NVIDIA GeForce GTX 1080 Ti vs NVIDIA P106-090 Comparison

Where Each One Wins

The benchmark data presents an unusually one-sided comparison. Across all three shared tests, the NVIDIA GeForce GTX 1080 Ti records a win, while the NVIDIA P106-090 does not register a single victory. The use-case split is therefore defined by absolute performance tiers rather than workload-specific strengths.

The GTX 1080 Ti dominates in every measurable category. In 3DMark Steel Nomad DX12, a demanding modern API test, the 1080 Ti scores 2231 against the P106-090's 509, a 338.3% advantage. This is not a marginal gap; it is a generational chasm. For any workload that stresses raw DX12 throughput, the 1080 Ti is the only viable option between the two.

In compute-oriented benchmarks, the story remains consistent. The Geekbench OpenCL score of 67929 for the 1080 Ti dwarfs the 21304 posted by the P106-090, a 218.9% delta. The 1080 Ti also leads in Geekbench Vulkan, scoring 40511 versus 18596, a 117.8% margin. These results indicate that the 1080 Ti is not merely faster in gaming-centric tests but also substantially stronger in general-purpose GPU compute.

The P106-090, by contrast, has no benchmark where it leads. Its role in the database appears to be that of a mining-oriented part, a conclusion supported by its lack of display outputs and its PCIe 1.0 x1 bus interface. The data shows no scenario where a user would prefer the P106-090 on performance grounds. However, its lower power footprint, 75 W versus 250 W, and its smaller die size suggest it was designed for efficiency in a narrow, non-gaming workload. The recorded benchmarks do not include mining-specific tests, so its intended advantage cannot be quantified here, but the absence of any competitive score in the shared tests is decisive.

Architecture Differences

The two GPUs share the Pascal architecture and the 16 nm TSMC process node, but diverge sharply in almost every other physical and functional characteristic.

The GTX 1080 Ti uses the GP102 chip, a large die of 471 mm² housing 11,800 million transistors. The P106-090 uses the GP106 chip, a much smaller 200 mm² die with 4,400 million transistors. Transistor density favors the 1080 Ti slightly, at 25.1M per mm² versus 22.0M per mm², but the absolute difference in transistor count is the dominant factor. The 1080 Ti has 3584 shading units, 224 texture mapping units, and 88 raster output units. The P106-090 has 768 shading units, 48 TMUs, and 48 ROPs. These are not proportional reductions; the 1080 Ti offers more than four times the shading units and nearly twice the ROP count relative to the P106-090's smaller configuration.

Memory architecture further separates the two. The 1080 Ti carries 11 GB of GDDR5X on a 352-bit bus, yielding 484.4 GB/s of bandwidth. The P106-090 has 3 GB of GDDR5 on a 192-bit bus, producing 192.2 GB/s. The 1080 Ti also runs its memory at a higher effective speed, 11 Gbps versus 8 Gbps. Clock speeds tell a similar story: the 1080 Ti has a base clock of 1481 MHz and a boost of 1582 MHz, while the P106-090 operates at 1354 MHz base and 1531 MHz boost. Although the P106-090's clocks are not dramatically lower, the massive difference in execution resources means the 1080 Ti's theoretical peak rates are far higher: 11.34 TFLOPS FP32, 354.4 GTexel/s texture rate, and 139.2 GPixel/s pixel rate, against the P106-090's 2.352 TFLOPS, 73.49 GTexel/s, and 73.49 GPixel/s.

The bus interface is a critical functional difference. The GTX 1080 Ti uses PCIe 3.0 x16, while the P106-090 uses PCIe 1.0 x1. This severely limits the P106-090's ability to transfer data to and from the host system, which is consistent with its mining orientation. The P106-090 also has no display outputs, making it unusable for any visual output task. The 1080 Ti provides 1x HDMI 2.0 and 3x DisplayPort 1.4a outputs. Power requirements differ accordingly: the 1080 Ti draws up to 250 W with a 1x 6-pin plus 1x 8-pin connector setup and a suggested 600 W PSU, while the P106-090 requires only 75 W, a single 6-pin connector, and a 250 W PSU. Both support DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4, but the 1080 Ti's far larger hardware resource pool makes that API support substantially more effective.

Head-to-Head Benchmarks

The shared benchmark suite consists of three tests, and the GTX 1080 Ti wins all of them by margins that range from roughly double to more than quadruple the P106-090's scores.

The largest relative gap appears in 3DMark Steel Nomad DX12. The 1080 Ti scores 2231, while the P106-090 scores 509. This is a 338.3% delta, meaning the 1080 Ti is more than four times faster in this specific DX12 workload. For context, the 1080 Ti's average benchmark score across all recorded tests is 15548, placing it in the 58th percentile of all GPUs. The P106-090's average is 13470, in the 54th percentile. The Steel Nomad result is not an outlier in direction, but it is the most extreme in magnitude.

In Geekbench OpenCL, the 1080 Ti posts 67929 against 21304, a 218.9% advantage. This test measures general compute performance, and the 1080 Ti's 3584 shading units versus 768 give it a natural advantage in parallel workloads. The delta here is slightly more than triple the P106-090's score, consistent with the raw FP32 throughput ratio of 11.34 TFLOPS to 2.352 TFLOPS, which is roughly 4.8 times. The smaller gap in the benchmark relative to the theoretical peak suggests that memory bandwidth or other bottlenecks partially constrain the 1080 Ti, but the outcome is never in doubt.

The closest contest is Geekbench Vulkan, where the 1080 Ti scores 40511 and the P106-090 scores 18596, a 117.8% delta. This is the only shared test where the P106-090 exceeds half of the 1080 Ti's score. Vulkan's lower overhead may allow the smaller GPU to utilize its resources more efficiently, but the absolute result still leaves the 1080 Ti more than twice as fast. The 1080 Ti also has a Geekbench Metal score of 30624 and Passmark scores of 18600 in G3D and 9632 in GPU Compute, none of which have P106-090 counterparts in the database.

The nearest rivals for the 1080 Ti in the database include the AMD Radeon R9 M380 (average score 15521, delta 0.2%), the AMD Radeon Pro W5500 (15679, delta -0.8%), the NVIDIA GeForce RTX 2060 (15290, delta 1.7%), and the NVIDIA GeForce GTX 580 (15283, delta 1.7%). These are all close competitors in average score, indicating that the 1080 Ti sits in a crowded mid-to-high performance band. The P106-090's nearest rivals are the NVIDIA GeForce GTX 570 (13515, delta -0.3%), the AMD Radeon Pro 555 (13407, delta 0.5%), the AMD Radeon HD 7770M (13536, delta -0.5%), and the AMD Radeon RX 9070 XT (13543, delta -0.5%). The P106-090 is essentially equivalent to these parts in average score, but the 1080 Ti is not in the same performance class as any of them.

The Verdict

The data is unambiguous. The NVIDIA GeForce GTX 1080 Ti is the superior GPU in every shared benchmark, by margins that range from 117.8% to 338.3%. Any user who needs a graphics card for gaming, rendering, or general compute should select the 1080 Ti without hesitation. Its 11 GB memory, 484.4 GB/s bandwidth, and 3584 shading units provide a level of performance that the P106-090 cannot approach.

The P106-090 has a narrower purpose. Its lack of display outputs, PCIe 1.0 x1 interface, and 75 W power draw indicate a design for mining operations where visual output and host bandwidth are irrelevant. The database does not include mining benchmarks, so its effectiveness in that specific role cannot be quantified here. For any task that appears in the shared test suite, the P106-090 is not a suitable alternative.

The percentile data reinforces the separation. The 1080 Ti sits at the 58th percentile of all GPUs with an average score of 15548, while the P106-090 is at the 54th percentile with 13470. The difference in average score is 2078 points, roughly 15%, but the head-to-head deltas are far larger because the shared tests are dominated by the 1080 Ti. The P106-090's average score is buoyed by its inclusion in a different benchmark pool, but when forced to compete directly, it falls far behind.

A buyer choosing between these two parts should consider the P106-090 only if the workload is mining-specific and the power budget is a priority. The 250 W power draw of the 1080 Ti versus 75 W for the P106-090 is a real operational difference, and the P106-090's lower PSU requirement of 250 W versus 600 W reduces system cost. But for any general-purpose use, the 1080 Ti is the only rational choice, and the benchmark data provides no counterargument.

FAQ

Q: Which GPU has a higher average benchmark score?

A: The NVIDIA GeForce GTX 1080 Ti has an average benchmark score of 15548, while the NVIDIA P106-090 has an average of 13470.

Q: How large is the performance gap in 3DMark Steel Nomad DX12?

A: The 1080 Ti scores 2231 in this test, and the P106-090 scores 509, a delta of 338.3% in favor of the 1080 Ti.

Q: Does the P106-090 win any shared benchmark?

A: No, the database records 3 wins for the 1080 Ti and 0 wins for the P106-090 across the three common tests.

Q: What is the memory configuration difference between the two cards?

A: The 1080 Ti has 11 GB of GDDR5X on a 352-bit bus with 484.4 GB/s bandwidth. The P106-090 has 3 GB of GDDR5 on a 192-bit bus with 192.2 GB/s bandwidth.

Q: Why might the P106-090 be used despite its lower performance?

A: The P106-090 has a 75 W power draw, a single 6-pin connector, and no display outputs, indicating a design for mining workloads. Its PCIe 1.0 x1 interface further limits general-purpose use.

Q: How do the two GPUs compare in Geekbench Vulkan?

A: The 1080 Ti scores 40511, and the P106-090 scores 18596, a 117.8% advantage for the 1080 Ti. This is the smallest relative gap among the shared tests.