NVIDIA GeForce GTX 1070 Ti vs NVIDIA P106-090 Comparison

The NVIDIA GeForce GTX 1070 Ti and the NVIDIA P106-090 are both Pascal-based GPUs, but they target entirely different realities. The 1070 Ti is a fully-featured consumer graphics card, while the P106-090 is a mining-focused board with no display outputs. Benchmark data shows the 1070 Ti is decisively faster in every shared test, making it the clear choice for any gaming or general-purpose workload. The P106-090’s only conceivable role is as a compute-only accelerator in a system where its PCIe 1.0 x1 interface and lack of video outputs are not limiting factors. For anyone building a system that needs to display an image, the choice is the 1070 Ti without question.

The Verdict

The data is unambiguous: the GeForce GTX 1070 Ti wins every benchmark in the head-to-head comparison, with a 100% win rate (3 wins, 0 losses). Its average benchmark score of 14277 places it in the 56th percentile of all GPUs, while the P106-090’s average of 13470 sits in the 54th percentile. In the 3DMark Steel Nomad DX12 test, the 1070 Ti scores 1471 against the P106-090’s 509, a 189% delta. This is not a near-run contest; it is a generational and architectural gap made visible.

The 1070 Ti is the only viable option for gamers, content creators, or anyone using a standard desktop workflow. The P106-090, with no display outputs, cannot drive a monitor. Its 3 GB memory and 192-bit bus are also restrictive, and its PCIe 1.0 x1 interface severely bottlenecks data transfer. The P106-090 is a niche compute card for systems with a separate GPU for display, and even then, its performance is far below the 1070 Ti. The 1070 Ti’s launch MSRP was 399 USD, and it offers a complete feature set in return. The P106-090 has no launch MSRP listed, reflecting its status as a specialized board.

Architecture Differences

Both GPUs use the Pascal architecture on TSMC’s 16 nm process, but they are built on different chips. The 1070 Ti uses the GP104 die, which contains 7,200 million transistors on a 314 mm² die (transistor density of 22.9M / mm²). The P106-090 uses the smaller GP106 chip, with 4,400 million transistors on a 200 mm² die (22.0M / mm²). This silicon difference explains the massive performance gap.

The 1070 Ti has 2432 shading units, 152 texture mapping units (TMUs), and 64 ROPs. The P106-090 is cut down to 768 shading units, 48 TMUs, and 48 ROPs. That is roughly a 68% reduction in shading units and TMUs. Clock speeds also favor the 1070 Ti: its base clock is 1607 MHz with a boost of 1683 MHz, while the P106-090 runs at 1354 MHz base and 1531 MHz boost. Both use GDDR5 memory at 8 Gbps effective, but the 1070 Ti has 8 GB on a 256-bit bus (256.3 GB/s bandwidth) versus the P106-090’s 3 GB on a 192-bit bus (192.2 GB/s).

The 1070 Ti is a full consumer card with display outputs (1x DVI, 1x HDMI 2.0, 3x DisplayPort 1.4a) and a PCIe 3.0 x16 interface. The P106-090 has no display outputs and uses a PCIe 1.0 x1 bus, which is a severe limitation for any data-intensive task. The 1070 Ti also supports Geekbench Metal, a test the P106-090 lacks entirely, highlighting the former’s broader ecosystem compatibility.

Head-to-Head Benchmarks

The head-to-head data covers three tests, and the 1070 Ti dominates all of them. In 3DMark Steel Nomad DX12, the 1070 Ti scores 1471 versus the P106-090’s 509, a 189% lead. This is a DirectX 12 workload that stresses modern gaming features, and the 1070 Ti’s higher shading unit count and memory bandwidth translate directly into a nearly threefold performance advantage.

The Geekbench OpenCL test shows a similar story: the 1070 Ti scores 49219 against 21304 for the P106-090, a 131% delta. OpenCL is a general-purpose compute API, and the 1070 Ti’s FP32 throughput of 8.186 TFLOPS dwarfs the P106-090’s 2.352 TFLOPS. The P106-090’s compute capability is not negligible, but it is in a different performance class.

The largest gap appears in Geekbench Vulkan, where the 1070 Ti scores 61006 and the P106-090 scores 18596, a 228.1% delta. Vulkan is a low-overhead API used in modern games and compute applications. The 1070 Ti’s advantage here is stark, indicating that it is not just faster in raw throughput but also better at handling the API’s parallel workload characteristics. The P106-090’s nearest rival data shows it trading blows with the GeForce GTX 570 (deltaPct -0.3) and AMD Radeon Pro 555 (deltaPct 0.5), while the 1070 Ti’s closest competitors are the Intel Iris Xe MAX Graphics (deltaPct -0.3) and AMD Radeon Vega 11 (deltaPct -0.5). The 1070 Ti is competing in a higher tier entirely.

FAQ

Q: Can the P106-090 be used for gaming?

A: No. The P106-090 has no display outputs, so it cannot connect to a monitor. It is a compute-only card, and its PCIe 1.0 x1 interface would severely hamper any gaming workload that requires frequent data transfers.

Q: How much faster is the GTX 1070 Ti in DirectX 12?

A: In the 3DMark Steel Nomad DX12 test, the 1070 Ti scores 1471 versus the P106-090’s 509, a 189% advantage. This indicates the 1070 Ti delivers nearly three times the performance in this DirectX 12 workload.

Q: What is the memory configuration difference?

A: The 1070 Ti has 8 GB of GDDR5 on a 256-bit bus with 256.3 GB/s bandwidth. The P106-090 has 3 GB of GDDR5 on a 192-bit bus with 192.2 GB/s bandwidth. The 1070 Ti has both more capacity and higher bandwidth.

Q: Which card has a higher clock speed?

A: The 1070 Ti has a base clock of 1607 MHz and a boost clock of 1683 MHz. The P106-090 runs at 1354 MHz base and 1531 MHz boost. The 1070 Ti is faster at both clock steps.

Q: Are these cards in the same performance percentile?

A: No. The 1070 Ti is in the 56th percentile of all GPUs with an average benchmark score of 14277, while the P106-090 is in the 54th percentile with an average score of 13470. The percentile ranks are close, but the actual score gap is significant.

Q: What is the power requirement difference?

A: The 1070 Ti has a TDP of 180 W and requires a 450 W power supply, using a single 8-pin connector. The P106-090 has a TDP of 75 W, requires a 250 W power supply, and uses a single 6-pin connector. The P106-090 is far less power-hungry.

Where Each One Wins

The GTX 1070 Ti wins in every measurable category. It is 189% faster in 3DMark Steel Nomad DX12, 131% faster in Geekbench OpenCL, and 228.1% faster in Geekbench Vulkan. It has higher shading unit counts, more TMUs, more ROPs, higher clocks, more memory, and greater bandwidth. It also has full display outputs and a PCIe 3.0 x16 interface, making it a drop-in solution for any desktop.

The P106-090’s only advantage is efficiency. Its 75 W TDP is less than half the 1070 Ti’s 180 W, and it requires a 250 W power supply versus 450 W. It also uses a single 6-pin connector instead of an 8-pin. In a system where power draw is the absolute priority and display output is handled by another GPU, the P106-090 could be used as a secondary compute device. Its PCIe 1.0 x1 interface, however, makes even that role questionable, as data transfer will be a bottleneck. The 1070 Ti is the superior card for gaming, rendering, or any task that needs a GPU to both compute and display.

Specification Differences

| Field | NVIDIA GeForce GTX 1070 Ti | NVIDIA P106-090 |

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

| Chip | GP104 | GP106 |

| Transistors | 7,200 million | 4,400 million |

| Die Size | 314 mm² | 200 mm² |

| Transistor Density | 22.9M / mm² | 22.0M / mm² |

| Base Clock | 1607 MHz | 1354 MHz |

| Boost Clock | 1683 MHz | 1531 MHz |

| Memory Size | 8 GB | 3 GB |

| Memory Bus Width | 256 bit | 192 bit |

| Memory Bandwidth | 256.3 GB/s | 192.2 GB/s |

| Shading Units | 2432 | 768 |

| TMUs | 152 | 48 |

| ROPs | 64 | 48 |

| Pixel Rate | 107.7 GPixel/s | 73.49 GPixel/s |

| Texture Rate | 255.8 GTexel/s | 73.49 GTexel/s |

| FP32 Performance | 8.186 TFLOPS | 2.352 TFLOPS |

| FP16 Performance | 127.9 GFLOPS (1:64) | 36.74 GFLOPS (1:64) |

| TDP | 180 W | 75 W |

| Power Connectors | 1x 8-pin | 1x 6-pin |

| Suggested PSU | 450 W | 250 W |

| Bus Interface | PCIe 3.0 x16 | PCIe 1.0 x1 |

| Display Outputs | 1x DVI, 1x HDMI 2.0, 3x DisplayPort 1.4a | No outputs |

| Release Date | 2017-11-01 | 2017-07-30 |