AMD Radeon PRO V620 vs NVIDIA Tesla P100 PCIe 12 GB Comparison

The Verdict

The benchmark data presents a clear overall winner: the AMD Radeon PRO V620. Its average benchmark score of 136,472 places it in the 96th percentile of all GPUs, while the NVIDIA Tesla P100 PCIe 12 GB sits at 79,396, ranking in the 92nd percentile. The single head-to-head test, Geekbench OpenCL, shows the AMD card delivering a score of 128,580 against the NVIDIA card's 79,396, a decisive 61.9% advantage.

However, the verdict is not simply "buy the faster card." The Tesla P100 is an older, end-of-life product from 2016 with a different design philosophy. Its launch MSRP was 4,599 USD, and its specs reflect a compute-oriented Pascal architecture. The Radeon PRO V620, released in 2021 and also end-of-life, is a more modern RDNA 2.0 part. For any new workload where the Radeon PRO V620 is compatible, the data suggests it is the superior choice due to its massive performance lead. The Tesla P100 should only be considered in legacy deployments where its specific feature set or existing infrastructure requires it; the performance gap recorded here is too large to recommend it for new purchases based on raw compute.

Where Each One Wins

The recorded data shows a single benchmark comparison, and the AMD Radeon PRO V620 wins that test outright. There are no benchmark tests in the database where the NVIDIA Tesla P100 PCIe 12 GB emerges victorious. The Radeon PRO V620's win in Geekbench OpenCL is substantial, and its average score across all benchmarks reinforces this dominance.

For workloads reliant on OpenCL compute, the AMD card is the clear winner. The Tesla P100's only advantage lies in its historical context and a lower power draw. Its 250 W TDP is 50 W less than the AMD card's 300 W, and it requires a single 8-pin power connector instead of two. This makes the Tesla P100 a lower-power, lower-performance option. The Radeon PRO V620, with its higher transistor count (26,800 million vs 15,300 million), newer process node, and higher clock speeds, simply outperforms the older NVIDIA part in every measured category.

Architecture Differences

The two cards are built on fundamentally different architectures from different eras. The AMD Radeon PRO V620 uses the Navi 21 chip based on RDNA 2.0 architecture, fabricated on a 7 nm process at TSMC. The NVIDIA Tesla P100 PCIe 12 GB uses the GP100 chip based on the older Pascal architecture, fabricated on a 16 nm process, also at TSMC. This process node difference is significant: the AMD card has a transistor density of 51.5M per mm², while the NVIDIA card has only 25.1M per mm². Even though the Tesla P100 has a larger die (610 mm² vs 520 mm²), the AMD card packs far more transistors into a smaller space, totaling 26,800 million versus 15,300 million.

The Radeon PRO V620 features 4,608 shading units, 288 texture mapping units, and 128 ROPs. It also includes 72 ray tracing cores, a feature entirely absent from the Tesla P100. Conversely, neither card has tensor cores. The AMD card's memory subsystem uses 32 GB of GDDR6 on a 256-bit bus, while the NVIDIA card uses 12 GB of HBM2 on a much wider 3072-bit bus. This results in similar memory bandwidth: the AMD card achieves 512.0 GB/s, while the NVIDIA card achieves 549.1 GB/s. The Tesla P100's HBM2 memory runs at just 715 MHz (1430 Mbps effective), while the Radeon PRO V620's GDDR6 runs at 2000 MHz (16 Gbps effective). The AMD card also supports PCIe 4.0 x16, while the NVIDIA card is limited to PCIe 3.0 x16. In terms of API support, the AMD card supports DirectX 12 Ultimate (12_2), Vulkan 1.4, and OpenGL 4.6. The NVIDIA card supports DirectX 12 (12_1), Vulkan 1.3, and OpenGL 4.6. Neither card has display outputs, indicating their pure compute-focused design.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The AMD Radeon PRO V620 has a significantly higher average benchmark score of 136,472 compared to the NVIDIA Tesla P100 PCIe 12 GB's 79,396. This places the AMD card in the 96th percentile of all GPUs, versus the NVIDIA card's 92nd percentile.

Q: How much faster is the AMD Radeon PRO V620 in the recorded head-to-head test?

A: In the Geekbench OpenCL test, the AMD Radeon PRO V620 scored 128,580, while the NVIDIA Tesla P100 PCIe 12 GB scored 79,396. This results in a 61.9% performance advantage for the AMD card.

Q: Do these cards support ray tracing?

A: The AMD Radeon PRO V620 includes 72 ray tracing cores, while the NVIDIA Tesla P100 PCIe 12 GB has no ray tracing cores listed in the database.

Q: What are the memory capacities of each card?

A: The AMD Radeon PRO V620 comes with 32 GB of GDDR6 memory on a 256-bit bus. The NVIDIA Tesla P100 PCIe 12 GB comes with 12 GB of HBM2 memory on a 3072-bit bus.

Q: What is the power consumption difference?

A: The AMD Radeon PRO V620 has a TDP of 300 W, while the NVIDIA Tesla P100 PCIe 12 GB has a lower TDP of 250 W. The AMD card requires two 8-pin power connectors, whereas the NVIDIA card requires only one.

Q: Which GPU has better memory bandwidth?

A: The NVIDIA Tesla P100 PCIe 12 GB has a slightly higher memory bandwidth of 549.1 GB/s compared to the AMD Radeon PRO V620's 512.0 GB/s. This is achieved through a much wider 3072-bit bus, despite the NVIDIA card's lower memory clock speed.

Head-to-Head Benchmarks

The database contains one direct comparison between these two GPUs: Geekbench OpenCL. This test is a strong indicator of general compute performance for OpenCL-based workloads.

In this test, the AMD Radeon PRO V620 recorded a score of 128,580. The NVIDIA Tesla P100 PCIe 12 GB recorded a score of 79,396. The delta is a substantial 61.9% in favor of the AMD card. This is not a marginal victory; it is a dominant performance gap. The AMD card's score is roughly 62% higher, meaning it can complete OpenCL compute tasks significantly faster than the NVIDIA card.

This result is consistent with the architectural differences. The AMD card has more shading units (4,608 vs 3,584), more TMUs (288 vs 224), and more ROPs (128 vs 96). Its boost clock of 2200 MHz is dramatically higher than the Tesla P100's 1329 MHz. The AMD card's FP32 throughput of 20.28 TFLOPS is more than double the NVIDIA card's 9.526 TFLOPS. Similarly, its texture rate of 633.6 GTexel/s is over double the NVIDIA card's 297.7 GTexel/s. While the Tesla P100 has a slight edge in memory bandwidth (549.1 GB/s vs 512.0 GB/s), this does little to close the gap in raw compute power. The data clearly shows that in a direct OpenCL compute comparison, the AMD Radeon PRO V620 is the superior performer by a wide margin.

Specification Differences

The following specifications differ between the AMD Radeon PRO V620 and the NVIDIA Tesla P100 PCIe 12 GB, based on the database records.

| Specification | AMD Radeon PRO V620 | NVIDIA Tesla P100 PCIe 12 GB |

| :--- | :--- | :--- |

| Chip | Navi 21 | GP100 |

| Architecture | RDNA 2.0 | Pascal |

| Process Node | 7 nm | 16 nm |

| Transistors | 26,800 million | 15,300 million |

| Die Size | 520 mm² | 610 mm² |

| Transistor Density | 51.5M / mm² | 25.1M / mm² |

| Base Clock | 1825 MHz | 1190 MHz |

| Boost Clock | 2200 MHz | 1329 MHz |

| Memory Clock | 2000 MHz (16 Gbps effective) | 715 MHz (1430 Mbps effective) |

| Memory Size | 32 GB | 12 GB |

| Memory Type | GDDR6 | HBM2 |

| Memory Bus Width | 256 bit | 3072 bit |

| Memory Bandwidth | 512.0 GB/s | 549.1 GB/s |

| Shading Units | 4608 | 3584 |

| TMUs | 288 | 224 |

| ROPs | 128 | 96 |

| Ray Tracing Cores | 72 | None |

| Pixel Rate | 281.6 GPixel/s | 127.6 GPixel/s |

| Texture Rate | 633.6 GTexel/s | 297.7 GTexel/s |

| FP32 Performance | 20.28 TFLOPS | 9.526 TFLOPS |

| FP16 Performance | 40.55 TFLOPS (2:1) | 19.05 TFLOPS (2:1) |

| TDP | 300 W | 250 W |

| Power Connectors | 2x 8-pin | 1x 8-pin |

| Suggested PSU | 700 W | 600 W |

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

| DirectX Support | 12 Ultimate (12_2) | 12 (12_1) |

| Vulkan Support | 1.4 | 1.3 |

| Dimensions (H x W) | 120 mm x 50 mm | Not specified |

| Release Date | 2021-11-03 | 2016-06-19 |

| Predecessor | Radeon Pro Vega | Tesla Maxwell |

| Successor | None | Tesla Volta |

| Launch MSRP | Not specified | 4,599 USD |