AMD Radeon Pro 575 vs AMD Radeon RX 590 GME Comparison

Where Each One Wins

The recorded data splits these two AMD parts cleanly by workload type. The AMD Radeon RX 590 GME takes both head-to-head benchmark wins, which gives it a clear overall advantage in the database’s direct comparisons. The AMD Radeon Pro 575, however, posts a higher average benchmark score across all recorded tests, which suggests it performs better in a broader mixed workload set even though it loses the direct matchups.

Looking at the individual tests, the RX 590 GME wins the Geekbench OpenCL test with a score of 36,294 against the Pro 575’s 34,596, a margin of 4.7%. That is a modest but consistent lead in a compute-oriented API. The Vulkan test tells a more dramatic story: the RX 590 GME scores 60,508 versus 37,878 for the Pro 575, a 37.4% advantage. If your workload leans on Vulkan, the choice is unambiguous.

The Pro 575’s average benchmark score of 39,555 versus the RX 590 GME’s 32,601 indicates that the Pro card performs better across the full database of tests, many of which may favor its professional positioning. The Pro 575 also sits at the 82nd percentile of all GPUs, while the RX 590 GME sits at the 77th percentile. So while the RX 590 GME dominates the two direct comparisons, the Pro 575 has a stronger overall standing in the broader benchmark universe.

For a use-case split, the RX 590 GME is the pick for Vulkan-heavy applications, where its lead is massive, and for OpenCL tasks where it holds a smaller but real edge. The Pro 575 is better suited for a mixed bag of professional or legacy workloads, as its higher average score and percentile suggest it handles a wider variety of tests without the same dramatic swings.

Architecture Differences

Both cards share the same fundamental architecture: GCN 4.0, built on a 14 nm process at GlobalFoundries. Both use a 5,700 million transistor design on a 232 mm² die, with a transistor density of 24.6M per mm². The chip names differ, Ellesmere for the Pro 575 and Polaris 20 for the RX 590 GME, but the underlying silicon is essentially the same family.

The key architectural difference is in the execution resources. The RX 590 GME has 2,304 shading units, 144 texture mapping units, and 32 ROPs. The Pro 575 has 2,048 shading units, 128 TMUs, and the same 32 ROPs. That means the RX 590 GME has 12.5% more shaders and TMUs, which directly feeds its higher compute throughput. The pixel rate confirms this: 45.44 GPixel/s for the RX 590 GME versus 35.07 GPixel/s for the Pro 575. Texture rate is 204.5 GTexel/s versus 140.3 GTexel/s, a 46% difference.

Clock behavior also differs. The RX 590 GME has a base clock of 1257 MHz and a boost clock of 1420 MHz. The Pro 575 has no recorded base or boost clock in the database, only a memory clock of 1695 MHz (6.8 Gbps effective). The RX 590 GME’s memory runs at 2000 MHz (8 Gbps effective), giving it a bandwidth of 256.0 GB/s versus 217.0 GB/s for the Pro 575. Both use 256-bit memory buses and GDDR5 memory, but the RX 590 GME has twice the capacity: 8 GB versus 4 GB.

FP32 compute is 6.543 TFLOPS for the RX 590 GME and 4.489 TFLOPS for the Pro 575, a 45.7% gap. FP16 is identical to FP32 on both, at a 1:1 ratio. The RX 590 GME draws 175 W versus 150 W for the Pro 575, which is the trade-off for the higher clocks and more shaders. The Pro 575 is an MXM module with no power connectors, while the RX 590 GME is a dual-slot card with a single 8-pin connector and a 450 W suggested PSU.

Head-to-Head Benchmarks

The database lists two direct comparisons. The first is Geekbench OpenCL. The RX 590 GME scores 36,294, the Pro 575 scores 34,596. The delta is 4.7% in favor of the RX 590 GME. That is a solid, repeatable lead in a compute API that many rendering and simulation tools use. The gap is not huge, but it is consistent with the RX 590 GME’s higher shader count and clock speed.

The second test is Geekbench Vulkan, and here the separation is far more pronounced. The RX 590 GME scores 60,508, while the Pro 575 manages 37,878. That is a 37.4% difference, the largest margin in any recorded metric between these two cards. Vulkan is a low-overhead API that scales well with raw compute resources, so the RX 590 GME’s 45.7% FP32 advantage translates almost directly into this benchmark result. The Pro 575’s lower clocks and fewer shaders leave it at a severe disadvantage in this workload.

The RX 590 GME wins both head-to-head tests, giving it a 2-0 record in the database. The Pro 575 has zero direct wins. That is a clean sweep, but it is notably the Pro 575’s average benchmark score across all tests is higher, which means the RX 590 GME’s wins are concentrated in these two specific APIs rather than representing a universal superiority.

The Verdict

From the recorded data, the AMD Radeon RX 590 GME is the stronger card for OpenCL and Vulkan workloads. It wins both direct comparisons, with a 4.7% lead in OpenCL and a 37.4% lead in Vulkan. If your application uses Vulkan, the RX 590 GME is the only rational choice between the two. Its higher shader count, clock speeds, and memory bandwidth all point in the same direction.

The AMD Radeon Pro 575, however, holds a higher average benchmark score of 39,555 versus 32,601, and a better percentile ranking at 82 versus 77. That suggests it performs more consistently across a wide range of tests, even if it loses the two specific ones recorded here. The Pro 575 also has a lower power draw at 150 W versus 175 W, which matters in constrained systems, and its MXM form factor is designed for portable or compact professional devices.

Who should pick which? If you are building or upgrading a desktop system with room for a dual-slot card and an 8-pin connector, and your primary workloads are Vulkan-based games or compute, the RX 590 GME is the data-backed choice. If you need a professional card for a mobile or compact workstation, or if your workload mix is broader and less Vulkan-dependent, the Pro 575’s higher average score and better percentile make it the safer bet. The data does not support a universal winner; it supports a workload-specific decision.

FAQ

Q: Which card wins in Vulkan performance?

A: The AMD Radeon RX 590 GME wins decisively, scoring 60,508 against the Pro 575’s 37,878 in Geekbench Vulkan, a 37.4% advantage.

Q: Is the AMD Radeon Pro 575 faster in any recorded benchmark?

A: No. The Pro 575 loses both head-to-head tests: OpenCL (34,596 versus 36,294) and Vulkan (37,878 versus 60,508). It has zero direct wins.

Q: Why does the Pro 575 have a higher average benchmark score if it loses the head-to-head tests?

A: The Pro 575’s average benchmark score is 39,555 across all recorded tests, while the RX 590 GME averages 32,601. The RX 590 GME only has two recorded benchmarks, both of which favor it, but the Pro 575 performs better across the broader database.

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

A: The RX 590 GME has 8 GB of GDDR5 with 256.0 GB/s bandwidth, while the Pro 575 has 4 GB of GDDR5 with 217.0 GB/s bandwidth. Both use a 256-bit bus.

Q: Which card has more shading units?

A: The RX 590 GME has 2,304 shading units, compared to 2,048 on the Pro 575, a 12.5% difference. It also has 144 TMUs versus 128.

Q: Do both cards support the same APIs?

A: Yes, both support DirectX 12 (12_0), OpenGL 4.6, and Vulkan 1.3. The API feature set is identical.

Specification Differences

| Specification | AMD Radeon Pro 575 | AMD Radeon RX 590 GME |

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

| Chip | Ellesmere | Polaris 20 |

| Generation | Radeon Pro Mac (500 Series) | Polaris (RX 500) |

| Base Clock | Not recorded | 1257 MHz |

| Boost Clock | Not recorded | 1420 MHz |

| Memory Clock | 1695 MHz (6.8 Gbps effective) | 2000 MHz (8 Gbps effective) |

| Memory Size | 4 GB | 8 GB |

| Memory Bandwidth | 217.0 GB/s | 256.0 GB/s |

| Shading Units | 2048 | 2304 |

| TMUs | 128 | 144 |

| Pixel Rate | 35.07 GPixel/s | 45.44 GPixel/s |

| Texture Rate | 140.3 GTexel/s | 204.5 GTexel/s |

| FP32 Compute | 4.489 TFLOPS | 6.543 TFLOPS |

| FP16 Compute | 4.489 TFLOPS (1:1) | 6.543 TFLOPS (1:1) |

| TDP | 150 W | 175 W |

| Slot Width | MXM Module | Dual-slot |

| Power Connectors | None | 1x 8-pin |

| Suggested PSU | Not recorded | 450 W |

| Display Outputs | Portable Device Dependent | 1x HDMI 2.0b, 3x DisplayPort 1.4a |

| Dimensions | Not recorded | 241 mm (9.5 inches) length |

| Release Date | 2017-06-04 | 2020-03-08 |

| Predecessor | Not recorded | Arctic Islands |

| Success