AMD Radeon Pro Vega II vs NVIDIA Quadro GP100 Comparison

AMD Radeon Pro Vega II and NVIDIA Quadro GP100 are both end-of-life professional workstation GPUs, but they represent very different generations of GPU architecture. The data shows a single common benchmark result, with the AMD card holding a decisive edge in raw compute performance according to that metric.

Head-to-Head Benchmarks

The only directly comparable benchmark in the data is Geekbench OpenCL. Here, the AMD Radeon Pro Vega II scores 99,048 points, while the NVIDIA Quadro GP100 scores 87,445 points. This represents a 13.3% victory for the AMD card, a substantial margin within professional compute benchmarks. The AMD card's raw FP32 throughput of 14.09 TFLOPS versus the NVIDIA card's 10.34 TFLOPS explains much of this gap, as OpenCL workloads tend to scale well with raw shader compute. The AMD card also holds an advantage in texture rate, with 440.3 GTexel/s compared to the NVIDIA card's 323.2 GTexel/s.

However, the NVIDIA Quadro GP100 fights back in a different metric. Its pixel rate is 138.5 GPixel/s, which is 25.8% higher than the AMD card's 110.1 GPixel/s, despite the AMD card having a higher boost clock of 1720 MHz versus 1443 MHz. This is because the NVIDIA card has 96 ROPs compared to just 64 ROPs on the AMD card. The benchmark results indicate that for pure rasterization throughput, the NVIDIA card is the stronger part, even if the single OpenCL test does not capture this advantage.

The average benchmark scores reinforce the AMD card's overall lead. The Radeon Pro Vega II has an average score of 109,617 across its three Geekbench tests (Metal, OpenCL, and Vulkan), while the Quadro GP100 has only one recorded benchmark with an average of 87,445. The AMD card also sits in the 94th percentile of all GPUs, just one point higher than the NVIDIA card's 93rd percentile. Looking at the nearest rivals for each card, the AMD card is 1% behind the AMD Radeon PRO W7900 (110,725 average score) and 2.1% ahead of the AMD Radeon Pro W6600X (107,342). The NVIDIA card, by contrast, is 0.4% ahead of the AMD Radeon PRO W7600 (87,108) but 4.6% behind the NVIDIA RTX A4500 (91,671).

Where Each One Wins

The AMD Radeon Pro Vega II is clearly the compute-oriented card. Its 13.3% OpenCL victory is supported by a 36.3% advantage in FP32 throughput (14.09 TFLOPS vs 10.34 TFLOPS) and a 36.2% advantage in FP16 throughput (28.18 TFLOPS vs 20.69 TFLOPS). The AMD card also has twice the memory capacity at 32 GB versus 16 GB, and a higher memory bandwidth of 825.3 GB/s versus 732.2 GB/s. For workloads that are memory-capacity bound, such as large dataset processing or high-resolution rendering, the AMD card offers significantly more headroom. The AMD card's 7 nm process node also gives it a transistor density of 40.0M per mm², compared to 25.1M per mm² on the NVIDIA card, indicating a more modern and efficient design.

The NVIDIA Quadro GP100 wins in rasterization-oriented tasks. Its 96 ROPs and 138.5 GPixel/s fill rate make it better suited for pixel-heavy workloads like traditional 3D rendering in viewports, where pixel throughput is the bottleneck. The NVIDIA card also has a more practical physical profile, being a dual-slot card with a 267 mm length and 111 mm height, whereas the AMD card is a quad-slot design. The NVIDIA card uses a standard PCIe 3.0 x16 interface and a single 8-pin power connector, making it far easier to integrate into existing workstations. The AMD card uses the Apple MPX bus interface, which restricts it primarily to Apple Mac Pro systems, and its 475 W TDP is more than double the NVIDIA card's 235 W TDP.

Architecture Differences

The architectural chasm between these two cards is significant. The AMD Radeon Pro Vega II is built on the GCN 5.1 architecture using the Vega 20 chip, while the NVIDIA Quadro GP100 uses the Pascal architecture with the GP100 chip. The process nodes are generations apart: AMD uses a 7 nm TSMC process, while NVIDIA uses a 16 nm TSMC process. This explains why the AMD chip, despite having fewer transistors (13,230 million vs 15,300 million), achieves higher clock speeds — 1574 MHz base and 1720 MHz boost versus 1304 MHz base and 1443 MHz boost on the NVIDIA card.

The compute unit configurations differ substantially. The AMD card has 4,096 shading units, 256 TMUs, and 64 ROPs. The NVIDIA card has 3,584 shading units, 224 TMUs, and 96 ROPs. This means AMD allocates more die area to shader and texture processing, while NVIDIA allocates more to pixel processing. Both cards use HBM2 memory with a 4096-bit memory bus, but the AMD card runs its memory at 806 MHz (1612 Mbps effective) versus the NVIDIA card's 715 MHz (1430 Mbps effective), resulting in the AMD card's higher bandwidth of 825.3 GB/s versus 732.2 GB/s.

Neither card has dedicated ray tracing cores or tensor cores. Both support DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.3. The AMD card was released on 2019-06-02, while the NVIDIA card was released earlier on 2016-09-30. The NVIDIA card lists its predecessor as Quadro Maxwell and its successor as Quadro Volta, while the AMD card has no listed predecessor or successor. The display outputs also differ: the AMD card offers 1x HDMI 2.0b and 4x Thunderbolt, while the NVIDIA card offers 1x DVI and 4x DisplayPort 1.4a.

FAQ

Q: Which card has higher raw compute performance?

A: The AMD Radeon Pro Vega II wins decisively. It scores 99,048 in Geekbench OpenCL versus 87,445 for the NVIDIA Quadro GP100, a 13.3% advantage. This is backed by higher FP32 throughput of 14.09 TFLOPS versus 10.34 TFLOPS.

Q: Does the NVIDIA card have any performance advantage?

A: Yes, in pixel throughput. The Quadro GP100 has a pixel rate of 138.5 GPixel/s, which is 25.8% higher than the AMD card's 110.1 GPixel/s, due to its larger count of 96 ROPs versus 64 ROPs.

Q: How do their memory configurations compare?

A: The AMD card offers 32 GB of HBM2 memory with 825.3 GB/s bandwidth. The NVIDIA card offers 16 GB of HBM2 memory with 732.2 GB/s bandwidth. Both use a 4096-bit memory bus.

Q: What are the power requirements for each card?

A: The AMD Radeon Pro Vega II has a TDP of 475 W and requires an 850 W power supply. The NVIDIA Quadro GP100 has a TDP of 235 W and requires a 550 W power supply.

Q: Are these cards compatible with standard PC motherboards?

A: The NVIDIA Quadro GP100 uses PCIe 3.0 x16 and a standard 8-pin power connector, making it broadly compatible. The AMD Radeon Pro Vega II uses the Apple MPX bus interface, limiting it to Apple Mac Pro systems.

Q: How do these cards rank against all other GPUs?

A: The AMD card is in the 94th percentile of all GPUs with an average benchmark score of 109,617. The NVIDIA card is in the 93rd percentile with an average score of 87,445.

The Verdict

The data points to a clear compute performance winner: the AMD Radeon Pro Vega II. Its 13.3% OpenCL lead, double memory capacity, and higher bandwidth make it the superior choice for compute-heavy workloads like scientific simulation, machine learning training, or large-scale data processing. The AMD card also holds a massive advantage in FP32 and FP16 throughput, making it the better option for general-purpose GPU computing.

The NVIDIA Quadro GP100 is the choice for users who prioritize rasterization performance and system integration flexibility. Its higher pixel rate and lower power draw (235 W vs 475 W) make it more suitable for traditional 3D viewport rendering and for workstations where power and cooling are constraints. The NVIDIA card's standard PCIe interface and dual-slot form factor also mean it can be installed in a wider range of systems, whereas the AMD card's Apple MPX interface restricts it to Mac Pro environments.

For users who need maximum compute capability and are working within the Apple ecosystem, the AMD Radeon Pro Vega II is the data-backed choice. For users who need broad hardware compatibility and pixel throughput, and who are willing to accept lower compute scores, the NVIDIA Quadro GP100 is the more practical option. The benchmark results show a 13.3% compute gap between them, but the architectural differences — particularly in ROP count and memory capacity — mean the right choice depends entirely on the specific workload profile.

Specification Differences

| Specification | AMD Radeon Pro Vega II | NVIDIA Quadro GP100 |

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

| Architecture | GCN 5.1 | Pascal |

| Process Node | 7 nm | 16 nm |

| Transistors | 13,230 million | 15,300 million |

| Die Size | 331 mm² | 610 mm² |

| Transistor Density | 40.0M / mm² | 25.1M / mm² |

| Base Clock | 1574 MHz | 1304 MHz |

| Boost Clock | 1720 MHz | 1443 MHz |

| Memory Size | 32 GB | 16 GB |

| Memory Clock | 806 MHz (1612 Mbps effective) | 715 MHz (1430 Mbps effective) |

| Memory Bandwidth | 825.3 GB/s | 732.2 GB/s |

| Shading Units | 4,096 | 3,584 |

| TMUs | 256 | 224 |

| ROPs | 64 | 96 |

| Pixel Rate | 110.1 GPixel/s | 138.5 GPixel/s |

| Texture Rate | 440.3 GTexel/s | 323.2 GTexel/s |

| FP32 | 14.09 TFLOPS | 10.34 TFLOPS |

| FP16 | 28.18 TFLOPS (2:1) | 20.69 TFLOPS (2:1) |

| TDP | 475 W | 235 W |

| Slot Width | Quad-slot | Dual-slot |

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

| Suggested PSU | 850 W | 550 W |

| Bus Interface | Apple MPX | PCIe 3.0 x16 |

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

| Release Date | 2019-06-02 | 2016-09-30 |

| Launch MSRP | 2,199 USD | None listed |