GPU Comparison
AMD Radeon Pro 560X
P106-090
PERFORMANCE BENCHMARKS
Analysis: AMD Radeon Pro 560X vs NVIDIA P106-090
FAQ
Q: Which GPU has the higher average benchmark score?
A: The AMD Radeon Pro 560X records an average benchmark score of 15,082, while the NVIDIA P106-090 records 13,470. The AMD part sits in the 57th percentile of all GPUs, compared to the 54th percentile for the NVIDIA part.
Q: How do the two cards compare in Geekbench OpenCL performance?
A: The NVIDIA P106-090 scores 21,304 in Geekbench OpenCL, which is 54.6% higher than the AMD Radeon Pro 560X's 9,663. This is the largest performance gap between the two in any recorded test.
Q: What is the memory configuration difference?
A: The AMD Radeon Pro 560X has 4 GB of GDDR5 on a 128-bit bus with 94.08 GB/s bandwidth. The NVIDIA P106-090 has 3 GB of GDDR5 on a 192-bit bus with 192.2 GB/s bandwidth, offering more than double the memory bandwidth.
Q: Which GPU has a higher pixel fill rate?
A: The NVIDIA P106-090 has a pixel rate of 73.49 GPixel/s, which is substantially higher than the AMD Radeon Pro 560X's 16.06 GPixel/s. This suggests the NVIDIA part is better suited for fill-rate-heavy workloads.
Q: What is the difference in shading unit counts?
A: The AMD Radeon Pro 560X has 1,024 shading units, while the NVIDIA P106-090 has 768. Despite fewer shading units, the NVIDIA part achieves higher FP32 throughput at 2.352 TFLOPS versus 2.056 TFLOPS for the AMD card.
Q: Are both GPUs still in production?
A: No, both are listed as end-of-life products. The AMD Radeon Pro 560X was released in July 2018, while the NVIDIA P106-090 was released in July 2017.
Where Each One Wins
The benchmark data shows a clear split between the two cards. The NVIDIA P106-090 wins both head-to-head tests recorded in the database, making it the stronger performer in raw compute and graphics API workloads. However, the AMD Radeon Pro 560X holds a higher average score across all recorded benchmarks, which reflects a broader set of tests including its strong Geekbench Metal result of 18,763.
For users running OpenCL workloads, the P106-090 is decisively faster, with a 54.6% lead over the Radeon Pro 560X. This makes the NVIDIA card the better choice for compute-heavy tasks that leverage OpenCL. In Vulkan applications, the P106-090 also wins, but by a narrower margin of 9.6%, suggesting that the AMD card remains competitive in this API.
The AMD Radeon Pro 560X wins in aggregate benchmark standing. Its average score of 15,082 places it just 0.1% above the NVIDIA GeForce GTX 660 Ti (15,063) and only 0.6% below the AMD Radeon RX 7600 (15,171). The P106-090's average of 13,470 puts it 0.3% below the NVIDIA GeForce GTX 570 (13,515) and 0.5% above the AMD Radeon Pro 555 (13,407).
For gaming or general-purpose use where display output is required, only the AMD card is viable, since the P106-090 has no display outputs. The P106-090 is explicitly a mining GPU, so its wins are confined to computational roles.
Architecture Differences
The two GPUs come from different architectural generations and foundries. The AMD Radeon Pro 560X uses the Polaris 21 chip built on GCN 4.0 architecture, manufactured on a 14 nm process at GlobalFoundries. The NVIDIA P106-090 uses the GP106 chip based on Pascal architecture, manufactured on a 16 nm process at TSMC.
Transistor counts differ notably. The NVIDIA chip packs 4,400 million transistors on a 200 mm² die, while the AMD chip has 3,000 million transistors on a 123 mm² die. This gives the AMD die a higher transistor density of 24.4M per mm² versus 22.0M per mm² for the NVIDIA die, despite the older process node.
The memory subsystems reflect different design priorities. The AMD card uses a 128-bit bus with 4 GB capacity, while the NVIDIA card uses a 192-bit bus with 3 GB capacity. The NVIDIA card's memory clock is 2002 MHz (8 Gbps effective), compared to 1470 MHz (5.9 Gbps effective) for the AMD card. This results in 192.2 GB/s bandwidth for the P106-090 versus 94.08 GB/s for the Radeon Pro 560X.
FP16 compute differs dramatically. The AMD Radeon Pro 560X achieves 2.056 TFLOPS with a 1:1 FP16 to FP32 ratio. The NVIDIA P106-090 has only 36.74 GFLOPS with a 1:64 ratio, meaning its FP16 performance is negligible by comparison. This makes the AMD card vastly superior for any workload using half-precision arithmetic.
API support also differs. The AMD card supports DirectX 12 (12_0), OpenGL 4.6, and Vulkan 1.3. The NVIDIA card supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4, giving it a slightly newer DirectX feature level and a newer Vulkan version.
Specification Differences
The AMD Radeon Pro 560X has a TDP of 75 W and is an IGP (integrated graphics processor) with no power connectors. The NVIDIA P106-090 also has a 75 W TDP but is a dual-slot card requiring a 1x 6-pin power connector and a suggested PSU of 250 W. The P106-090 is 250 mm (9.8 inches) long.
Bus interface differs significantly. The AMD card uses PCIe 3.0 x8, while the NVIDIA card uses PCIe 1.0 x1. The PCIe 1.0 x1 interface on the P106-090 severely limits data transfer between the GPU and host system, which can bottleneck performance in certain workloads despite the GPU's raw compute advantages.
Display outputs also set them apart. The AMD Radeon Pro 560X's outputs are listed as portable device dependent, meaning it is designed for laptops where the display connection varies by model. The NVIDIA P106-090 has no display outputs at all, confirming its purpose as a dedicated mining card.
Shading unit counts, texture mapping units, and ROPs all differ. The AMD card has 1,024 shading units, 64 TMUs, and 16 ROPs. The NVIDIA card has 768 shading units, 48 TMUs, and 48 ROPs. The higher ROP count on the NVIDIA card contributes to its much higher pixel rate of 73.49 GPixel/s.
Clock speeds show the NVIDIA card running higher. The P106-090 has a base clock of 1354 MHz and a boost clock of 1531 MHz. The Radeon Pro 560X has no base or boost clock listed in the database, only its memory clock. The texture fill rate is also higher on the NVIDIA card at 73.49 GTexel/s versus 64.26 GTexel/s on the AMD card.
Head-to-Head Benchmarks
The database records two direct head-to-head tests between these GPUs. Both use Geekbench but target different APIs.
In Geekbench OpenCL, the NVIDIA P106-090 scores 21,304 against the AMD Radeon Pro 560X's 9,663. This is a 54.6% difference, making it the largest margin in any comparison between the two. The OpenCL result aligns with the P106-090's higher FP32 throughput (2.352 TFLOPS versus 2.056 TFLOPS) and its much higher memory bandwidth (192.2 GB/s versus 94.08 GB/s). The AMD card's GCN architecture does not compensate for these deficits in this test.
In Geekbench Vulkan, the NVIDIA P106-090 scores 18,596 against 16,819 for the AMD Radeon Pro 560X. The 9.6% margin is far smaller than in OpenCL. This suggests that while the NVIDIA card retains the advantage, the AMD architecture narrows the gap when running Vulkan workloads. The AMD card's 1:1 FP16 support may help in some Vulkan compute paths, though the overall result still favors NVIDIA.
The AMD Radeon Pro 560X also has a Geekbench Metal score of 18,763, which is not directly compared here since the P106-090 has no Metal result recorded. This Metal score contributes to the AMD card's higher average benchmark score of 15,082.
For context on the AMD card's position, its average score is 0.1% above the NVIDIA GeForce GTX 660 Ti (15,063) and 0.6% below the AMD Radeon RX 7600 (15,171). The NVIDIA P106-090's average is 0.3% below the NVIDIA GeForce GTX 570 (13,515) and 0.5% above the AMD Radeon Pro 555 (13,407).
The win count is 2 for the NVIDIA P106-090 and 0 for the AMD Radeon Pro 560X in head-to-head tests.
The Verdict
The NVIDIA P106-090 is the better compute performer in the recorded head-to-head benchmarks. It wins both OpenCL and Vulkan tests, with a massive 54.6% lead in OpenCL and a solid 9.6% lead in Vulkan. If the workload is pure computational throughput using these APIs, the P106-090 is the clear choice.
However, the P106-090 has severe practical limitations. It has no display outputs, so it cannot drive a monitor. It uses a PCIe 1.0 x1 bus interface, which will bottleneck data transfers in many real-world scenarios. It is a dual-slot card requiring a 6-pin power connector and a 250 W PSU. Its 3 GB of VRAM is smaller than the AMD card's 4 GB. These factors make it unsuitable for general desktop use.
The AMD Radeon Pro 560X is a mobile IGP with no power connectors and portable-device-dependent outputs. It fits into laptop designs where power efficiency and integration matter. Its 4 GB VRAM and 1:1 FP16 support give it advantages in specific workloads, particularly those using half-precision arithmetic. Its higher average benchmark score (15,082 versus 13,470) and higher percentile ranking (57th versus 54th) indicate better overall standing across a broader benchmark suite.
For a user choosing between these two for a mining or compute-only rig where display output is unnecessary, the P106-090 offers better raw performance in OpenCL and Vulkan. For any scenario requiring a usable display output, a laptop integration, or half-precision compute, the Radeon Pro 560X is the appropriate pick. The data shows that the P106-090 wins on speed, while the Radeon Pro 560X wins on versatility and aggregate benchmark position.