AMD Radeon 660M vs NVIDIA P106-090 Comparison

The AMD Radeon 660M and NVIDIA P106-090 occupy very different corners of the GPU landscape, yet their average benchmark scores land within a few percent of each other. The Radeon 660M is a modern integrated processor from the Rembrandt line, built for portability, while the P106-090 is a dedicated mining card from 2017 with no display outputs. Across the two shared head-to-head tests, the NVIDIA card wins both, but the AMD part holds its own in overall standing. This comparison highlights how architecture, memory, and intended use case matter more than raw average scores alone.

Where Each One Wins

The NVIDIA P106-090 wins every head-to-head benchmark recorded between the two. In Geekbench OpenCL, the P106-090 scores 21,304 against the Radeon 660M’s 12,876, a margin of 39.6 percent. In Geekbench Vulkan, the NVIDIA card posts 18,596 versus 14,748, a 20.7 percent lead. This makes the P106-090 the clear choice for compute-heavy workloads that rely on OpenCL or Vulkan acceleration, especially given its dedicated 3 GB of GDDR5 memory and 192-bit bus.

The AMD Radeon 660M, while losing both direct comparisons, is not without its strengths. Its architecture is significantly newer, and it draws far less power. The 660M has a TDP of 40 W and requires no power connectors, making it suitable for thin-and-light portable devices. The P106-090, by contrast, has a 75 W TDP and needs a 6-pin power connector. For scenarios where power efficiency and system integration matter—such as laptops or compact desktops—the Radeon 660M wins on practical grounds, even if it loses on raw compute scores.

Looking at broader standings, the Radeon 660M sits at the 55th percentile among all GPUs, while the P106-090 sits at the 54th. The average benchmark score for the AMD part is 13,812, slightly ahead of the P106-090’s 13,470. The AMD card’s nearest rivals include the NVIDIA RTX A2000 Mobile (13,821, a 0.1 percent delta) and the AMD Radeon RX 570X (13,871, a 0.4 percent delta). The NVIDIA card’s nearest rivals include the GeForce GTX 570 (13,515, a 0.3 percent delta) and the Radeon Pro 555 (13,407, a 0.5 percent delta). These margins are tiny, suggesting both GPUs perform in a similar mid-range tier despite their very different designs.

Architecture Differences

The architectural gap between these two is vast. The AMD Radeon 660M uses the RDNA 2.0 architecture on a 6 nm process at TSMC, with 13,100 million transistors packed into a 208 mm² die. That works out to a transistor density of 63.0 million per square millimeter. The NVIDIA P106-090 uses the older Pascal architecture on a 16 nm process, also at TSMC, but with only 4,400 million transistors on a 200 mm² die, giving a density of 22.0 million per square millimeter. The Radeon 660M is more than three times denser, reflecting its modern fabrication.

The Radeon 660M is an integrated graphics processor, part of the Rembrandt mobile generation. It has 384 shading units, 24 texture mapping units, and 16 raster output units. It also includes 6 ray tracing cores, a feature the P106-090 lacks entirely. The P106-090, built on the GP106 chip, has 768 shading units, 48 TMUs, and 48 ROPs—double the shading units and triple the ROPs of the AMD part. Neither GPU has tensor cores.

Memory architecture is another major divide. The Radeon 660M uses system-shared memory, with its size, type, and bus width all listed as “System Shared,” and bandwidth described as “System Dependent.” The P106-090 has dedicated 3 GB of GDDR5 memory on a 192-bit bus, with 192.2 GB/s of bandwidth. This dedicated memory gives the NVIDIA card a significant advantage in memory-bound tasks, as it does not compete with the CPU for bandwidth.

Clock speeds tell a different story. The Radeon 660M runs at a base of 1500 MHz and boosts to 1900 MHz. The P106-090 has a lower base of 1354 MHz and a boost of 1531 MHz. Higher clocks help the AMD part close some of the gap in raw throughput, but the NVIDIA card’s larger shader count and dedicated memory prevail in the benchmarks. The Radeon 660M also supports DirectX 12 Ultimate (12_2), while the P106-090 only reaches DirectX 12 (12_1). Both support OpenGL 4.6 and Vulkan 1.4.

Head-to-Head Benchmarks

The most lopsided result comes in Geekbench OpenCL. The P106-090 scores 21,304, which is 39.6 percent higher than the Radeon 660M’s 12,876. This is a substantial lead, likely driven by the NVIDIA card’s 768 shading units and dedicated memory bandwidth. The Radeon 660M’s system-shared memory and lower shader count hurt it in this compute-oriented test.

The Vulkan test is closer but still favors NVIDIA. The P106-090 scores 18,596, beating the Radeon 660M’s 14,748 by 20.7 percent. Vulkan tends to favor efficient driver overhead and raw throughput, both of which benefit the dedicated Pascal card. The Radeon 660M’s newer RDNA 2 architecture and higher boost clock—1900 MHz versus 1531 MHz—help it stay within a more competitive range here, but the shader and memory advantages of the P106-090 remain decisive.

It is worth remembering the Radeon 660M has only two recorded benchmarks—both Geekbench tests—while the P106-090 has three, including a 3DMark Steel Nomad DX12 score of 509. That extra test does not factor into the head-to-head comparison, but it does contribute to the NVIDIA card’s average score of 13,470. The AMD part’s average is 13,812, which is 2.5 percent higher, despite losing both direct matchups. This discrepancy stems from the different test sets each GPU has been subjected to.

FAQ

Q: Which GPU has a higher average benchmark score?

A: The AMD Radeon 660M has an average benchmark score of 13,812, while the NVIDIA P106-090 averages 13,470. The AMD part leads by roughly 2.5 percent in this metric.

Q: How much faster is the NVIDIA P106-090 in OpenCL?

A: The P106-090 scores 21,304 in Geekbench OpenCL, which is 39.6 percent higher than the Radeon 660M’s 12,876.

Q: Does the Radeon 660M support ray tracing?

A: Yes, the Radeon 660M includes 6 ray tracing cores. The NVIDIA P106-090 has no ray tracing cores listed.

Q: What is the memory configuration of each GPU?

A: The Radeon 660M uses system-shared memory with a system-dependent bandwidth. The P106-090 has 3 GB of GDDR5 memory on a 192-bit bus with 192.2 GB/s of bandwidth.

Q: Which GPU has a higher transistor density?

A: The Radeon 660M has a transistor density of 63.0 million per mm², compared to the P106-090’s 22.0 million per mm². The AMD part uses a 6 nm process, while the NVIDIA card uses 16 nm.

Q: Can the P106-090 be used for display output?

A: No, the P106-090 has no display outputs. It was designed for mining, whereas the Radeon 660M’s display outputs are portable-device dependent.

Specification Differences

| Specification | AMD Radeon 660M | NVIDIA P106-090 |

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

| Architecture | RDNA 2.0 | Pascal |

| Process Node | 6 nm | 16 nm |

| Transistors | 13,100 million | 4,400 million |

| Die Size | 208 mm² | 200 mm² |

| Transistor Density | 63.0M / mm² | 22.0M / mm² |

| Base Clock | 1500 MHz | 1354 MHz |

| Boost Clock | 1900 MHz | 1531 MHz |

| Memory Size | System Shared | 3 GB |

| Memory Type | System Shared | GDDR5 |

| Memory Bus Width | System Shared | 192 bit |

| Memory Bandwidth | System Dependent | 192.2 GB/s |

| Shading Units | 384 | 768 |

| TMUs | 24 | 48 |

| ROPs | 16 | 48 |

| Ray Tracing Cores | 6 | None |

| Pixel Rate | 30.40 GPixel/s | 73.49 GPixel/s |

| Texture Rate | 45.60 GTexel/s | 73.49 GTexel/s |

| FP32 Performance | 1,459.2 GFLOPS | 2.352 TFLOPS |

| FP16 Performance | 2.918 TFLOPS (2:1) | 36.74 GFLOPS (1:64) |

| TDP | 40 W | 75 W |

| Slot Width | IGP | Dual-slot |

| Power Connectors | None | 1x 6-pin |

| Suggested PSU | None listed | 250 W |

| Bus Interface | PCIe 4.0 x8 | PCIe 1.0 x1 |

| Display Outputs | Portable Device Dependent | No outputs |

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

| Release Date | 2022-01-03 | 2017-07-30 |

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

The two GPUs differ in nearly every measurable specification. The Radeon 660M is newer, denser, more power-efficient, and supports more advanced APIs, but it relies on shared system memory and has fewer shading units. The P106-090 offers dedicated memory, more shaders, and higher pixel and texture rates, but it is older, less efficient, and lacks ray tracing. Benchmark results indicate the NVIDIA card delivers higher raw compute performance in the tests they share, while the AMD part achieves a slightly higher average score across its own benchmark set.