AMD Radeon RX 590 vs NVIDIA GeForce RTX 2080 Ti Comparison

The NVIDIA GeForce RTX 2080 Ti and AMD Radeon RX 590 occupy the same 74th percentile of all GPUs, and their average benchmark scores are nearly identical at 30,459 and 30,457 respectively. This places them as direct rivals, with the RTX 2080 Ti holding a negligible 0% deltaPct lead over the RX 590. However, that aggregate parity masks a dramatic divergence in specific workloads, particularly in modern DirectX 12 scenarios where the Turing card's architecture and feature set pull far ahead. The data tells a story of two very different designs that happen to land at the same overall performance tier, but achieve that status through entirely different means.

Head-to-Head Benchmarks

The only shared benchmark between the two cards is the 3DMark Steel Nomad DX12 test, and the results are lopsided. The RTX 2080 Ti scores 3,537 points, while the RX 590 manages just 1,081 points. That is a 227.2% advantage for the NVIDIA card, meaning it delivers more than three times the performance in this particular workload. This is not a marginal win; it is a decisive generational gap in how each architecture handles DirectX 12's rendering pipeline, particularly around asynchronous compute and draw call efficiency.

Beyond that single head-to-head test, the benchmark suites diverge. The RTX 2080 Ti has a full set of scores across multiple APIs: it hits 134,929 in Geekbench OpenCL, 132,930 in Geekbench Vulkan, 21,548 in Passmark G3D, and 10,056 in Passmark GPU Compute. The RX 590, by contrast, only provides data for Geekbench Metal, where it scores 59,832, and the aforementioned 3DMark test. This lack of overlapping data means the Steel Nomad result is the only direct apples-to-apples comparison available, and it heavily favors the RTX 2080 Ti. In older DirectX 11 and DirectX 9 Passmark tests, the RTX 2080 Ti posts scores of 190 and 235 respectively, while its DirectX 10 score sits at 153 and DirectX 12 at 84 — figures that have no RX 590 counterpart to compare against.

What the aggregate data suggests is that the RX 590's overall average score is buoyed by its Metal performance and possibly other unlisted workloads, while the RTX 2080 Ti's average comes from a broader spread of tests. The 0% deltaPct between them in the nearestRivals list indicates that, on average, they are statistically indistinguishable in overall performance. Yet the 3DMark result reveals that when a workload is heavily threaded and uses modern API features, the RTX 2080 Ti is in a different league entirely. The RX 590's 1,081 score is less than a third of the RTX 2080 Ti's 3,537, and no amount of averaging can hide that chasm in raw compute throughput under DX12.

Architecture Differences

The architectural gap between these two cards is vast. The RTX 2080 Ti uses the TU102 chip built on TSMC's 12 nm process, packing 18,600 million transistors onto a 754 mm² die with a transistor density of 24.7M per mm². The RX 590 uses the Polaris 30 chip, also on a 12 nm process but fabricated by GlobalFoundries, with just 5,700 million transistors on a 232 mm² die at 24.6M per mm². While the process node and transistor density are nearly identical, the RTX 2080 Ti crams over three times as many transistors into a much larger physical area, which translates directly into more compute resources.

The RTX 2080 Ti is built on NVIDIA's Turing architecture, which introduces dedicated hardware for ray tracing and AI. It has 68 RT cores and 544 tensor cores, features that are entirely absent from the RX 590's GCN 4.0 architecture. The RX 590 has no RT cores and no tensor cores, making it a pure rasterization card. This is reflected in their API support: the RTX 2080 Ti supports DirectX 12 Ultimate (12_2) and Vulkan 1.4, while the RX 590 tops out at DirectX 12 (12_0) and Vulkan 1.3. The newer DirectX 12 Ultimate feature set includes hardware-accelerated ray tracing and mesh shaders, which the RX 590 cannot accelerate at all.

Shader resources also differ massively. The RTX 2080 Ti has 4,352 shading units, 272 texture mapping units (TMUs), and 88 raster operation pipelines (ROPs). The RX 590 has 2,304 shading units, 144 TMUs, and only 32 ROPs — roughly half the shading units and TMUs, and just over a third of the ROPs. This shows up in theoretical throughput: the RTX 2080 Ti delivers 13.45 TFLOPS FP32 and 26.90 TFLOPS FP16 (at a 2:1 ratio), while the RX 590 manages 7.119 TFLOPS FP32 and the same 7.119 TFLOPS FP16 (at a 1:1 ratio). The RTX 2080 Ti also has a much higher pixel rate at 136.0 GPixel/s versus 49.44 GPixel/s, and a texture rate of 420.2 GTexel/s versus 222.5 GTexel/s.

Memory is another clear divider. The RTX 2080 Ti has 11 GB of GDDR6 on a 352-bit bus, yielding 616.0 GB/s of bandwidth. The RX 590 has 8 GB of GDDR5 on a 256-bit bus, for 256.0 GB/s. That is a 2.4x bandwidth advantage for the NVIDIA card, which is critical for high-resolution textures and compute workloads. The RX 590's memory clock is 2000 MHz (8 Gbps effective), while the RTX 2080 Ti runs at 1750 MHz (14 Gbps effective) — the wider bus and faster memory type more than compensate for the lower clock speed. The RTX 2080 Ti also has a higher TDP at 250 W versus 175 W, requiring a 600 W suggested PSU and dual 8-pin connectors, while the RX 590 only needs a 450 W PSU and a single 8-pin connector.

The Verdict

From the data, the choice is clear for anyone prioritizing modern DirectX 12 gaming or compute workloads: the RTX 2080 Ti is the superior card, with a 227.2% lead in the only shared benchmark. Its Turing architecture brings hardware ray tracing, tensor cores, and DirectX 12 Ultimate support that the RX 590 simply cannot match. The RTX 2080 Ti also has more than double the shading units, nearly triple the ROPs, and 2.4x the memory bandwidth, all of which translate into headroom for future titles that leverage these features.

However, the aggregate benchmark scores tell a different story. Both cards sit at the 74th percentile, and their average scores are within 2 points of each other. This suggests that in older or lighter workloads, or in the specific mix of tests that feed the average, the RX 590 is far more competitive than the 3DMark result implies. The RX 590 also draws 75 W less power, needs a smaller PSU, and has a smaller die (232 mm² versus 754 mm²), which historically correlates with lower manufacturing cost and potentially better availability. For a builder targeting legacy DirectX 11 games or Vulkan titles that do not stress the RTX 2080 Ti's unique hardware, the RX 590 offers a viable path to similar average performance.

The verdict hinges on workload relevance. If the target is modern AAA titles with DX12 or ray tracing, the RTX 2080 Ti is the only rational choice — the RX 590's 1,081 Steel Nomad score indicates it will struggle in those scenarios. If the target is esports titles, older games, or general compute where the Metal and OpenCL scores matter, the RX 590's performance is statistically equivalent to the RTX 2080 Ti on average, making it a pragmatic pick for those specific use cases. The data does not support calling the RX 590 a "better value" — that is a pricing discussion outside this analysis — but it does support calling it a capable card for a narrower set of workloads.

Specification Differences

• Chip & Architecture: NVIDIA TU102 on Turing vs AMD Polaris 30 on GCN 4.0
• Foundry: TSMC vs GlobalFoundries
• Transistors: 18,600 million vs 5,700 million
• Die Size: 754 mm² vs 232 mm²
• Base Clock: 1350 MHz vs 1469 MHz (RX 590 has higher base clock)
• Memory: 11 GB GDDR6 vs 8 GB GDDR5
• Memory Bus: 352 bit vs 256 bit
• Memory Bandwidth: 616.0 GB/s vs 256.0 GB/s
• Memory Clock: 1750 MHz (14 Gbps effective) vs 2000 MHz (8 Gbps effective)
• Shading Units: 4352 vs 2304
• TMUs: 272 vs 144
• ROPs: 88 vs 32
• RT Cores: 68 vs none
• Tensor Cores: 544 vs none
• Pixel Rate: 136.0 GPixel/s vs 49.44 GPixel/s
• Texture Rate: 420.2 GTexel/s vs 222.5 GTexel/s
• FP32: 13.45 TFLOPS vs 7.119 TFLOPS
• FP16: 26.90 TFLOPS (2:1) vs 7.119 TFLOPS (1:1)
• TDP: 250 W vs 175 W
• Power Connectors: 2x 8-pin vs 1x 8-pin
• Suggested PSU: 600 W vs 450 W
• Display Outputs: 1x HDMI 2.0, 3x DisplayPort 1.4a, 1x USB Type-C vs 1x HDMI 2.0b, 3x DisplayPort 1.4a
• DirectX Support: 12 Ultimate (12_2) vs 12 (12_0)
• Vulkan Support: 1.4 vs 1.3
• Dimensions: 267 mm length vs 241 mm length
• Launch MSRP: 999 USD vs 279 USD

FAQ

Q: Which card has a higher average benchmark score?

A: The RTX 2080 Ti has an average score of 30,459, while the RX 590 has 30,457. The deltaPct between them is 0%, meaning they are statistically tied on average.

Q: How much faster is the RTX 2080 Ti in 3DMark Steel Nomad DX12?

A: The RTX 2080 Ti scores 3,537 versus the RX 590's 1,081, a 227.2% advantage for the NVIDIA card.

Q: Does the RX 590 support hardware ray tracing?

A: No. The RX 590 has no RT cores and no tensor cores, while the RTX 2080 Ti has 68 RT cores and 544 tensor cores.

Q: What is the memory bandwidth difference?

A: The RTX 2080 Ti has 616.0 GB/s of bandwidth from 11 GB of GDDR6 on a 352-bit bus. The RX 590 has 256.0 GB/s from 8 GB of GDDR5 on a 256-bit bus.

Q: Which card has a higher base clock speed?

A: The RX 590 has a base clock of 1469 MHz, which is higher than the RTX 2080 Ti's 1350 MHz. However, both cards have the same boost clock of 1545 MHz.

Q: What are the power requirements for each card?

A: The RTX 2080 Ti has a 250 W TDP and requires a 600 W PSU with dual 8-pin connectors. The RX 590 has a 175 W TDP and requires a 450 W PSU with a single 8-pin connector.

Where Each One Wins

The RTX 2080 Ti wins decisively in modern DirectX 12 workloads, as evidenced by its 227.2% lead in Steel Nomad. It also wins in any scenario that leverages its unique hardware: ray tracing via its 68 RT cores, AI acceleration via its 544 tensor cores, and DirectX 12 Ultimate features. Its 616.0 GB/s memory bandwidth and 13.45 TFLOPS FP32 make it the choice for high-resolution texture-heavy games and compute tasks like machine learning inference. The 11 GB VRAM capacity also provides more headroom for large datasets or texture packs, and its 420.2 GTexel/s texture rate and 136.0 GPixel/s pixel rate are overkill for any current rasterization task.

The RX 590 wins in power efficiency, drawing 75 W less than the RTX 2080 Ti, and in scenarios where its 1:1 FP16 ratio is beneficial — it delivers 7.119 TFLOPS FP16, which is the same as its FP32, whereas the RTX 2080 Ti's FP16 is a 2:1 rate that may not be exposed in all APIs. Its higher base clock of 1469 MHz suggests it may have better responsiveness in lightly threaded workloads that are clock-bound rather than throughput-bound. The RX 590 also wins in the average benchmark comparison, where it matches the RTX 2080