AMD Radeon RX 6750 XT vs NVIDIA P106-100 Comparison

# AMD Radeon RX 6750 XT vs NVIDIA P106-100: Head-to-Head Benchmark Analysis

The AMD Radeon RX 6750 XT (Navi 22, RDNA 2.0) and the NVIDIA P106-100 (GP106, Pascal) occupy very different positions in the database. The RX 6750 XT is a full-scale gaming and rendering card from the Radeon RX 6000 series, while the P106-100 is a mining-oriented product stripped of display outputs. The recorded benchmarks show a clear split: the AMD card wins two of the three head-to-head tests, and the NVIDIA card wins one. The RX 6750 XT takes the 3DMark Steel Nomad DX12 test with a score of 2548 against 899 for the P106-100, a delta of 183.4% in favor of AMD. The NVIDIA card hits back in Geekbench OpenCL, scoring 35951 versus 16360, a gap of 54.5% in its favor. The final head-to-head, Geekbench Vulkan, goes to AMD again: 98089 against 32897, a 198.2% advantage. Overall, the database records two wins for the RX 6750 XT and one win for the P106-100.

The average benchmark score across all recorded tests places the RX 6750 XT at 26011, against 23249 for the P106-100, a difference of roughly 11.9% in favor of AMD. In percentile terms, the RX 6750 XT sits at the 71st percentile of all GPUs in the database, while the P106-100 sits at the 68th percentile. These are close, mid-range positions, not top-tier scores, but the split in workloads is what really separates these two.

Head-to-Head Benchmarks

The data contains three head-to-head benchmark comparisons. The first is 3DMark Steel Nomad DX12, and the RX 6750 XT delivers 2548 points. The P106-100 manages 899. That is a delta of 183.4%, a decisive result for the AMD card. In absolute terms, the RX 6750 XT produces more than two and a half times the frames of the P106-100 in this test. The RX 6750 XT is the clear winner here, and this is the kind of workload the AMD card is built for: a modern 3D API, high geometry throughput, and the full 2560 shading units of the Navi 22 chip.

The second test is Geekbench OpenCL. Here the P106-100 wins, reversing the first result. The NVIDIA card scores 35951 in OpenCL, the RX 6750 XT scores 16360. That is a delta of 54.5% in favor of the P106-100. The P106-100 is a Pascal card with 1280 shading units and a very high memory clock, and OpenCL compute is a classic NVIDIA strength. The RX 6750 XT has a much lower FP32 throughput in this database recording, and the result aligns with that.

The third benchmark, Geekbench Vulkan, goes back to AMD. The RX 6750 XT records 98089, against 32897 for the P106-100, a delta of 198.2%. Vulkan is a modern low-level API, and the RDNA 2 architecture with its 40 ray tracing cores and 2560 shading units handles this type of workload with clear superiority. The P106-100 is based on the older Pascal design, which has no dedicated ray tracing or tensor cores, and its 1280 shading units are arranged in a configuration that pushes less geometry per cycle.

So the head-to-head results in two wins for the RX 6750 XT, one win for the P106-100. The wins for the AMD card are in modern 3D and a modern cross-API compute/geometry workload. The win for NVIDIA is in a compute-level OpenCL test. That is a typical split: the AMD card is the more balanced, more feature-rich product, while the NVIDIA P106-100 is a specialized mining part with a narrow driver feature set (no display outputs, no ray tracing, etc.) that still does well in raw compute.

Where Each One Wins

3D gaming and modern APIs: The RX 6750 XT is the card to pick. The 3DMark Steel Nomad DX12 result (2548 against 8996) is a 183.4% lead. The card has 40 ray tracing cores, 160 texture mapping units, and 64 render output processors, and the Navi 22 chip is built to execute the full DX12 Ultimate feature set, including DirectX Raytracing. The P106-100 has no ray tracing cores and 48 ROPs and 80 TMUs. In a modern API load, the AMD card simply has more of the relevant silicon to throw at the scene.

Compute and OpenCL: The P106-100 is the pick. The Geekbench OpenCL score of 35951 is 54.5% higher than the 16360 of the RX 6750 XT. OpenCL is a generic compute workload, and the GP106 chip in the NVIDIA card has a very high FP32 throughput and a driver that is well optimized for this type of load. The RX 6750 XT also has FP32 (13.31 TFLOPS) but the NVIDIA card records a higher raw score in this database entry.

Vulkan and cross-API geometry: The RX 6750 XT wins again, with a Geekbench Vulkan score of 98089 vs 32897 for the P106-100. That is a 198.2% advantage. This test includes geometry and compute, and the AMD chip with its 2560 shading units and 40 RT cores is far more suitable for the new explicit graphics API than the older Pascal-based part.

So the use case is split: the RX 6750 XT is for gamers and content creators who run modern 3D APIs and want the full feature set (ray tracing, DirectX 12 Ultimate, Vulkan 1.4). The P106-100 is for compute-heavy compute workloads, especially OpenCL / OpenCL-like or CUDA, and it does not need a display output (it has no display outputs at all, the database lists "no outputs").

Architecture Differences

The RX 6750 XT is an RDNA 2.0 chip, manufactured at 7 nm at TSMC. It packs 17,200 million transistors into a 335 mm² die, yielding a transistor density of 51.3 million per mm². The P106-100 is a GP106 chip, manufactured at 16 nm at TSMC, with 4,400 million transistors on a 200 mm² die, density 22.0 million per mm².

The RX 6750 XT has 2560 shading units, 160 texture memory units, and 64 ROPs. It also has 40 ray tracing cores, but no tensor cores. The P106-100 has 1280 shading units, 80 TMUs, and 48 ROPs; it has no RT cores and no tensor cores. The memory subsystem is a major differentiator: the RX 6750 XT carries 12 GB of GDDR6 on a 192-bit bus, with a bandwidth of 432.0 GB/s. The P106-100 has 6 GB of GDDR5 on a 192-bit bus, 192.2 GB/s.

The clocks tell the story of their intent. The RX 6750 XT has a base of 2150 MHz, a boost of 2600 MHz, and a game clock of 2495 MHz. The memory runs at 2250 MHz effectively (18 Gbps effective). The P106-100 has a base of 1506 MHz, boost 1709 MHz, memory 2002 MHz (8 Gbps effective). The AMD chip is clocked substantially higher, and that is why it gets its throughput advantage in geometry-heavy workloads.

The P106 also has a significantly lower power profile: the database lists the RX 6750 XT with a TDP of 250 W, requiring a 600 W power supply and one 6-pin plus one 8-pin power connector. The P106-100 has a 120 W TDP, a 300 W PSU, and a single 6-pin connector. The P106 is a dual-slot card, as is the RX 6750 XT.

The P106-100 has no display outputs, which is key for a mining part; the RX 6750 XT has one HDMI 2.1 output and three DisplayPort 1.4a outputs. The P106-100 is classified as End-of-life in the database; the RX 6750 XT is also End-of-life. Both are no longer in active production.

The Verdict

The data in the database is quite clear on which part should be used for which task. Choose the AMD Radeon RX 6750 XT if you are building a system that will run modern DirectX 12 and Vulkan titles, or any workload that uses the full feature set of a modern GPU: ray tracing enabled titles, high-resolution 3D rendering, or compute in Vulkan. It has a 183.4% winning margin in 3DMark Steel Nomad, a 198.2% winning margin in Geekbench Vulkan, and a 71st percentile overall position. It is the more balanced card, with 2560 shading units, 40 RT cores, and 12 GB of memory.

Choose the NVIDIA P106-100 for raw compute throughput in OpenCL or similar compute APIs, or for a mining/compute card where no display output is needed. It wins Geekbench OpenCL by 54.5%, sits at the 68th percentile, and has a lower power draw (120 W vs 250 W) and a lower PSU recommendation (300 W vs 600 W). It is a mining part with no display outputs, so it is a board for computing, not for gaming.

FAQ

Q: Which card is faster in 3DMark Steel Nomad DX12?

A: The AMD Radeon RX 6750 XT is faster, scoring 2548 vs 899, a 183.4% advantage. The NVIDIA P106-100 cannot keep up with the AMD part in this modern DX12 workload.

Q: Which card wins in Geekbench OpenCL?

A: The NVIDIA P106-100 wins Geekbench OpenCL, 35951 vs 16360, a 54.5% lead. This is a compute-heavy test, and the Pascal chip with its high FP32 throughput takes the win.

Q: Does the RX 6750 XT have ray tracing cores?

A: Yes. The RX 6750 XT has 40 ray tracing cores. The P106-100 has no ray tracing cores.

Q: How much memory does each card have?

A: The RX 6750 XT has 12 GB of GDDR6 on a 192-bit bus with 432.0 GB/s of bandwidth. The P106-100 has 6 GB of GDDR5 on a 192-bit bus with 192.2 GB/s of bandwidth.

Q: What is the power consumption of each card?

A: The RX 6750 XT is rated at 250 W TDP and comes with a 600 W PSU recommendation. The P106-100 is rated at 120 W TDP and comes with a 300 W PSU recommendation.

Q: Which card has display outputs?

A: The RX 6750 XT has one HDMI 2.1 and three DisplayPort 1.4a outputs. The P106-100 has no display outputs.