AMD Radeon RX 590 vs NVIDIA GeForce RTX 3080 Comparison

FAQ

Q: How do the average benchmark scores of the AMD Radeon RX 590 and NVIDIA GeForce RTX 3080 compare?

A: The RX 590 has an average benchmark score of 24,744, while the RTX 3080 scores 23,172. This puts the RX 590 at a 6.8% higher average score in the database, despite the RTX 3080 winning the head-to-head test.

Q: Which GPU has the higher percentile ranking among all GPUs?

A: The RX 590 ranks in the 70th percentile, while the RTX 3080 sits in the 68th percentile. The data indicates the RX 590 outperforms nearly 70% of all GPUs, whereas the RTX 3080 outperforms roughly 68%.

Q: What is the difference in process node size between the two cards?

A: The RX 590 uses a 12 nm process at GlobalFoundries, while the RTX 3080 uses an 8 nm process at Samsung. The RTX 3080's smaller node allows for significantly higher transistor density: 45.1M transistors per mm² versus 24.6M for the RX 590.

Q: How do memory specifications differ between the two?

A: The RX 590 has 8 GB of GDDR5 on a 256-bit bus with 256.0 GB/s bandwidth. The RTX 3080 has 10 GB of GDDR6X on a 320-bit bus with 760.3 GB/s bandwidth — roughly triple the memory bandwidth.

Q: What is the head-to-head result in the 3DMark Steel Nomad DX12 test?

A: The RTX 3080 scores 4,407 versus the RX 590's 1,081, a delta of -75.5% for the RX 590. This is the only shared benchmark in the data, and the RTX 3080 wins it decisively.

Q: Which card has a higher FP32 compute throughput?

A: The RTX 3080 delivers 29.77 TFLOPS FP32, while the RX 590 delivers 7.119 TFLOPS. The RTX 3080 is roughly 4.2 times higher in raw single-precision compute.

The Verdict

The data presents a split verdict. In the only direct head-to-head benchmark available — 3DMark Steel Nomad DX12 — the NVIDIA GeForce RTX 3080 wins outright with a score of 4,407 against 1,081 for the AMD Radeon RX 590, a 75.5% margin. This is not a close contest in that specific workload.

However, the overall database picture complicates matters. The RX 590's average benchmark score of 24,744 exceeds the RTX 3080's 23,172, and its percentile rank of 70 beats the RTX 3080's 68. The RX 590's nearest rivals include the NVIDIA RTX A5000 Mobile (delta -0.1%), Intel Arc A350M (delta +0.4%), and AMD Radeon RX 6600 XT (delta +1.2%). The RTX 3080's nearest rivals — NVIDIA P106-100, AMD Radeon Pro Vega 16, AMD Radeon RX 6600M, and AMD Radeon R9 M290X — all sit within 0.4% of its average score.

For a buyer prioritizing the 3DMark Steel Nomad DX12 workload, the RTX 3080 is the clear choice. For a buyer looking at aggregate multi-test performance across the database, the RX 590 appears marginally stronger. The RTX 3080 offers substantially higher raw specifications — 8,704 shading units, 29.77 TFLOPS FP32, and 760.3 GB/s bandwidth — but the benchmark aggregate does not reflect that advantage in average score.

The verdict from the data: pick the RTX 3080 for DX12-heavy workloads and raw compute; pick the RX 590 if the broader benchmark average and percentile ranking matter more to you. The RTX 3080's launch MSRP is 699 USD; the RX 590's launch MSRP is 279 USD.

Head-to-Head Benchmarks

The single head-to-head result in the data is the 3DMark Steel Nomad DX12 test. The RTX 3080 scores 4,407, and the RX 590 scores 1,081. The delta percentage is -75.5%, meaning the RX 590 trails by three-quarters of the RTX 3080's score. This is a massive gap. In this DX12 workload, the RTX 3080 delivers roughly four times the performance of the RX 590.

The RTX 3080's other benchmark results give context. Its PassMark G3D score is 25,086, and its Geekbench OpenCL score is 152,423. Its Geekbench Vulkan score is 33,620. The RX 590's Geekbench Metal score is 48,407. These are different test suites, so direct comparison is not possible, but the RTX 3080's OpenCL score of 152,423 dwarfs the RX 590's Metal score of 48,407 in absolute terms.

The RTX 3080 also shows consistent performance across PassMark API tests: 258 in DX9, 207 in DX11, 170 in DX10, and 100 in DX12 (the latter normalized to 100 as a baseline in that suite). Its PassMark GPU Compute score is 14,397. The RX 590 has no equivalent PassMark entries in the data, so cross-suite comparisons are limited.

The takeaway from the head-to-head data is clear: in the one shared test, the RTX 3080 wins by a 75.5% margin, and its standalone scores across OpenCL, Vulkan, and PassMark are consistently strong. The RX 590's aggregate average is higher, but that average comes from a different set of benchmarks that may weight older or different workloads more heavily.

Specification Differences

| Specification | AMD Radeon RX 590 | NVIDIA GeForce RTX 3080 |

|---|---|---|

| Process Node | 12 nm | 8 nm |

| Foundry | GlobalFoundries | Samsung |

| Transistors | 5,700 million | 28,300 million |

| Die Size | 232 mm² | 628 mm² |

| Transistor Density | 24.6M / mm² | 45.1M / mm² |

| Base Clock | 1469 MHz | 1440 MHz |

| Boost Clock | 1545 MHz | 1710 MHz |

| Memory Size | 8 GB GDDR5 | 10 GB GDDR6X |

| Memory Bus Width | 256 bit | 320 bit |

| Memory Bandwidth | 256.0 GB/s | 760.3 GB/s |

| Shading Units | 2304 | 8704 |

| TMUs | 144 | 272 |

| ROPs | 32 | 96 |

| RT Cores | None | 68 |

| Tensor Cores | None | 272 |

| Pixel Rate | 49.44 GPixel/s | 164.2 GPixel/s |

| Texture Rate | 222.5 GTexel/s | 465.1 GTexel/s |

| FP32 | 7.119 TFLOPS | 29.77 TFLOPS |

| FP16 | 7.119 TFLOPS (1:1) | 29.77 TFLOPS (1:1) |

| TDP | 175 W | 320 W |

| Power Connectors | 1x 8-pin | 1x 12-pin |

| Suggested PSU | 450 W | 700 W |

| Bus Interface | PCIe 3.0 x16 | PCIe 4.0 x16 |

| Display Outputs | 1x HDMI 2.0b, 3x DisplayPort 1.4a | 1x HDMI 2.1, 3x DisplayPort 1.4a |

| DirectX Support | 12 (12_0) | 12 Ultimate (12_2) |

| Vulkan Support | 1.3 | 1.4 |

| Length | 241 mm (9.5 inches) | 285 mm (11.2 inches) |

| Release Date | 2018-11-14 | 2020-08-31 |

The RTX 3080 has roughly five times the transistor count, nearly triple the memory bandwidth, and over four times the FP32 throughput. It also adds RT cores (68) and tensor cores (272), which the RX 590 lacks entirely. The RX 590 has a higher base clock (1469 MHz vs 1440 MHz) but a lower boost clock (1545 MHz vs 1710 MHz).

Architecture Differences

The RX 590 uses the Polaris 30 chip based on GCN 4.0 architecture, built on GlobalFoundries' 12 nm process. The RTX 3080 uses the GA102 chip based on Ampere architecture, built on Samsung's 8 nm process. These are fundamentally different designs from different eras.

The transistor counts tell the story of the architectural gap. The RTX 3080 packs 28,300 million transistors into a 628 mm² die, while the RX 590 has 5,700 million in 232 mm². The transistor density is nearly double on the RTX 3080: 45.1M per mm² versus 24.6M. This density difference directly enables the RTX 3080's 8,704 shading units versus 2,304 on the RX 590.

Memory architecture differs completely. The RX 590 uses GDDR5 with 8 GB capacity and 256.0 GB/s bandwidth. The RTX 3080 uses GDDR6X with 10 GB capacity and 760.3 GB/s bandwidth. The 320-bit bus on the RTX 3080 is 25% wider than the RX 590's 256-bit bus, but the memory type change accounts for most of the bandwidth jump.

The RTX 3080 includes dedicated ray tracing hardware (68 RT cores) and tensor cores (272) for AI workloads. The RX 590 has neither. This is a structural difference, not just a performance gap — the RTX 3080 can accelerate workloads that the RX 590 cannot accelerate at all.

API support also differs. The RTX 3080 supports DirectX 12 Ultimate (12_2) and Vulkan 1.4, while the RX 590 supports DirectX 12 (12_0) and Vulkan 1.3. Both support OpenGL 4.6. The RTX 3080's PCIe 4.0 x16 interface doubles the bandwidth of the RX 590's PCIe 3.0 x16.

The RX 590's FP16 throughput matches its FP32 (1:1 ratio). The RTX 3080 also has a 1:1 ratio, but at a much higher absolute level: 29.77 TFLOPS FP16 versus 7.119 TFLOPS.

Where Each One Wins

NVIDIA GeForce RTX 3080 wins in DX12 workloads. The 3DMark Steel Nomad DX12 result is the only direct comparison, and the RTX 3080 wins by 75.5%. The RX 590's DirectX 12_0 support is older than the RTX 3080's 12_2, which may explain part of this gap.

NVIDIA GeForce RTX 3080 wins in raw compute throughput. At 29.77 TFLOPS FP32, it offers over four times the RX 590's 7.119 TFLOPS. Its PassMark GPU Compute score of 14,397 further reflects this strength.

NVIDIA GeForce RTX 3080 wins in memory-bound scenarios. With 760.3 GB/s bandwidth versus 256.0 GB/s, the RTX 3080 can feed its 8,704 shading units far more effectively. The 10 GB GDDR6X capacity also exceeds the RX 590's 8 GB GDDR5.

AMD Radeon RX 590 wins in aggregate benchmark average. The RX 590's 24,744 average score beats the RTX 3080's 23,172. Its 70th percentile rank also edges out the RTX 3080's 68th.

AMD Radeon RX 590 wins in power efficiency. At 175 W TDP versus 320 W, the RX 590 draws nearly half the power. Its suggested PSU of 450 W is also lower than the RTX 3080's 700 W requirement.

AMD Radeon RX 590 wins in physical footprint. At 241 mm length, it fits in smaller cases than the RTX 3080's 285 mm length. The RX 590 also uses a single 8-pin power connector versus the RTX 3080's 12-pin.

The use-case split is clear. For gamers running DX12 titles, compute-heavy workloads, or ray-traced content, the RTX 3080 is the data-backed choice. For users with power or space constraints, or those whose workload matches the benchmark aggregate, the RX 590 holds its own. The RTX 3080's nearest rivals all sit within 0.4% of its average score, suggesting it competes in a tight cluster, while the RX 590's nearest rivals span a wider range (up to 1.9% delta), indicating more variance in its competitive position.