AMD Radeon Pro 580 vs NVIDIA GeForce RTX 4090 Mobile Comparison

The NVIDIA GeForce RTX 4090 Mobile and AMD Radeon Pro 580 represent two very different generations of mobile graphics, separated by nearly six years of architectural evolution. The data shows a dominant performance gap, with the RTX 4090 Mobile leading by massive margins in shared benchmarks, yet the Radeon Pro 580 holds its own in its specific legacy niche. This analysis breaks down the architectural chasm, head-to-head results, and specification differences to determine which GPU suits which workload.

FAQ

Q: How much faster is the NVIDIA GeForce RTX 4090 Mobile than the AMD Radeon Pro 580 in OpenCL?

A: The RTX 4090 Mobile scores 180,831 in Geekbench OpenCL, which is 370.2% higher than the Radeon Pro 580’s 38,457. This is the largest performance differential recorded in the head-to-head data.

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA GeForce RTX 4090 Mobile averages 43,667 across all recorded benchmarks, placing it in the 84th percentile of all GPUs. The AMD Radeon Pro 580 averages 40,318, sitting in the 82nd percentile, a difference of roughly 8.3% in aggregate performance.

Q: Does the AMD Radeon Pro 580 support ray tracing or tensor cores?

A: No. The Radeon Pro 580 has null values for both RT cores and tensor cores, indicating it lacks dedicated hardware for these features. The RTX 4090 Mobile includes 76 RT cores and 304 tensor cores, enabling hardware-accelerated ray tracing and AI workloads.

Q: What is the transistor density difference between the two chips?

A: The RTX 4090 Mobile’s AD103 chip, built on a 5 nm process, packs 121.1 million transistors per mm². The Radeon Pro 580’s Ellesmere die, on a 14 nm process, achieves only 24.6 million transistors per mm², a nearly 5x density advantage for NVIDIA.

Q: Which GPU has the higher memory bandwidth?

A: The RTX 4090 Mobile provides 576.0 GB/s of bandwidth from 16 GB of GDDR6 memory on a 256-bit bus. The Radeon Pro 580 offers 217.0 GB/s from 8 GB of GDDR5 memory on the same 256-bit bus width.

Q: Is the Radeon Pro 580 still in production?

A: No. The AMD Radeon Pro 580 is marked as end-of-life, with a release date of 2017-06-04. The RTX 4090 Mobile remains active, released on 2023-01-02.

Architecture Differences

The foundational gap between these GPUs is their fabrication process. The RTX 4090 Mobile uses TSMC’s 5 nm node, while the Radeon Pro 580 relies on GlobalFoundries’ 14 nm process. This translates directly into transistor counts: NVIDIA’s AD103 die contains 45,900 million transistors on a 379 mm² package, versus AMD’s Ellesmere with just 5,700 million transistors on a 232 mm² die. The density difference is stark—121.1M transistors per mm² for NVIDIA against 24.6M for AMD.

Architecturally, the RTX 4090 Mobile is built on Ada Lovelace, the latest GeForce 40-series design, while the Radeon Pro 580 uses GCN 4.0 from the Radeon Pro Mac (500 Series) generation. This is more than a generational label; it represents a shift in compute philosophy. Ada Lovelace integrates dedicated RT cores (76 on this chip) and tensor cores (304), enabling hardware acceleration for ray tracing and deep learning inference. GCN 4.0 has no such dedicated units, with RT and tensor core fields listed as null.

The compute resources diverge sharply. The RTX 4090 Mobile fields 9,728 shading units, 304 TMUs, and 112 ROPs. The Radeon Pro 580 has 2,304 shading units, 144 TMUs, and 32 ROPs—roughly a quarter of the shader count and less than a third of the ROPs. Pixel rate tells the story: 189.8 GPixel/s for NVIDIA versus 38.40 GPixel/s for AMD. Texture rate is 515.3 GTexel/s against 172.8 GTexel/s.

Memory architecture also differs. Both use a 256-bit bus, but the RTX 4090 Mobile pairs 16 GB of GDDR6 with an 18 Gbps effective data rate, yielding 576.0 GB/s. The Radeon Pro 580 has 8 GB of GDDR5 at 6.8 Gbps effective, for 217.0 GB/s. API support favors NVIDIA with DirectX 12 Ultimate (12_2) and Vulkan 1.4, while AMD tops out at DirectX 12 (12_0) and Vulkan 1.3. Both support OpenGL 4.6.

Head-to-Head Benchmarks

The shared benchmark suite between these two GPUs consists of only two tests: Geekbench OpenCL and Geekbench Vulkan. In both, the RTX 4090 Mobile wins decisively, recording 2 wins to 0 for the Radeon Pro 580.

In OpenCL, the RTX 4090 Mobile scores 180,831 against the Radeon Pro 580’s 38,457. That is a delta of 370.2%, meaning NVIDIA delivers nearly 4.7 times the raw compute throughput in this cross-platform API. This aligns with the FP32 peak: 32.98 TFLOPS for NVIDIA versus 5.530 TFLOPS for AMD, a 6x theoretical advantage that real-world OpenCL results partially reflect.

The Vulkan test shows a similarly lopsided result. The RTX 4090 Mobile hits 170,774, while the Radeon Pro 580 manages 43,285. The delta here is 294.5%, or roughly 3.9x the score. Interestingly, the Radeon Pro 580’s Vulkan score of 43,285 is actually higher than its OpenCL score of 38,457, suggesting it performs relatively better in Vulkan’s lower-overhead API. However, the absolute gap remains enormous—NVIDIA’s Vulkan score is 4x larger than AMD’s best result.

Looking at the broader benchmark context, the RTX 4090 Mobile also has data for DirectX 10/11/12 tests, where it scores 173, 262, and 107 respectively in Passmark, plus a G3D score of 27,212 and a compute score of 12,347. The Radeon Pro 580 lacks these DirectX results entirely, only offering a Geekbench Metal score of 39,213, which the RTX 4090 Mobile cannot run (no Metal data). This makes the head-to-head comparison incomplete but the direction unambiguous: in every overlapping test, NVIDIA holds a commanding lead.

Specification Differences

| Specification | NVIDIA GeForce RTX 4090 Mobile | AMD Radeon Pro 580 |

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

| Architecture | Ada Lovelace | GCN 4.0 |

| Process Node | 5 nm | 14 nm |

| Foundry | TSMC | GlobalFoundries |

| Transistors | 45,900 million | 5,700 million |

| Die Size | 379 mm² | 232 mm² |

| Base Clock | 1335 MHz | 1100 MHz |

| Boost Clock | 1695 MHz | 1200 MHz |

| Memory Size | 16 GB | 8 GB |

| Memory Type | GDDR6 | GDDR5 |

| Memory Clock | 2250 MHz / 18 Gbps effective | 1695 MHz / 6.8 Gbps effective |

| Memory Bandwidth | 576.0 GB/s | 217.0 GB/s |

| Shading Units | 9728 | 2304 |

| TMUs | 304 | 144 |

| ROPs | 112 | 32 |

| RT Cores | 76 | null |

| Tensor Cores | 304 | null |

| Pixel Rate | 189.8 GPixel/s | 38.40 GPixel/s |

| Texture Rate | 515.3 GTexel/s | 172.8 GTexel/s |

| FP32 Performance | 32.98 TFLOPS | 5.530 TFLOPS |

| FP16 Performance | 32.98 TFLOPS (1:1) | 5.530 TFLOPS (1:1) |

| TDP | 120 W | 185 W |

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

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

| Vulkan Support | 1.4 | 1.3 |

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

| Release Date | 2023-01-02 | 2017-06-04 |

The TDP figures are counterintuitive: the far more powerful RTX 4090 Mobile draws 120 W, while the older Radeon Pro 580 requires 185 W. This efficiency leap is a direct consequence of the 5 nm node versus 14 nm. Both are listed as IGP with no power connectors, indicating they are soldered mobile parts. The bus interface also differs, with NVIDIA using PCIe 4.0 x16 against AMD’s PCIe 3.0 x16.

The Verdict

The data makes one thing clear: the NVIDIA GeForce RTX 4090 Mobile is in a different performance class entirely. Its 370.2% OpenCL lead and 294.5% Vulkan lead over the Radeon Pro 580 are not incremental gains—they represent a generational leap that redefines what is possible from a mobile GPU. For any workload that relies on FP32 compute, ray tracing, or tensor operations, the RTX 4090 Mobile is the only rational choice. Its 32.98 TFLOPS of FP32 performance, 76 RT cores, and 304 tensor cores provide capabilities the Radeon Pro 580 simply cannot offer, as evidenced by the latter’s null values for those dedicated units.

The Radeon Pro 580, however, is not without context. Its average benchmark score of 40,318 places it in the 82nd percentile of all GPUs, a respectable showing for a 2017 part. It even edges out the RTX 4090 Mobile in percentile rank among its nearest rivals: the Radeon Pro 580 sits just 0.6% ahead of the Radeon Pro WX 7100 and 1.9% ahead of the RTX A500 Mobile. But its nearest rival list includes the GeForce RTX 5070 at nearly identical average scores, suggesting modern mid-range GPUs have caught up to this legacy part.

For buyers, the verdict depends on workload. If the task demands maximum compute, modern API support, or ray tracing, the RTX 4090 Mobile wins outright—there is no metric in the data where the Radeon Pro 580 comes close. The RTX 4090 Mobile’s 84th percentile ranking and 43,667 average score speak to its position near the top of the mobile GPU stack. For legacy Mac-based systems or applications optimized for GCN architecture, the Radeon Pro 580 remains a functional, end-of-life option, but its 5.530 TFLOPS FP32 ceiling and lack of hardware ray tracing make it a poor fit for any forward-looking workload. The data is unambiguous: choose NVIDIA for performance, choose AMD only if locked into an older ecosystem.