AMD Radeon RX 590 vs Intel Arc B570 Comparison

Where Each One Wins

The recorded benchmark data splits cleanly between two very different GPU generations. The AMD Radeon RX 590 holds a single benchmark entry, while the Intel Arc B570 carries a broader suite of results across multiple test suites. Based on the available measurements, the Intel part wins the only direct head-to-head comparison, but the AMD card's average benchmark score sits higher than the Intel card's average, which complicates any simple "winner" narrative.

Looking at the aggregate numbers, the RX 590 records an average benchmark score of 24,744 across its two tests, placing it in the 70th percentile of all GPUs in the database. The Arc B570, despite winning the head-to-head test, has a lower average benchmark score of 20,556 and sits in the 65th percentile. This discrepancy stems from the different test sets each card was subjected to. The RX 590's results come from 3DMark Steel Nomad (DX12) and Geekbench Metal, while the Arc B570 was tested across 3DMark Steel Nomad, Geekbench OpenCL, Geekbench Vulkan, and four Passmark sub-tests (DX9, DX10, DX11, DX12), plus Passmark G2D, G3D, and GPU Compute.

In the only shared test, 3DMark Steel Nomad DX12, the Arc B570 scores 2,649 against the RX 590's 1,081, a delta of -59.2% from the AMD card's perspective. That is a decisive victory for Intel's architecture in a modern DX12 workload. However, the RX 590's Geekbench Metal score of 48,407 suggests strong performance in Apple's graphics API, a domain where the Arc B570 has no recorded data. For users prioritizing Metal-based workflows, the RX 590 is the only option with measured evidence.

The Arc B570's Passmark results reveal a mixed profile. Its G3D score of 14,195 is respectable, and its GPU Compute score of 7,281 indicates solid compute throughput. But its DirectX 10 score of 65 and DirectX 11 score of 118 are notably low, suggesting that legacy DX10/DX11 titles may not be a strength. The DX9 score of 164 is higher relative to the other legacy tests, yet still far below modern expectations. The G2D score of 661 reflects 2D rendering performance, which is not a primary gaming metric but does factor into desktop and productivity workloads.

The AMD card's nearest rivals include the NVIDIA RTX A5000 Mobile (average 24,763, delta -0.1%), Intel Arc A350M (24,647, delta 0.4%), AMD Radeon RX 6600 XT (24,442, delta 1.2%), and NVIDIA GeForce GTX 1630 (24,277, delta 1.9%). The RX 590 sits essentially at parity with the mobile RTX A5000 and slightly ahead of the Arc A350M. The Intel Arc B570's nearest rivals are the NVIDIA GeForce RTX 3070 Mobile (20,534, delta 0.1%), Intel Arc A750 (20,582, delta -0.1%), NVIDIA Quadro M4000M (20,480, delta 0.4%), and AMD Radeon R9 M390X (20,662, delta -0.5%). The Arc B570 is effectively tied with the RTX 3070 Mobile and Arc A750, all within half a percent of each other.

Architecture Differences

The two cards represent fundamentally different design philosophies separated by six years of silicon evolution. The RX 590 uses the Polaris 30 chip built on GlobalFoundries' 12 nm process, employing the GCN 4.0 architecture. It packs 5,700 million transistors into a 232 mm² die, yielding a transistor density of 24.6 million per square millimeter. The Arc B570, by contrast, uses Intel's BMG-G21 chip based on the Xe2-HPG architecture, fabricated by TSMC on a 5 nm process. It contains 19,600 million transistors across a 272 mm² die, achieving a density of 72.1 million per square millimeter, nearly three times the RX 590's density.

Clock speeds differ substantially. The RX 590 runs at a 1,469 MHz base and 1,545 MHz boost, while the Arc B570 operates at a flat 2,500 MHz for both base and boost. That is a 955 MHz advantage in boost clock for Intel. Memory configurations also diverge: the RX 590 uses 8 GB of GDDR5 on a 256-bit bus with 2,000 MHz memory clock (8 Gbps effective), delivering 256.0 GB/s bandwidth. The Arc B570 uses 10 GB of GDDR6 on a 160-bit bus with 2,375 MHz memory clock (19 Gbps effective), delivering 380.0 GB/s. Despite the narrower bus, the faster memory type gives Intel a 124 GB/s bandwidth advantage.

Shader resources are identical in count: both cards have 2,304 shading units and 144 texture mapping units. The ROP count differs dramatically, however. The RX 590 has 32 ROPs, while the Arc B570 has 80 ROPs, a 2.5x increase. This directly impacts pixel throughput: the RX 590 achieves 49.44 GPixel/s versus the Arc B570's 200.0 GPixel/s. Texture rates also favor Intel, 360.0 GTexel/s against 222.5 GTexel/s.

Compute capabilities show the generational gap. The RX 590 delivers 7.119 TFLOPS of FP32 and the same 7.119 TFLOPS for FP16 at a 1:1 ratio. The Arc B570 offers 11.52 TFLOPS FP32 and 23.04 TFLOPS FP16 at a 2:1 ratio, meaning Intel's card has both higher absolute FP32 and far superior FP16 throughput. The Arc B570 also includes 18 ray tracing cores, a feature entirely absent from the RX 590, which has no RT cores listed. Tensor cores are absent from both cards.

Power and interface specs also shift. The RX 590 has a 175 W TDP, while the Arc B570 draws 150 W despite being substantially faster. Both are dual-slot cards with a single 8-pin power connector and a 450 W suggested PSU. The RX 590 uses PCIe 3.0 x16, while the Arc B570 moves to PCIe 4.0 x8. Display outputs differ: the RX 590 offers 1x HDMI 2.0b and 3x DisplayPort 1.4a, while the Arc B570 provides 1x HDMI 2.1a and 3x DisplayPort 2.1, supporting newer display standards.

API support reflects the architectural update. The RX 590 supports DirectX 12 (12_0), OpenGL 4.6, and Vulkan 1.3. The Arc B570 supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, adding ray tracing and mesh shader support through the 12_2 feature level. Physical dimensions also differ: the RX 590 is 241 mm (9.5 inches) long with no listed height or width, while the Arc B570 is 272 mm (10.7 inches) long and 115 mm (4.5 inches) tall. Production status separates them further: the RX 590 is end-of-life, released on 2018-11-14, while the Arc B570 is active, released on 2025-01-15.

FAQ

Q: Which card has the higher average benchmark score?

A: The AMD Radeon RX 590 records an average benchmark score of 24,744, placing it in the 70th percentile of all GPUs. The Intel Arc B570 has an average of 20,556, placing it in the 65th percentile.

Q: How do the two cards compare in the only shared benchmark test?

A: In 3DMark Steel Nomad DX12, the Intel Arc B570 scores 2,649 while the RX 590 scores 1,081. The delta is -59.2% from the AMD card's perspective, meaning Intel wins by a wide margin.

Q: Does the RX 590 have any ray tracing capability?

A: No. The RX 590 has no ray tracing cores listed in the database. The Intel Arc B570 includes 18 ray tracing cores.

Q: What is the memory bandwidth difference between the two cards?

A: The RX 590 provides 256.0 GB/s bandwidth using 8 GB of GDDR5 on a 256-bit bus. The Arc B570 provides 380.0 GB/s bandwidth using 10 GB of GDDR6 on a 160-bit bus, a 124 GB/s advantage for Intel.

Q: Which card supports newer display outputs?

A: The Intel Arc B570 supports 1x HDMI 2.1a and 3x DisplayPort 2.1. The RX 590 supports 1x HDMI 2.0b and 3x DisplayPort 1.4a.

Q: How do the ROP counts differ?

A: The RX 590 has 32 ROPs, while the Arc B570 has 80 ROPs. This contributes to the Intel card's 200.0 GPixel/s pixel rate versus the AMD card's 49.44 GPixel/s.

Specification Differences

The table below highlights only the fields where the two cards differ in the recorded data.

| Specification | AMD Radeon RX 590 | Intel Arc B570 |

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

| Chip | Polaris 30 | BMG-G21 |

| Architecture | GCN 4.0 | Xe2-HPG |

| Generation | Polaris (RX 500) | Battlemage (Arc 5) |

| Process Node | 12 nm | 5 nm |

| Foundry | GlobalFoundries | TSMC |

| Transistors | 5,700 million | 19,600 million |

| Die Size | 232 mm² | 272 mm² |

| Transistor Density | 24.6M / mm² | 72.1M / mm² |

| Base Clock | 1469 MHz | 2500 MHz |

| Boost Clock | 1545 MHz | 2500 MHz |

| Memory Clock | 2000 MHz, 8 Gbps effective | 2375 MHz, 19 Gbps effective |

| Memory Size | 8 GB | 10 GB |

| Memory Type | GDDR5 | GDDR6 |

| Memory Bus Width | 256 bit | 160 bit |

| Memory Bandwidth | 256.0 GB/s | 380.0 GB/s |

| ROPs | 32 | 80 |

| RT Cores | None | 18 |

| Pixel Rate | 49.44 GPixel/s | 200.0 GPixel/s |

| Texture Rate | 222.5 GTexel/s | 360.0 GTexel/s |

| FP32 | 7.119 TFLOPS | 11.52 TFLOPS |

| FP16 | 7.119 TFLOPS (1:1) | 23.04 TFLOPS (2:1) |

| TDP | 175 W | 150 W |

| Bus Interface | PCIe 3.0 x16 | PCIe 4.0 x8 |

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

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

| Vulkan Support | 1.3 | 1.4 |

| Length | 241 mm (9.5 inches) | 272 mm (10.7 inches) |

| Height | Not listed | 115 mm (4.5 inches) |

| Production Status | End-of-life | Active |

| Release Date | 2018-11-14 | 2025-01-15 |

| Predecessor | Arctic Islands | Alchemist |

| Successor | Vega | None |

| Launch MSRP | 279 USD | 219 USD |

Head-to-Head Benchmarks

The database contains exactly one shared benchmark between these two cards: 3DMark Steel Nomad DX12. In this test, the Intel Arc B570 scores 2,649 against the RX 590's 1,081. The delta percentage of -59.2% indicates the AMD card falls short by nearly three-fifths of the Intel score. This is the only direct comparison available, and it favors Intel overwhelmingly.

To contextualize this result, consider the nearest rivals for each card. The RX 590's average score of 24,744 places it within 0.1% of the NVIDIA RTX A5000 Mobile (24,763) and 0.4% ahead of the Intel Arc A350M (24,647). The Arc B570's average of 20,556 places it within 0.1% of the NVIDIA GeForce RTX 3070 Mobile (20,534) and within 0.1% of the Intel Arc A750 (20,582). The RX 590's average sits roughly 20% higher than the Arc B570's, yet the Arc B570 dominates the one test where both have data.

The RX 590's other recorded benchmark, Geekbench Metal at 48,407, has no counterpart for the Arc B570. The Arc B570's Geekbench OpenCL score of 83,514 and Vulkan score of 96,844 have no RX 590 equivalents. These are separate ecosystems, and the data does not permit direct comparisons across these APIs.

Among the Arc B570's Passmark results, the G3D score of 14,195 stands out as the most game-relevant figure. The GPU Compute score of 7,281 suggests strong parallel processing. The DX11 score of 118 is the weakest point, indicating poor legacy DirectX 11 performance. By contrast, the DX9 score of 164 and DX10 score of 65 show that older DX9 titles fare better than DX10 or DX11 ones, though none of these numbers approach the modern DX12 performance level implied by the 3DMark result.

The wins tally is unambiguous: the Arc B570 claims 1 win in the head-to-head set, while the RX 590 claims 0. However, the average score gap tells a different story. The RX 590's 70th percentile ranking versus the Arc B570's 65th percentile suggests that, across the full spectrum of recorded tests, the older AMD card holds a slight overall standing. The Intel card's single-test victory is decisive in that specific workload, but the broader dataset shows the RX 590 maintaining a higher aggregate position among all GPUs. This tension between a narrow head-to-head loss and a higher average score is typical when comparing cards from different eras with different test coverage.