AMD Radeon RX 7900 XT vs NVIDIA GeForce GTX 590 Comparison

# AMD Radeon RX 7900 XT vs NVIDIA GeForce GTX 590

The head-to-head benchmark record between these two cards is remarkably lopsided, but not in the way the hardware generation gap might suggest. The NVIDIA GeForce GTX 590 wins the single shared test — Geekbench OpenCL — by a razor-thin 0.6% margin, scoring 12,830 against the RX 7900 XT’s 12,756. That result, however, is the only direct comparison available, and it paints a picture of near-identical compute performance in one specific workload while the architectural chasm between a 2011 dual-GPU flagship and a 2022 RDNA 3 part remains vast everywhere else.

Head-to-Head Benchmarks

The only overlapping benchmark in the data is Geekbench OpenCL, and the outcome is a statistical tie. The GTX 590 edges out the RX 7900 XT by 74 points (12,830 vs 12,756), a delta of -0.6% from the AMD card’s perspective. This is within the noise of typical run-to-run variance, and the result likely reflects OpenCL’s ability to leverage the GTX 590’s dual GF110 dies in a compute-heavy scenario that does not stress modern features like ray tracing or mesh shaders.

Looking at the broader benchmark suite, the RX 7900 XT’s other scores reveal where its real strengths lie. In 3DMark Steel Nomad DX12, it posts 5,623 points — a workload that the GTX 590 cannot even attempt, given its DirectX 12 (11_0) API level. The Vulkan gap is even more pronounced: the RX 7900 XT scores 71,068 in Geekbench Vulkan, while the GTX 590 has no Vulkan support at all (the API field is null). Passmark scores for the RX 7900 XT — 29,009 in G3D, 16,828 in GPU Compute, 1,232 in G2D — have no counterpart in the GTX 590’s data, which only includes the single OpenCL result.

What this means in practice: in the one test both cards can run, the 2011 card is competitive. In every modern API test that the RX 7900 XT supports, the GTX 590 is either absent or obsolete by design. The 0.6% OpenCL delta is an anomaly, not a trend — a legacy compute workload where raw FP32 throughput on two older dies can still trade blows with a modern single-die design.

Architecture Differences

The architectural divide here is generational, not incremental. The RX 7900 XT uses the Navi 31 chip built on TSMC’s 5 nm process, with 57,700 million transistors packed into a 529 mm² die. The GTX 590 uses the GF110 chip on TSMC’s 40 nm node, with just 3,000 million transistors on a 520 mm² die. That is a 19.2x transistor count difference on nearly identical physical area, yielding transistor densities of 109.1M per mm² for the RX 7900 XT versus 5.8M per mm² for the GTX 590.

The memory subsystems tell a similar story. The RX 7900 XT has 20 GB of GDDR6 on a 320-bit bus, delivering 800.0 GB/s of bandwidth. The GTX 590 has 1,536 MB of GDDR5 on a 384-bit bus, delivering 164.0 GB/s. The bandwidth advantage is 4.9x in favor of AMD, and the capacity advantage is 13.3x. Clock speeds differ as well: the RX 7900 XT runs a base of 1,387 MHz, a boost of 2,394 MHz, and a game clock of 2,025 MHz, while the GTX 590’s memory runs at 854 MHz (3.4 Gbps effective) and its core clocks are not listed in the data.

Compute resources are equally divergent. The RX 7900 XT has 5,376 shading units, 336 TMUs, 192 ROPs, and 84 ray tracing cores. The GTX 590 has 512 shading units, 64 TMUs, and 48 ROPs, with no ray tracing hardware. Pixel rate is 459.6 GPixel/s for AMD versus 19.46 GPixel/s for NVIDIA — a 23.6x difference. Texture rate is 804.4 GTexel/s versus 38.91 GTexel/s — a 20.7x difference. FP32 compute is 51.48 TFLOPS against 1,244.2 GFLOPS, a 41.4x gap. The RX 7900 XT also supports FP16 at 103.0 TFLOPS (2:1 ratio); the GTX 590 has no FP16 data.

API support further separates them. The RX 7900 XT runs DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The GTX 590 runs DirectX 12 (11_0) and OpenGL 4.6, but has no Vulkan support. The RX 7900 XT uses PCIe 4.0 x16; the GTX 590 uses PCIe 2.0 x16. Power draw is higher on the NVIDIA card at 365 W TDP versus 300 W, though the RX 7900 XT suggests a 700 W PSU versus 750 W for the GTX 590. Both are dual-slot cards with 2x 8-pin power connectors.

Where Each One Wins

The GTX 590 wins in exactly one measurable dimension: the Geekbench OpenCL score, where it beats the RX 7900 XT by 0.6%. It also has a 384-bit memory bus versus 320-bit, though the RX 7900 XT’s much faster GDDR6 more than compensates. The GTX 590 is slightly shorter (279 mm vs 276 mm? No — the RX 7900 XT is 276 mm, the GTX 590 is 279 mm) and thinner (40 mm vs 51 mm), which could matter in tight chassis, though both are dual-slot designs.

The RX 7900 XT wins everywhere else. It has 13.3x more memory capacity, 4.9x more bandwidth, 10.5x more shading units, 5.25x more TMUs, 4x more ROPs, 23.6x higher pixel rate, 20.7x higher texture rate, and 41.4x higher FP32 compute. It adds ray tracing cores, Vulkan 1.4 support, DirectX 12 Ultimate, and PCIe 4.0. It consumes 65 W less power. It has a 5 nm process node versus 40 nm, giving it a 18.8x transistor density advantage. It supports modern display outputs including HDMI 2.1a, DisplayPort 2.1, and USB Type-C, versus the GTX 590’s 3x DVI and 1x mini-DisplayPort.

In any workload that uses modern APIs, hardware-accelerated ray tracing, or large memory buffers, the RX 7900 XT is categorically superior. The GTX 590’s only niche is legacy OpenCL compute where its dual-die configuration can still produce competitive numbers.

FAQ

Q: Which card wins the only shared benchmark?

A: The NVIDIA GeForce GTX 590 wins Geekbench OpenCL with a score of 12,830 versus the AMD Radeon RX 7900 XT’s 12,756, a margin of 0.6%.

Q: Does the GTX 590 support Vulkan?

A: No. The GTX 590’s Vulkan field is null, while the RX 7900 XT supports Vulkan 1.4.

Q: How much more memory does the RX 7900 XT have?

A: The RX 7900 XT has 20 GB of GDDR6, while the GTX 590 has 1,536 MB of GDDR5 — a 13.3x capacity difference.

Q: What is the FP32 compute difference?

A: The RX 7900 XT delivers 51.48 TFLOPS, while the GTX 590 delivers 1,244.2 GFLOPS. The AMD card is 41.4x higher.

Q: Which card draws more power?

A: The GTX 590 has a 365 W TDP, while the RX 7900 XT has a 300 W TDP. The GTX 590 also recommends a 750 W PSU versus 700 W for the RX 7900 XT.

Q: Do both cards support DirectX 12?

A: Yes, but at different levels. The RX 7900 XT supports DirectX 12 Ultimate (12_2), while the GTX 590 supports DirectX 12 (11_0).

The Verdict

The data is unambiguous: the AMD Radeon RX 7900 XT is the superior card by every meaningful metric except one legacy compute benchmark. The 0.6% Geekbench OpenCL win for the GTX 590 is a statistical curiosity, not a competitive result — it does not offset the RX 7900 XT’s 41.4x FP32 advantage, 23.6x pixel rate advantage, or 20.7x texture rate advantage.

For modern gaming, content creation, or compute workloads, the RX 7900 XT is the only viable choice. It supports ray tracing, Vulkan 1.4, DirectX 12 Ultimate, has 20 GB of memory, and runs on a 5 nm process with 57,700 million transistors. The GTX 590 is a 2011 dual-GPU design with 3,000 million transistors on 40 nm, no ray tracing, no Vulkan, and 1,536 MB of memory.

The GTX 590’s percentile ranking of 52 versus the RX 7900 XT’s 55 reflects their overall standing in the GPU database, but that 3-point gap understates the real-world difference. The RX 7900 XT’s average benchmark score is 13,745, versus 12,830 for the GTX 590. The nearest rivals tell the story: the RX 7900 XT sits within 0.5-0.9% of cards like the AMD Radeon 660M and NVIDIA RTX A2000 Mobile, while the GTX 590 sits within 0.1-0.4% of cards like the AMD FirePro W5100 and NVIDIA GeForce GTX 670.

Anyone choosing between these two should pick the RX 7900 XT unless they specifically need a legacy OpenCL compute result — and even then, the 0.6% difference is negligible. The GTX 590’s launch MSRP was 699 USD; the RX 7900 XT’s launch MSRP was 899 USD. That 200 USD gap is the only real advantage the older card has, and it is dwarfed by the performance chasm.

Specification Differences

| Specification | AMD Radeon RX 7900 XT | NVIDIA GeForce GTX 590 |

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

| Architecture | RDNA 3.0 | Fermi 2.0 |

| Process Node | 5 nm | 40 nm |

| Transistors | 57,700 million | 3,000 million |

| Die Size | 529 mm² | 520 mm² |

| Transistor Density | 109.1M / mm² | 5.8M / mm² |

| Memory Size | 20 GB | 1536 MB |

| Memory Type | GDDR6 | GDDR5 |

| Memory Bus Width | 320 bit | 384 bit |

| Memory Bandwidth | 800.0 GB/s | 164.0 GB/s |

| Shading Units | 5376 | 512 |

| TMUs | 336 | 64 |

| ROPs | 192 | 48 |

| RT Cores | 84 | null |

| Pixel Rate | 459.6 GPixel/s | 19.46 GPixel/s |

| Texture Rate | 804.4 GTexel/s | 38.91 GTexel/s |

| FP32 | 51.48 TFLOPS | 1,244.2 GFLOPS |

| FP16 | 103.0 TFLOPS (2:1) | null |

| TDP | 300 W | 365 W |

| Suggested PSU | 700 W | 750 W |

| Bus Interface | PCIe 4.0 x16 | PCIe 2.0 x16 |

| DirectX | 12 Ultimate (12_2) | 12 (11_0) |

| Vulkan | 1.4 | null |

| Display Outputs | 1x HDMI 2.1a, 2x DisplayPort 2.1, 1x USB Type-C | 3x DVI, 1x mini-DisplayPort |

| Length | 276 mm | 279 mm |

| Height | 110 mm | 111 mm |

| Width | 51 mm | 40 mm |

| Release Date | 2022-11-02 | 2011-03-23 |

| Launch MSRP | 899 USD | 699 USD |