AMD Radeon RX 460 vs NVIDIA RTX PRO 6000 Blackwell Comparison

Head-to-Head Benchmarks

The recorded data shows two graphics cards that are separated by roughly a decade of architectural progress, yet the benchmark comparison is complicated by the fact that each card was tested under different workloads. The AMD Radeon RX 460, a product from the Arctic Islands generation, has an average benchmark score of 18,373 across three OpenCL, Vulkan, and Metal tests. The NVIDIA RTX PRO 6000 Blackwell, by contrast, has an average score of 16,408 from a single 3DMark Steel Nomad DX12 run. On raw average scores, the RX 460 sits 12% higher, but that gap reflects the different test suites rather than a true performance advantage.

Looking at the individual results, the RX 460 delivers 17,065 in Geekbench Metal, 17,855 in Geekbench OpenCL, and 20,198 in Geekbench Vulkan. Its strongest showing is in Vulkan, where it exceeds its own average by roughly 10%. The RTX PRO 6000 Blackwell posts 16,408 in 3DMark Steel Nomad DX12, a demanding modern workload that stresses ray tracing and mesh shading features the RX 460 does not support. The nearest rivals for the RX 460 illustrate how tightly grouped its results are: the Intel Arc A770M scores 18,383, a delta of -0.1%, while the AMD Radeon Pro 5700 sits 1% behind at 18,189, and the NVIDIA GeForce RTX 3060 Mobile trails by 1.2% with 18,159. The FirePro D500 leads the group at 18,533, a 0.9% advantage over the RX 460.

For the RTX PRO 6000 Blackwell, the rival comparisons show a similarly tight cluster. The AMD Radeon PRO W7500 matches it almost exactly with 16,415 and a 0% delta, while the AMD Radeon RX 5700 XT comes in at 16,361, just 0.3% behind. The AMD Radeon Pro 5600M is 0.4% off the pace at 16,351, and the NVIDIA GeForce RTX 5090 D V2 leads by 0.6% with 16,504. These deltas are all within measurement noise, indicating that the RTX PRO 6000 Blackwell's single benchmark result places it in a competitive mid-pack position among cards that are far less expensive and far less powerful on paper.

The most striking observation is that the RTX PRO 6000 Blackwell, despite having 27 times the shading units, 13 times the texture units, 12 times the ROPs, and 48 times the memory capacity, does not dominate the RX 460 in the recorded benchmark averages. This is because the tests are entirely different. The RX 460's scores come from compute-style Geekbench workloads that favor its GCN architecture's raw throughput, while the RTX PRO 6000 Blackwell's score comes from a DX12 gaming benchmark that includes features the RX 460 cannot execute. The data does not provide a direct apples-to-apples comparison, so the head-to-head must be interpreted with that limitation in mind.

Where Each One Wins

The RX 460 wins in compute-oriented synthetic benchmarks. Its Geekbench Vulkan score of 20,198 is its best result, and its OpenCL score of 17,855 and Metal score of 17,065 show consistent performance across multiple API frontends. These scores place it in the 62nd percentile of all GPUs, which is respectable for a low-power card from 2016. The RX 460's nearest rival, the Intel Arc A770M, is only 0.1% ahead, meaning the RX 460 is competitive with a much newer mobile GPU in these specific workloads.

The RTX PRO 6000 Blackwell wins in modern DirectX 12 Ultimate workloads. Its 3DMark Steel Nomad DX12 score of 16,408 comes from a test that leverages DirectX 12 Ultimate features, including ray tracing and mesh shaders, which the RX 460 cannot handle because it only supports DirectX 12 (12_0). The RTX PRO 6000 Blackwell sits in the 59th percentile of all GPUs, slightly below the RX 460's 62nd percentile, but again, these percentiles are derived from different test sets and should not be read as a direct performance ordering.

For users running legacy compute applications that rely on OpenCL or Vulkan compute, the RX 460 presents a workable if dated option. For users running current-generation DX12 games or professional 3D applications that use ray tracing, the RTX PRO 6000 Blackwell is the only one of the two that can execute those workloads at all. The RX 460's 2 GB of GDDR5 memory limits it to light tasks, while the RTX PRO 6000 Blackwell's 96 GB of GDDR7 opens up datasets that would be impossible on the older card.

Architecture Differences

The two cards come from entirely different architectural lineages. The RX 460 uses the Baffin chip built on GCN 4.0, manufactured on a 14 nm process at GlobalFoundries. It packs 3,000 million transistors into a 123 mm² die, giving a transistor density of 24.4 million per square millimeter. The RTX PRO 6000 Blackwell uses the GB202 chip on Blackwell 2.0, fabricated by TSMC on a 5 nm process, with 92,200 million transistors spread across a 750 mm² die, yielding a density of 122.9 million per square millimeter. The density difference is a factor of 5, reflecting the process node advantage of the newer part.

The compute resources are vastly different. The RX 460 has 896 shading units, 56 texture mapping units, and 16 ROPs, with no dedicated ray tracing or tensor cores. The RTX PRO 6000 Blackwell has 24,064 shading units, 752 TMUs, 192 ROPs, 188 ray tracing cores, and 752 tensor cores. The FP32 throughput tells the story: 2.150 TFLOPS for the RX 460 versus 126.0 TFLOPS for the RTX PRO 6000 Blackwell, a 58-fold difference. Both cards have 1:1 FP16 to FP32 ratios, meaning the RX 460 delivers 2.150 TFLOPS FP16 and the RTX PRO 6000 Blackwell delivers 126.0 TFLOPS FP16.

Memory architecture is another major divergence. The RX 460 uses 2 GB of GDDR5 on a 128-bit bus with 112.0 GB/s of bandwidth, running at 1750 MHz (7 Gbps effective). The RTX PRO 6000 Blackwell uses 96 GB of GDDR7 on a 512-bit bus with 1.79 TB/s of bandwidth, also running at 1750 MHz but with 28 Gbps effective. The memory clock is the same base value, but the newer memory type and wider bus produce a 16-fold increase in bandwidth.

Pixel and texture rates follow the same pattern: the RX 460 manages 19.20 GPixel/s and 67.20 GTexel/s, while the RTX PRO 6000 Blackwell reaches 502.5 GPixel/s and 1,968.0 GTexel/s. The RTX PRO 6000 Blackwell also supports DirectX 12 Ultimate (12_2), Vulkan 1.4, and has four DisplayPort 2.1b outputs, whereas the RX 460 is limited to DirectX 12 (12_0), Vulkan 1.3, and a single DVI, one HDMI 2.0b, and one DisplayPort 1.4a. The bus interface also differs: PCIe 3.0 x8 for the RX 460 versus PCIe 5.0 x16 for the RTX PRO 6000 Blackwell.

Power requirements are in different leagues. The RX 460 has a 75 W TDP, requires no power connectors, and a suggested 250 W PSU. The RTX PRO 6000 Blackwell has a 600 W TDP, needs a single 16-pin connector, and a suggested 1000 W PSU. The RX 460 measures 170 mm (6.7 inches) in length, while the RTX PRO 6000 Blackwell is 304 mm (12 inches) long, 137 mm (5.4 inches) tall, and 40 mm (1.6 inches) wide. Both are dual-slot cards, but they occupy very different physical spaces in a chassis.

FAQ

Q: Which card has a higher average benchmark score?

A: The AMD Radeon RX 460 has an average benchmark score of 18,373, while the NVIDIA RTX PRO 6000 Blackwell has an average of 16,408. However, these averages come from different test suites, so they are not directly comparable.

Q: Can the RX 460 run modern DirectX 12 Ultimate games?

A: No. The RX 460 supports DirectX 12 (12_0), while the RTX PRO 6000 Blackwell supports DirectX 12 Ultimate (12_2). The newer API features, including ray tracing and mesh shaders, are only available on the RTX PRO 6000 Blackwell.

Q: What is the memory capacity difference?

A: The RX 460 has 2 GB of GDDR5 on a 128-bit bus with 112.0 GB/s bandwidth. The RTX PRO 6000 Blackwell has 96 GB of GDDR7 on a 512-bit bus with 1.79 TB/s bandwidth.

Q: How do their transistor counts compare?

A: The RX 460 has 3,000 million transistors on a 123 mm² die, while the RTX PRO 6000 Blackwell has 92,200 million transistors on a 750 mm² die.

Q: Which card is more efficient per watt?

A: The RX 460 has a 75 W TDP and delivers 2.150 TFLOPS FP32, while the RTX PRO 6000 Blackwell has a 600 W TDP and delivers 126.0 TFLOPS FP32. The RTX PRO 6000 Blackwell delivers significantly more performance per watt despite its higher total power draw.

Q: What are the nearest rivals for each card?

A: For the RX 460, the closest rival is the Intel Arc A770M with a 0.1% delta, followed by the AMD FirePro D500 at 0.9% and the AMD Radeon Pro 5700 at 1%. For the RTX PRO 6000 Blackwell, the AMD Radeon PRO W7500 matches it exactly, the AMD Radeon RX 5700 XT is 0.3% behind, and the NVIDIA GeForce RTX 5090 D V2 is 0.6% ahead.

The Verdict

The data supports a clear conclusion: these cards serve completely different purposes, and the benchmark numbers should not be used to declare a winner. The RX 460 is an end-of-life product from 2016, built for light gaming and basic compute, with a 62nd percentile standing among all GPUs. The RTX PRO 6000 Blackwell is an active workstation card from 2025, built for professional workloads, with a 59th percentile ranking. The RX 460's higher average score is an artifact of its compute-oriented test suite, not a sign of real superiority.

For users with legacy applications that rely on OpenCL or Vulkan compute and who do not need ray tracing or large memory pools, the RX 460 can still handle basic tasks, but its 2 GB memory and 112.0 GB/s bandwidth will bottleneck any serious workload. For users running modern DX12 applications, professional 3D rendering, or AI inference with tensor cores, the RTX PRO 6000 Blackwell is the only viable choice, offering 126.0 TFLOPS FP32, 96 GB of GDDR7, and 1.79 TB/s of bandwidth. The RX 460 supports Vulkan 1.3 and DirectX 12 (12_0); the RTX PRO 6000 Blackwell supports Vulkan 1.4 and DirectX 12 Ultimate (12_2). The choice is not about which card is faster in a generic sense, but about which workload you need to run.

Specification Differences

| Specification | AMD Radeon RX 460 | NVIDIA RTX PRO 6000 Blackwell |

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

| Chip | Baffin | GB202 |

| Architecture | GCN 4.0 | Blackwell 2.0 |

| Process Node | 14 nm | 5 nm |

| Foundry | GlobalFoundries | TSMC |

| Transistors | 3,000 million | 92,200 million |

| Die Size | 123 mm² | 750 mm² |

| Transistor Density | 24.4M / mm² | 122.9M / mm² |

| Base Clock | 1090 MHz | 1590 MHz |

| Boost Clock | 1200 MHz | 2617 MHz |

| Memory Size | 2 GB | 96 GB |

| Memory Type | GDDR5 | GDDR7 |

| Memory Bus Width | 128 bit | 512 bit |

| Memory Bandwidth | 112.0 GB/s | 1.79 TB/s |

| Shading Units | 896 | 24064 |

| TMUs | 56 | 752 |

| ROPs | 16 | 192 |

| RT Cores | None | 188 |

| Tensor Cores | None | 752 |

| Pixel Rate | 19.20 GPixel/s | 502.5 GPixel/s |

| Texture Rate | 67.20 GTexel/s | 1,968.0 GTexel/s |

| FP32 | 2.150 TFLOPS | 126.0 TFLOPS |

| FP16 | 2.150 TFLOPS (1:1) | 126.0 TFLOPS (1:1) |

| TDP | 75 W | 600 W |

| Power Connectors | None | 1x 16-pin |

| Suggested PSU | 250 W | 1000 W |

| Bus Interface | PCIe 3.0 x8 | PCIe 5.0 x16 |

| Display Outputs | 1x DVI, 1x HDMI 2.0b, 1x DisplayPort 1.4a | 4x DisplayPort 2.1b |

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

| OpenGL | 4.6 | 4.6 |

| Vulkan | 1.3 | 1.4 |

| Length | 170 mm (6.7 inches) | 304 mm (12 inches) |

| Height | Not specified | 137 mm (5.4 inches) |

| Width | Not specified | 40 mm (1.6 inches) |

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

| Release Date | 2016-08-07 | 2025-03-17 |

| Successor | Polaris | None |