AMD Radeon RX 590 vs AMD Radeon RX Vega M GH Comparison

AMD Radeon RX Vega M GH and AMD Radeon RX 590 are two very different implementations of GCN 4.0, and the benchmark data shows a clear split in their strengths. The Vega M GH posts a higher average benchmark score of 29,197, placing it in the 74th percentile of all GPUs, while the RX 590 averages 24,744, sitting in the 70th percentile. That headline gap of roughly 18% in average score is driven by different test workloads, and the architecture data explains why.

Head-to-Head Benchmarks

The recorded benchmarks do not include a direct head-to-head test between these two cards, but the individual benchmark scores and the rival comparisons tell a consistent story. The Vega M GH’s strongest recorded result is a Geekbench Vulkan score of 31,268, which is its highest single-test output. Its Geekbench OpenCL score is 27,125. The RX 590’s highest recorded score is a Geekbench Metal result of 48,407, which is a much larger number than anything the Vega M GH produced, though Metal is a macOS-specific API and not directly comparable to the Vulkan or OpenCL results.

The average benchmark score of 29,197 for the Vega M GH places it nearly level with the AMD FirePro W8000, which averages 29,211, a delta of 0%. It is also 0.1% ahead of the Intel Arc A370M (29,175), 0.7% ahead of the AMD Radeon RX 470 (28,996), and 1.1% ahead of the AMD Radeon RX 6800M (28,874). These are tight margins, indicating the Vega M GH performs in a cluster of mid-range GPUs where small percentage differences separate the field.

The RX 590’s average of 24,744 is 0.1% behind the NVIDIA RTX A5000 Mobile (24,763), 0.4% ahead of the Intel Arc A350M (24,647), 1.2% ahead of the AMD Radeon RX 6600 XT (24,442), and 1.9% ahead of the NVIDIA GeForce GTX 1630 (24,277). This places the RX 590 in a lower performance tier than the Vega M GH, with the average score gap between the two cards being roughly 4,453 points, or about 18% in favor of the Vega M GH.

Looking at the individual test results, the RX 590’s Metal score of 48,407 is substantially higher than the Vega M GH’s Vulkan score of 31,268, but these tests measure different API paths and different driver optimizations. The Vega M GH’s Geekbench OpenCL result of 27,125 is closer to its Vulkan result, suggesting consistent compute performance across API boundaries. The RX 590 does not have a recorded OpenCL or Vulkan score in the database, so cross-API comparisons are limited to what the database provides.

Architecture Differences

The architectural split between these two cards is substantial. Both use GCN 4.0 architecture, but they are built on different process nodes. The Vega M GH uses a 14 nm process from GlobalFoundries, while the RX 590 uses a 12 nm process, also from GlobalFoundries. The smaller process node on the RX 590 allows for more transistors: 5,700 million versus 5,000 million on the Vega M GH. The die sizes also differ, with the RX 590 measuring 232 mm² against the Vega M GH’s 208 mm². Transistor density is nearly identical, at 24.6 million per mm² for the RX 590 and 24.0 million per mm² for the Vega M GH, so the RX 590’s advantage is mostly a matter of having a larger die.

The Vega M GH is built on the Polaris 22 chip, while the RX 590 uses the Polaris 30 chip. The Vega M GH’s generation is listed as “Vega (Vega M)”, and the RX 590’s generation is “Polaris (RX 500)”. The RX 590’s predecessor is listed as “Arctic Islands” and its successor as “Vega”, which places it in a direct lineage. The Vega M GH has no listed predecessor or successor.

Clock speeds favor the RX 590 by a wide margin. The RX 590 runs at a base clock of 1469 MHz and boosts to 1545 MHz, while the Vega M GH runs at 1063 MHz base and 1190 MHz boost. The RX 590 also has faster memory clocks: 2000 MHz with 8 Gbps effective transfer, while the Vega M GH runs memory at 800 MHz with 1600 Mbps effective. These clock differences are partially offset by the memory types, as the Vega M GH uses HBM2 with a 1024-bit bus, while the RX 590 uses GDDR5 with a 256-bit bus.

The memory bandwidth figures reflect this trade-off. The RX 590 achieves 256.0 GB/s bandwidth, which is higher than the Vega M GH’s 204.8 GB/s, but the Vega M GH’s HBM2 memory has a much wider bus. The Vega M GH has 4 GB of memory, while the RX 590 has 8 GB, a significant capacity difference for modern workloads.

The shader and texture configurations also differ. The RX 590 has 2304 shading units, 144 texture mapping units, and 32 raster operation units. The Vega M GH has 1536 shading units, 96 texture mapping units, and 64 ROPs. This means the RX 590 has 50% more shading units and 50% more TMUs, but the Vega M GH has twice the ROP count. The pixel rate reflects this: the Vega M GH achieves 76.16 GPixel/s, while the RX 590 achieves 49.44 GPixel/s. The texture rate flips the result: the RX 590 hits 222.5 GTexel/s, while the Vega M GH reaches 114.2 GTexel/s.

The FP32 compute numbers follow the shading unit counts. The RX 590 delivers 7.119 TFLOPS, while the Vega M GH delivers 3.656 TFLOPS. Both cards have a 1:1 FP16 to FP32 ratio, so the FP16 numbers are identical to their FP32 counts.

Where Each One Wins

The Vega M GH wins in pixel throughput and memory bus width. Its 76.16 GPixel/s pixel rate is roughly 54% higher than the RX 590’s 49.44 GPixel/s, which indicates stronger fill-rate-bound performance in scenarios that stress the ROPs. The 1024-bit HBM2 memory bus, despite the lower clock speed, provides a different memory access pattern than the RX 590’s 256-bit GDDR5 bus, which can benefit certain compute workloads that require high random-access bandwidth.

The RX 590 wins in raw compute throughput, texture processing, and memory capacity. Its 7.119 TFLOPS of FP32 performance is nearly double the Vega M GH’s 3.656 TFLOPS, and its 222.5 GTexel/s texture rate is about 95% higher. The 8 GB memory capacity is double the Vega M GH’s 4 GB, which matters for games and applications that exceed 4 GB of video memory. The RX 590 also has a higher boost clock, 1545 MHz versus 1190 MHz, and a higher base clock, 1469 MHz versus 1063 MHz.

The Vega M GH has a much lower TDP of 100 W, while the RX 590 is rated at 175 W. This makes the Vega M GH significantly more power-efficient on paper, though it is designed as an integrated graphics processor (IGP) with a portable-device-dependent display output, while the RX 590 is a dual-slot card with a PCIe 3.0 x16 interface and a 1x 8-pin power connector. The RX 590 also has a suggested PSU rating of 450 W, while the Vega M GH has none listed.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The AMD Radeon RX Vega M GH has an average benchmark score of 29,197, compared to 24,744 for the AMD Radeon RX 590, a difference of about 4,453 points in favor of the Vega M GH.

Q: How do the two GPUs compare in FP32 compute performance?

A: The RX 590 delivers 7.119 TFLOPS of FP32 compute, which is roughly 95% higher than the Vega M GH’s 3.656 TFLOPS. Both cards have a 1:1 FP16 to FP32 ratio.

Q: What is the memory configuration difference?

A: The Vega M GH uses 4 GB of HBM2 memory on a 1024-bit bus with 204.8 GB/s bandwidth. The RX 590 uses 8 GB of GDDR5 memory on a 256-bit bus with 256.0 GB/s bandwidth.

Q: Which GPU has a higher pixel fill rate?

A: The Vega M GH has a pixel rate of 76.16 GPixel/s, which is about 54% higher than the RX 590’s 49.44 GPixel/s, due to the Vega M GH having 64 ROPs compared to the RX 590’s 32 ROPs.

Q: What are the power consumption figures?

A: The Vega M GH is rated at 100 W TDP, while the RX 590 is rated at 175 W TDP. The RX 590 also has a suggested PSU of 450 W, while the Vega M GH has no PSU suggestion listed.

Q: How do these GPUs rank against their nearest rivals?

A: The Vega M GH is within 1.1% of its nearest rivals, with the AMD FirePro W8000 at 0% delta, the Intel Arc A370M at 0.1%, the AMD Radeon RX 470 at 0.7%, and the AMD Radeon RX 6800M at 1.1%. The RX 590 ranges from 0.1% behind the NVIDIA RTX A5000 Mobile to 1.9% ahead of the NVIDIA GeForce GTX 1630.

The Verdict

The data indicates that the Vega M GH is the better choice for workloads that favor pixel throughput and memory bandwidth efficiency, given its 76.16 GPixel/s pixel rate and 1024-bit HBM2 bus. Its 100 W TDP also makes it the more power-efficient option. However, the RX 590 is the stronger card for compute-heavy tasks, with 7.119 TFLOPS of FP32 performance, a 222.5 GTexel/s texture rate, and 8 GB of memory capacity. The RX 590’s 175 W TDP and dual-slot design indicate it is intended for desktop use, while the Vega M GH is an integrated graphics solution for portable devices.

Users who prioritize raw compute throughput and need more than 4 GB of video memory should choose the RX 590. Users who need a lower-power integrated solution with superior pixel fill rate and a wider memory bus should choose the Vega M GH. The benchmark averages favor the Vega M GH, but the RX 590’s individual strengths in compute and texture processing are clear from the architecture data.

Specification Differences

| Specification | AMD Radeon RX Vega M GH | AMD Radeon RX 590 |

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

| Chip | Polaris 22 | Polaris 30 |

| Process Node | 14 nm | 12 nm |

| Transistors | 5,000 million | 5,700 million |

| Die Size | 208 mm² | 232 mm² |

| Transistor Density | 24.0M / mm² | 24.6M / mm² |

| Base Clock | 1063 MHz | 1469 MHz |

| Boost Clock | 1190 MHz | 1545 MHz |

| Memory Clock | 800 MHz, 1600 Mbps effective | 2000 MHz, 8 Gbps effective |

| Memory Size | 4 GB | 8 GB |

| Memory Type | HBM2 | GDDR5 |

| Memory Bus Width | 1024 bit | 256 bit |

| Memory Bandwidth | 204.8 GB/s | 256.0 GB/s |

| Shading Units | 1536 | 2304 |

| TMUs | 96 | 144 |

| ROPs | 64 | 32 |

| Pixel Rate | 76.16 GPixel/s | 49.44 GPixel/s |

| Texture Rate | 114.2 GTexel/s | 222.5 GTexel/s |

| FP32 | 3.656 TFLOPS | 7.119 TFLOPS |

| FP16 | 3.656 TFLOPS (1:1) | 7.119 TFLOPS (1:1) |

| TDP | 100 W | 175 W |

| Slot Width | IGP | Dual-slot |

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

| Suggested PSU | None | 450 W |

| Bus Interface | IGP | PCIe 3.0 x16 |

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

| Release Date | 2018-01-31 | 2018-11-14 |

| Launch MSRP | None | 279 USD |

| Average Benchmark Score | 29,197 | 24,744 |

| Percentile | 74 | 70 |