AMD Radeon RX 6800S vs NVIDIA P106-100 Comparison

The AMD Radeon RX 6800S and NVIDIA P106-100 represent two distinct eras of GPU design, separated by nearly five years of architectural evolution. The RX 6800S is a modern mobile RDNA 2.0 part, while the P106-100 is a Pascal-based mining card stripped of display outputs. Benchmark data reveals a decisive performance gap, with the RX 6800S winning all three shared head-to-head tests by substantial margins, yet the P106-100 holds its own in the overall percentile rankings, suggesting the comparison is more nuanced than raw speed alone.

Head-to-Head Benchmarks

The most lopsided result comes from the Geekbench Vulkan test, where the RX 6800S scores 81,272 against the P106-100's 32,897. That is a 147% delta — more than double the performance. This gap likely reflects the architectural chasm between RDNA 2.0's modern async compute and Pascal's older design, though the data does not specify the exact cause. The RX 6800S also dominates Geekbench OpenCL, posting 73,202 versus 35,951, a 103.6% advantage. This compute-oriented benchmark shows the RX 6800S delivering roughly twice the raw throughput, which aligns with its higher FP32 rating of 8.602 TFLOPS compared to the P106-100's 4.375 TFLOPS.

In the 3DMark Steel Nomad DX12 test, the RX 6800S scores 1,689 against 899 for the P106-100, an 87.9% lead. While this is the smallest relative margin of the three tests, it still represents a near-doubling of performance. The P106-100's DirectX 12 (12_1) API support, lacking the 12_2 Ultimate features found on the RX 6800S, may explain part of this deficit in a modern DX12 workload. Notably, the RX 6800S also has a PassMark G3D score of 15,908 and a 3DMark Steel Nomad score of 1,689, while the P106-100 does not have comparable entries in those specific test categories, limiting direct cross-referencing.

Despite these decisive wins, the average benchmark scores tell a different story. The RX 6800S averages 24,063 across its benchmark suite, while the P106-100 averages 23,249. That is a mere 3.5% difference. The RX 6800S sits at the 69th percentile of all GPUs, while the P106-100 is at the 68th. This near-parity in overall standing suggests the P106-100's few available benchmark scores are disproportionately strong in certain legacy or compute workloads, even if the modern DX12 and Vulkan tests favor the AMD part overwhelmingly.

Where Each One Wins

The RX 6800S wins every shared benchmark outright, so the question becomes where the P106-100 might still find relevance. The data shows the P106-100's nearest rival is the AMD Radeon Pro Vega 16 with a 0% delta, meaning it performs exactly on par with that workstation card. It also trails the RX 6600M by only 0.1% and the R9 M290X by 0.1%. This clustering suggests the P106-100 occupies a specific performance tier that, while far below the RX 6800S in modern tests, is competitive with older mobile gaming GPUs.

For the RX 6800S, its nearest rival is the NVIDIA GeForce RTX 2080 SUPER, which scores 24,170 against the AMD card's 24,063 — a 0.4% deficit. The RX 6800S also sits 0.7% behind the GTX 780 Ti and 0.9% behind the GTX 1630, yet it leads the AMD Radeon RX 9070 by 0.8%. This positions the RX 6800S in a strange middle ground: it matches desktop GPUs from prior generations despite being an integrated mobile part with a 100 W TDP. The P106-100, with its 120 W TDP and dual-slot design, cannot claim similar parity with high-end desktop parts.

The use-case split is stark. The RX 6800S is built for gaming and modern API workloads, evidenced by its 12 Ultimate (12_2) DirectX support, 32 ray-tracing cores, and Vulkan 1.4 compatibility. The P106-100, with no display outputs and a PCIe 1.0 x16 interface, is clearly designed for compute or mining tasks where video output is irrelevant. Its Geekbench OpenCL score of 35,951, while half the RX 6800S's result, is still respectable for a card that lacks any RT or tensor core acceleration.

Architecture Differences

The two GPUs are built on fundamentally different process nodes and architectures. The RX 6800S uses TSMC's 7 nm process with 11,060 million transistors on a 237 mm² die, yielding a transistor density of 46.7 million per mm². The P106-100 uses TSMC's 16 nm process with 4,400 million transistors on a 200 mm² die, giving a density of just 22.0 million per mm². This density advantage is a key factor in the RX 6800S's higher performance per watt, though the TDP figures (100 W for AMD, 120 W for NVIDIA) show the P106-100 actually consumes more power.

Architecturally, the RX 6800S employs RDNA 2.0 with 2,048 shading units, 128 texture mapping units, and 64 ROPs. It also includes 32 dedicated ray-tracing cores, a feature entirely absent from the P106-100. The NVIDIA card uses Pascal with 1,280 shading units, 80 TMUs, and 48 ROPs, and has no RT or tensor cores. The FP16 compute rates are particularly telling: the RX 6800S achieves 17.20 TFLOPS with a 2:1 ratio, while the P106-100 manages only 68.36 GFLOPS with a 1:64 ratio. This means the AMD card is over 250 times faster at half-precision math, a metric that matters for AI and compute workloads.

Memory subsystems also diverge sharply. The RX 6800S has 8 GB of GDDR6 on a 128-bit bus with 256.0 GB/s bandwidth. The P106-100 has 6 GB of GDDR5 on a 192-bit bus with 192.2 GB/s bandwidth. Despite the narrower bus, the AMD card's newer GDDR6 memory runs at 16 Gbps effective versus 8 Gbps for the NVIDIA card, yielding 33% more bandwidth. The RX 6800S also supports PCIe 4.0 x8, while the P106-100 is limited to PCIe 1.0 x16 — a potentially significant bottleneck for the older card in data-transfer-heavy workloads.

FAQ

Q: Which GPU is faster in DirectX 12 Ultimate workloads?

A: The RX 6800S scores 1,689 in 3DMark Steel Nomad DX12, an 87.9% lead over the P106-100's 899. The AMD card also supports DirectX 12 Ultimate (12_2), while the NVIDIA card only reaches DirectX 12 (12_1).

Q: Does the P106-100 have any advantage in compute tasks?

A: No shared benchmark shows the P106-100 ahead. The RX 6800S leads Geekbench OpenCL by 103.6% and Vulkan by 147%. However, the P106-100's average benchmark score of 23,249 places it at the 68th percentile, only one point below the RX 6800S's 69th percentile, indicating it competes well in unspecified legacy tests.

Q: Why does the P106-100 have no display outputs?

A: The P106-100 is categorized under "Mining GPUs" and lists "No outputs" for display connections. It uses a PCIe 1.0 x16 interface and a 1x 6-pin power connector, suggesting it was designed for compute or mining rigs rather than traditional desktop use.

Q: How do their memory configurations compare?

A: The RX 6800S has 8 GB of GDDR6 on a 128-bit bus delivering 256.0 GB/s, while the P106-100 has 6 GB of GDDR5 on a 192-bit bus delivering 192.2 GB/s. The AMD card's memory runs at 16 Gbps effective versus 8 Gbps for the NVIDIA card.

Q: Which GPU has better modern API support?

A: The RX 6800S supports DirectX 12 Ultimate (12_2), Vulkan 1.4, and includes 32 RT cores. The P106-100 supports DirectX 12 (12_1) and Vulkan 1.4 but has no RT cores. Both support OpenGL 4.6.

Q: Are these GPUs still in production?

A: Both are listed as "End-of-life." The RX 6800S was released on 2022-01-03, while the P106-100 was released on 2017-06-18.

Specification Differences

| Specification | AMD Radeon RX 6800S | NVIDIA P106-100 |

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

| Architecture | RDNA 2.0 | Pascal |

| Process Node | 7 nm | 16 nm |

| Transistors | 11,060 million | 4,400 million |

| Die Size | 237 mm² | 200 mm² |

| Transistor Density | 46.7M / mm² | 22.0M / mm² |

| Base Clock | 1800 MHz | 1506 MHz |

| Boost Clock | 2100 MHz | 1709 MHz |

| Memory Size | 8 GB GDDR6 | 6 GB GDDR5 |

| Memory Bus Width | 128 bit | 192 bit |

| Memory Bandwidth | 256.0 GB/s | 192.2 GB/s |

| Shading Units | 2048 | 1280 |

| TMUs | 128 | 80 |

| ROPs | 64 | 48 |

| RT Cores | 32 | None |

| FP32 Performance | 8.602 TFLOPS | 4.375 TFLOPS |

| FP16 Performance | 17.20 TFLOPS (2:1) | 68.36 GFLOPS (1:64) |

| TDP | 100 W | 120 W |

| Slot Width | IGP | Dual-slot |

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

| Bus Interface | PCIe 4.0 x8 | PCIe 1.0 x16 |

| Display Outputs | Portable Device Dependent | No outputs |

| Release Date | 2022-01-03 | 2017-06-18 |

The Verdict

The data makes one conclusion unavoidable: the RX 6800S is categorically faster than the P106-100 in every measurable shared benchmark. Its 147% lead in Vulkan, 103.6% lead in OpenCL, and 87.9% lead in 3DMark Steel Nomad are not marginal victories — they are generational gaps. The RX 6800S also brings modern features like ray tracing, DirectX 12 Ultimate, and PCIe 4.0 that the P106-100 simply cannot offer.

For a user choosing between these two, the RX 6800S is the only sensible option for gaming or any modern graphics workload. Its 8 GB of GDDR6 memory and 256.0 GB/s bandwidth, combined with 32 RT cores, make it suitable for current titles and future-proofing. The P106-100, with no display outputs and a PCIe 1.0 interface, is functionally limited to headless compute tasks. Its 6 GB of GDDR5 and 192.2 GB/s bandwidth are adequate for older workloads, and its 68th percentile ranking shows it still competes with GPUs like the RX 6600M and R9 M290X.

However, the P106-100's near-parity in average benchmark score — 23,249 versus 24,063 — hints that it may excel in specific legacy or compute benchmarks not shared in this comparison. For a mining rig or a dedicated compute server where video output is unnecessary, the P106-100's 120 W TDP and dual-slot design might still be serviceable. But for any interactive use, the RX 6800S wins by every metric in the data, making the choice clear. The verdict is not close: the RX 6800S is the superior GPU, and the P106-100's only remaining value lies in narrowly scoped, output-free compute roles.