RTX 3080 measures 285 mm in length, 112 mm in height, and 40 mm in width. The RTX 2070 SUPER is shorter at 267 mm, slightly taller at 116 mm, and thinner at 35 mm.

Display outputs differ. The RTX 3080 offers 1x HDMI 2.1 and 3x DisplayPort 1.4a. The RTX 2070 SUPER offers 1x HDMI 2.0, 3x DisplayPort 1.4a, and adds 1x USB Type-C. Both cards support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

Release dates and production status are identical in kind: both are end-of-life products. The RTX 3080 launched on 2020-08-31, while the RTX 2070 SUPER launched on 2019-07-08. The RTX 3080's predecessor is GeForce 20 and its successor is GeForce 40. The RTX 2070 SUPER's predecessor is GeForce 10 and its successor is GeForce 30.

FAQ

Q: Which card has the higher average benchmark score?

A: The RTX 3080 has an average benchmark score of 23172, while the RTX 2070 SUPER has an average of 20282. The RTX 3080 also sits at the 68th percentile among all GPUs, compared to the 65th percentile for the RTX 2070 SUPER.

Q: How large is the performance gap in DirectX 12 workloads?

A: In the PassMark DirectX 12 test, the RTX 3080 scores 100 versus the RTX 2070 SUPER's 67, a 49.3% advantage. In the 3DMark Steel Nomad DX12 test, the gap is 166.9% in favor of the RTX 3080.

Q: Does the RTX 2070 SUPER win any benchmarks?

A: Yes, it wins the Geekbench Vulkan test with a score of 90637, while the RTX 3080 scores 33620. This represents a 62.9% advantage for the RTX 2070 SUPER.

Q: What are the transistor density figures for each card?

A: The RTX 3080 has a transistor density of 45.1M per mm², while the RTX 2070 SUPER has 25.0M per mm². The RTX 3080 also carries more total transistors at 28,300 million versus 13,600 million.

Q: How do the memory bandwidth figures compare?

A: The RTX 3080 achieves 760.3 GB/s with 10 GB of GDDR6X on a 320-bit bus. The RTX 2070 SUPER achieves 448.0 GB/s with 8 GB of GDDR6 on a 256-bit bus.

Q: Which card has more tensor cores?

A: The RTX 2070 SUPER has 320 tensor cores, while the RTX 3080 has 272. Despite fewer tensor cores, the RTX 3080 delivers higher FP16 throughput at 29.77 TFLOPS versus 18.12 TFLOPS for the RTX 2070 SUPER.

The Verdict

The benchmark data is unambiguous in its overall direction. The RTX 3080 wins 9 of 10 head-to-head tests, with margins ranging from 15.7% to 166.9%. Its average benchmark score of 23172 exceeds the RTX 2070 SUPER's 20282 by a meaningful margin. Users seeking maximum performance in DirectX 11, DirectX 12, OpenCL, or general 3D workloads should choose the RTX 3080.

The RTX 2070 SUPER retains a clear niche in Vulkan-based applications, where its 90637 score far outstrips the RTX 3080's 33620. For developers or users specifically targeting Vulkan, the older card's advantage is substantial and cannot be ignored.

The architecture data supports the benchmark results. The RTX 3080's Ampere design, with its 8 nm process, 28,300 million transistors, and 8704 shading units, provides the hardware foundation for its dominant performance. The RTX 2070 SUPER's Turing architecture, with 13,600 million transistors and 2560 shading units, remains capable but operates at a lower performance tier.

The RTX 3080 also leads in memory bandwidth (760.3 GB/s versus 448.0 GB/s) and pixel rate (164.2 GPixel/s versus 113.3 GPixel/s). These are not marginal differences; they represent fundamental capacity gaps. The RTX 2070 SUPER's higher clock speeds (1770 MHz boost versus 1710 MHz) and larger tensor core count (320 versus 272) do not compensate for the RTX 3080's structural advantages.

For anyone weighing these two cards today, the data points decisively toward the RTX 3080 as the superior performer across most workloads. The RTX 2070 SUPER is the correct choice only in the narrow scenario where Vulkan performance is the primary requirement, and even then, the overall package of the RTX 3080 offers far more capability elsewhere. The recorded measurements show a generational shift that favors the newer architecture in nearly every measurable dimension.