AMD Radeon RX 590 vs NVIDIA GeForce RTX 5060 Comparison

The data compares two GPUs separated by almost seven years of architecture, process technology, and feature sets. The NVIDIA GeForce RTX 5060 is a modern Blackwell 2.0 part built on a 5 nm process, while the AMD Radeon RX 590 is an end-of-life Polaris 30 card from the GCN 4.0 era on a 12 nm node. Their average benchmark scores, 26331 for the RTX 5060 and 24744 for the RX 590, place them just 6.4% apart overall, but that aggregate figure hides a decisive split: in the only head-to-head test available, the difference is not marginal but massive.

Where Each One Wins

The benchmark results are starkly one-sided in the available data. The RTX 5060 wins the sole head-to-head comparison, the 3DMark Steel Nomad DX12 test, with a score of 3628 against the RX 590’s 1081. That represents a 235.6% advantage, meaning the NVIDIA card delivers more than three times the performance in this modern DirectX 12 workload. For any application that stresses current-generation graphics APIs—ray-traced titles, high-fidelity DX12 games, or professional 3D rendering—the RTX 5060 is categorically in a different class.

The RX 590, however, does not post a single winning score in the head-to-head data. Its wins are contextual rather than absolute. The card’s 70th percentile ranking among all GPUs, compared to the RTX 5060’s 72nd percentile, suggests that in legacy or lighter workloads, the older card remains serviceable. Its GCN 4.0 architecture, while lacking dedicated ray tracing or tensor cores, still delivers 7.119 TFLOPS of FP32 throughput and a 256.0 GB/s memory bandwidth across a 256-bit bus. For older DirectX 9, 10, or 11 titles, or for compute tasks that do not leverage newer instruction sets, the RX 590 can hold its own—but the data does not show it winning any modern benchmark.

The RTX 5060’s wins extend across the board in its own benchmark suite. It posts a Geekbench OpenCL score of 112787 and a Vulkan score of 113321, both far beyond the RX 590’s single Geekbench Metal result of 48407. In PassMark tests, the RTX 5060 scores 20891 in G3D and 10899 in GPU Compute, while the RX 590 has no comparable entries. The NVIDIA card’s 19.18 TFLOPS FP32 performance, 448.0 GB/s bandwidth on GDDR7 memory, and 30 RT cores give it a structural advantage that no legacy title can overcome.

The Verdict

The choice between these two cards is not a matter of preference; it is a matter of era. The data shows the RTX 5060 is the only rational pick for any user running current or future software. Its 235.6% lead in 3DMark Steel Nomad is not an outlier but a reflection of architectural superiority: Blackwell 2.0 supports DirectX 12 Ultimate (12_2), while the RX 590 is limited to DirectX 12 (12_0). The RTX 5060 also brings Vulkan 1.4 support versus the RX 590’s Vulkan 1.3, and its 72nd percentile ranking versus the RX 590’s 70th is the narrowest gap in the entire comparison—the real spread is in modern workloads.

For users on a legacy system with a PCIe 3.0 slot and a 450 W PSU, the RX 590 might still function, but it is end-of-life. Its 175 W TDP and 5,700 million transistors on a 232 mm² die reflect a 2018 design that cannot be upgraded. The RTX 5060, by contrast, is active production, uses a 145 W TDP, and requires only a 300 W suggested PSU. It is smaller (181 mm² die, 21,900 million transistors on 5 nm), cooler, and faster. Unless a specific workload requires the RX 590’s 256-bit memory bus for legacy bandwidth-heavy compute—and no benchmark in the data confirms such a case—the RTX 5060 is the definitive answer.

The only argument for the RX 590 is cost, but that is outside the scope of this analysis. The RTX 5060’s launch MSRP is 299 USD, while the RX 590 launched at 279 USD. For a 20 USD difference at launch, the RTX 5060 offers 235.6% more performance in the head-to-head test, higher average scores, and a modern feature set. The verdict is unambiguous: choose the RTX 5060 for any new purchase or upgrade.

Head-to-Head Benchmarks

The sole head-to-head benchmark, 3DMark Steel Nomad DX12, tells the entire story. The RTX 5060 scores 3628, while the RX 590 scores 1081. The delta is 235.6% in favor of NVIDIA. This is not a marginal improvement; it is a generational leap. Steel Nomad is a demanding test designed to scale with modern GPU features, and the RX 590’s lack of RT cores and tensor cores—both present on the RTX 5060 (30 and 120, respectively)—explains the gulf. The RTX 5060’s 19.18 TFLOPS FP32 output is 2.7 times the RX 590’s 7.119 TFLOPS, which aligns closely with the 3.36x score difference.

In the broader benchmark suite, the RTX 5060’s Geekbench OpenCL score of 112787 and Vulkan score of 113321 dwarf the RX 590’s Geekbench Metal score of 48407. While Metal is a different API (Apple’s), the magnitude of the gap—roughly 2.3x in favor of NVIDIA—is consistent with the Steel Nomad result. The PassMark scores for the RTX 5060 (G3D: 20891, GPU Compute: 10899) further confirm its compute and rasterization dominance. The RX 590 has no PassMark entries in the data, so direct comparison there is impossible, but the average benchmark score gap (26331 vs 24744, a 6.4% difference) is misleading; it is dragged down by the RX 590’s inclusion of only two benchmarks, both of which are less demanding than Steel Nomad.

The RTX 5060’s nearest rivals in the data are the AMD Radeon 860M (avg score 26401, -0.3% delta), NVIDIA GeForce MX550 (26421, -0.3%), AMD Radeon RX 5700 XT 50th Anniversary (26553, -0.8%), and AMD Radeon RX 6750 XT (26011, +1.2%). This places the RTX 5060 in a performance tier just below the RX 5700 XT 50th Anniversary, but with modern features the older card lacks. The RX 590’s nearest rivals are the NVIDIA RTX A5000 Mobile (24763, -0.1%), Intel Arc A350M (24647, +0.4%), AMD Radeon RX 6600 XT (24442, +1.2%), and NVIDIA GeForce GTX 1630 (24277, +1.9%). This shows the RX 590 sits in a much lower performance class, roughly on par with entry-level mobile GPUs from the same era.

The 3DMark Steel Nomad result is the only direct comparison, but it is sufficient. A 235.6% delta in a modern DX12 test is a categorical win. The RTX 5060’s 448.0 GB/s memory bandwidth on a 128-bit GDDR7 bus outperforms the RX 590’s 256.0 GB/s on a 256-bit GDDR5 bus, demonstrating that newer memory technology compensates for a narrower bus. The RTX 5060’s 299.6 GTexel/s texture rate and 119.9 GPixel/s pixel rate also overwhelm the RX 590’s 222.5 GTexel/s and 49.44 GPixel/s, respectively. Every measurable metric favors the RTX 5060.

FAQ

Q: Is the RTX 5060 always faster than the RX 590?

A: In the only head-to-head benchmark available, 3DMark Steel Nomad DX12, the RTX 5060 scores 3628 versus the RX 590's 1081, a 235.6% advantage. The RTX 5060 also has higher average benchmark scores (26331 vs 24744) and a higher percentile ranking (72nd vs 70th). No benchmark in the data shows the RX 590 winning.

Q: What is the architectural difference that explains the performance gap?

A: The RTX 5060 uses the Blackwell 2.0 architecture on a 5 nm TSMC process with 21,900 million transistors, while the RX 590 uses GCN 4.0 on a 12 nm GlobalFoundries process with 5,700 million transistors. The RTX 5060 also has 30 RT cores and 120 tensor cores, which the RX 590 lacks entirely.

Q: Does the RX 590 have any advantages in memory bandwidth?

A: The RX 590 has a wider 256-bit memory bus compared to the RTX 5060's 128-bit bus. However, the RTX 5060's GDDR7 memory at 448.0 GB/s bandwidth is 75% higher than the RX 590's GDDR5 at 256.0 GB/s, compensating for the narrower bus.

Q: What API support differences exist between the two cards?

A: The RTX 5060 supports DirectX 12 Ultimate (12_2) and Vulkan 1.4, while the RX 590 is limited to DirectX 12 (12_0) and Vulkan 1.3. This means the RTX 5060 can run the latest DX12 Ultimate features, which the RX 590 cannot.

Q: How do their power requirements compare?

A: The RTX 5060 has a 145 W TDP and requires a 300 W suggested PSU, while the RX 590 has a 175 W TDP and requires a 450 W suggested PSU. Both use a dual-slot design and a single 8-pin power connector.

Q: Which card is better for a new build in 2025?

A: The RTX 5060 is the only viable choice for a new build. It is in active production, released in 2025, and delivers 235.6% more performance in the head-to-head test. The RX 590 is end-of-life, released in 2018, and its 70th percentile ranking places it near entry-level mobile GPUs like the Intel Arc A350M.

Architecture Differences

The RTX 5060 and RX 590 are separated by more than time; they are fundamentally different designs. The RTX 5060 is built on the Blackwell 2.0 architecture using TSMC's 5 nm process, packing 21,900 million transistors into a 181 mm² die for a density of 121.0 million transistors per mm². The RX 590 uses GCN 4.0 on GlobalFoundries' 12 nm process, with 5,700 million transistors spread across a larger 232 mm² die, yielding just 24.6 million transistors per mm². This density difference—nearly 5x—is the root cause of the performance gap.

The memory subsystems reflect the generational leap. The RTX 5060 uses 8 GB of GDDR7 memory on a 128-bit bus, achieving 448.0 GB/s bandwidth. The RX 590 uses 8 GB of GDDR5 on a 256-bit bus, achieving 256.0 GB/s. Despite having twice the bus width, the RX 590 delivers 42.9% less bandwidth because GDDR7 is far more efficient. The RTX 5060's memory clock is 1750 MHz (28 Gbps effective), while the RX 590's is 2000 MHz (8 Gbps effective).

Compute resources also diverge sharply. The RTX 5060 has 3840 shading units, 120 TMUs, and 48 ROPs, plus 30 RT cores and 120 tensor cores. The RX 590 has 2304 shading units, 144 TMUs, and 32 ROPs, with no RT or tensor cores. The NVIDIA card's FP32 throughput is 19.18 TFLOPS versus 7.119 TFLOPS for AMD, a 2.7x difference. Pixel and texture rates follow suit: 119.9 GPixel/s and 299.6 GTexel/s for the RTX 5060, versus 49.44 GPixel/s and 222.5 GTexel/s for the RX 590.

The feature sets are equally divergent. The RTX 5060 supports PCIe 5.0 x8, while the RX 590 uses PCIe 3.0 x16. Display outputs differ: the RTX 5060 has 1x HDMI 2.1b and 3x DisplayPort 2.1b, while the RX 590 has 1x HDMI 2.0b and 3x DisplayPort 1.4a. API support is a clear differentiator, with the RTX 5060 offering DirectX 12 Ultimate (12_2) and Vulkan 1.4, versus the RX 590's DirectX 12 (12_0) and Vulkan 1.3. The RTX 5060 also has a lower TDP (145 W vs 175 W) and a lower suggested PSU (300 W vs 450 W), despite being dramatically faster. Both cards are dual-slot with a single 8-pin connector and share the same 241 mm length. The RTX 5060 is 111 mm tall and 40 mm wide; the RX 590's height and width are not listed. The only category where the RX 590 is not inferior is its 256-bit memory bus, but as the bandwidth figures show, that advantage is nullified by newer memory technology.