Intel Arc A750 vs NVIDIA P106-100 Comparison

Head-to-Head Benchmarks

The recorded data shows a decisive sweep in favor of the Intel Arc A750 across all three shared benchmark tests. The largest gap appears in the 3DMark Steel Nomad DX12 workload, where the Intel card scores 2612 against 899 for the NVIDIA P106-100, a delta of 65.6%. This test stresses modern DirectX 12 rendering pipelines, and the margin indicates a fundamental difference in raw geometry and shading throughput between the two architectures.

In compute-oriented workloads, the Intel Arc A750 continues its dominance. The Geekbench OpenCL score of 98554 versus 35951 represents a 63.5% advantage, while the Vulkan result of 85631 versus 32897 is a 61.6% lead. These are not narrow victories; the Intel card more than doubles the NVIDIA card's output in every single measured test. The pattern is consistent: whether the workload is graphics-bound or compute-bound, the Arc A750 holds a commanding lead.

The NVIDIA P106-100 recorded zero wins in the head-to-head comparison. Its best relative showing is in Geekbench Vulkan, where the 61.6% deficit is slightly smaller than the OpenCL gap, but it remains a substantial loss. The database also shows the P106-100 has no display outputs, which limits its practical utility to compute or mining scenarios, whereas the Arc A750 includes 1x HDMI 2.1 and 3x DisplayPort 2.0 outputs.

Looking at the broader benchmark context, the NVIDIA card's average benchmark score of 23249 places it at the 68th percentile of all GPUs, while the Intel card's average of 20582 sits at the 66th percentile. This is a curious inversion: the Arc A750 wins every head-to-head test, yet its aggregate average is pulled down by additional benchmark entries that are not shared with the P106-100. The Intel card's PassMark scores include a DirectX 9 result of 181, DirectX 10 at 65, DirectX 11 at 72, and DirectX 12 at 70, along with a G2D score of 732, G3D score of 12534, and GPU compute score of 5368. These legacy DX9 and DX10 numbers are particularly low relative to the G3D figure, which drags the overall average down.

The nearest rival comparisons reinforce this split. The P106-100's closest competitors are separated by a tenth of a percent or less: the AMD Radeon Pro Vega 16 matches it exactly at 23250, the AMD Radeon RX 6600M trails by 0.1%, and the AMD Radeon R9 M290X also trails by 0.1%. The Arc A750's nearest rivals include the Intel Arc B570 at 0.1% ahead, the NVIDIA GeForce RTX 3070 Mobile at 0.2% ahead, and the AMD Radeon R9 M390X at 0.4% behind. The delta values confirm that both cards sit in tightly contested performance tiers relative to their peers, despite the massive gap between them.

The Verdict

The data supports only one conclusion for performance: the Intel Arc A750 is categorically faster than the NVIDIA P106-100 in every shared workload. A 65.6% lead in 3DMark Steel Nomad DX12, a 63.5% lead in OpenCL, and a 61.6% lead in Vulkan leave no ambiguity. Users seeking maximum compute or graphics throughput should choose the Arc A750.

However, the P106-100 has a narrow but real niche. It draws 120 W with a suggested PSU of 300 W, compared to 225 W and 550 W for the Arc A750. For systems with strict power delivery constraints, the NVIDIA card is the lighter load. It is also physically smaller at 250 mm or 9.8 inches in length, while the Arc A750's dimensions are not recorded in the database. The P106-100 requires only a single 6-pin power connector, while the Arc A750 needs a 6-pin and an 8-pin. Neither card is current production; both are listed as end-of-life.

The Arc A750 launched with an MSRP of 289 USD, which can be stated once as a reference point. The P106-100 has no recorded launch MSRP. Buyers should weigh the performance gap against the power and physical requirements. For compute tasks where power density matters and display output is irrelevant, the P106-100 could suffice. For anything requiring modern graphics performance, display connectivity, or high-throughput compute, the Arc A750 is the only defensible pick from this data.

Architecture Differences

The two GPUs come from different architectural generations and process nodes. The NVIDIA P106-100 uses the GP106 chip based on the Pascal architecture, manufactured on a 16 nm process at TSMC. It integrates 4,400 million transistors on a 200 mm² die, yielding a transistor density of 22.0 million per square millimeter. The Intel Arc A750 uses the DG2-512 chip based on the Xe-HPG architecture, manufactured on a 6 nm process, also at TSMC. It packs 21,700 million transistors on a 406 mm² die, for a density of 53.4 million per square millimeter. The density difference is stark: Intel packs more than twice as many transistors per area, a direct result of the newer node.

The memory subsystems differ substantially. The P106-100 has 6 GB of GDDR5 on a 192-bit bus, delivering 192.2 GB/s of bandwidth. The Arc A750 has 8 GB of GDDR6 on a 256-bit bus, delivering 512.0 GB/s. That is over 2.6 times the memory bandwidth, which explains part of the compute and graphics performance gap. The memory clocks also differ: the NVIDIA card runs at 2002 MHz with 8 Gbps effective, while the Intel card runs at 2000 MHz with 16 Gbps effective. The effective data rate doubling, combined with the wider bus, drives the bandwidth advantage.

Execution resources are heavily skewed toward the Intel card. The Arc A750 has 3584 shading units, 224 texture mapping units, and 112 raster output units. The P106-100 has 1280 shading units, 80 TMUs, and 48 ROPs. The Intel card also includes 28 ray tracing cores, while the NVIDIA card has none. Pixel rate for the Arc A750 is 268.8 GPixel/s versus 82.03 GPixel/s, and texture rate is 537.6 GTexel/s versus 136.7 GTexel/s. Floating point performance follows the same pattern: the Arc A750 delivers 17.20 TFLOPS of FP32 and 34.41 TFLOPS of FP16 at a 2:1 ratio, while the P106-100 delivers 4.375 TFLOPS of FP32 and a heavily reduced 68.36 GFLOPS of FP16 at a 1:64 ratio. The FP16 situation is particularly notable: the Pascal card is effectively crippled for half-precision work, while the Intel card excels at it.

Clock speeds also favor Intel. The Arc A750 has a base clock of 2050 MHz and a boost of 2400 MHz, versus 1506 MHz base and 1709 MHz boost for the P106-100. The bus interface differs as well: the P106-100 uses PCIe 1.0 x16, an unusual choice that likely reflects its mining-oriented design, while the Arc A750 uses PCIe 4.0 x16. The NVIDIA card's PCIe 1.0 interface is a severe bottleneck for any workload that transfers data over the bus, further limiting its usefulness outside of pure compute tasks.

API support is broadly similar with one key exception. Both cards support DirectX 12, OpenGL 4.6, and Vulkan 1.4. But the Intel card supports DirectX 12 Ultimate (12_2), which includes features like ray tracing and mesh shaders, while the NVIDIA card is limited to DirectX 12 (12_1). The Arc A750 also has display outputs (1x HDMI 2.1, 3x DisplayPort 2.0), while the P106-100 has none. The P106-100's generation is listed as "Mining GPUs," which explains the missing display outputs and the older PCIe interface. The Arc A750 belongs to the Alchemist (Arc 7) generation, with a predecessor of Xe Graphics and a successor of Battlemage.

FAQ

Q: Which GPU has higher raw performance in the shared benchmarks?

A: The Intel Arc A750 wins all three head-to-head tests. It leads by 65.6% in 3DMark Steel Nomad DX12, 63.5% in Geekbench OpenCL, and 61.6% in Geekbench Vulkan.

Q: What is the memory bandwidth difference?

A: The Intel Arc A750 has 512.0 GB/s of bandwidth from 8 GB of GDDR6 on a 256-bit bus. The NVIDIA P106-100 has 192.2 GB/s from 6 GB of GDDR5 on a 192-bit bus.

Q: Does the NVIDIA P106-100 support display output?

A: No. The database lists its display outputs as "No outputs." The Intel Arc A750 has 1x HDMI 2.1 and 3x DisplayPort 2.0.

Q: How do the FP16 capabilities compare?

A: The Arc A750 delivers 34.41 TFLOPS of FP16 at a 2:1 ratio. The P106-100 delivers only 68.36 GFLOPS at a 1:64 ratio, a massive difference in half-precision throughput.

Q: What are the power requirements for each card?

A: The P106-100 has a TDP of 120 W with a suggested PSU of 300 W and one 6-pin connector. The Arc A750 has a TDP of 225 W with a suggested PSU of 550 W and one 6-pin plus one 8-pin connector.

Q: Which card has ray tracing cores?

A: Only the Intel Arc A750. It includes 28 ray tracing cores, while the P106-100 has none recorded.

Where Each One Wins

The Intel Arc A750 wins in every measurable performance category from the shared tests. For 3D rendering workloads, its 65.6% lead in Steel Nomad DX12 makes it the clear choice. For general-purpose compute through OpenCL, the 63.5% margin is decisive. For Vulkan-based applications, the 61.6% advantage holds. The Arc A750 also wins on memory bandwidth (512.0 GB/s versus 192.2 GB/s), FP32 throughput (17.20 TFLOPS versus 4.375 TFLOPS), FP16 throughput (34.41 TFLOPS versus 68.36 GFLOPS), and pixel/texture rates (268.8 GPixel/s versus 82.03 GPixel/s, and 537.6 GTexel/s versus 136.7 GTexel/s). It adds ray tracing support, DirectX 12 Ultimate features, and display outputs. Any workload that touches modern graphics APIs or requires high-throughput compute will favor the Arc A750.

The NVIDIA P106-100 wins only in power efficiency and physical footprint. Its 120 W TDP is nearly half of the Arc A750's 225 W, and its 300 W suggested PSU is far below the 550 W required by the Intel card. It needs just one 6-pin connector versus the 6-pin and 8-pin pair on the Arc A750. Its 250 mm length is compact, and the database records no length for the Arc A750, but the P106-100's smaller die (200 mm² versus 406 mm²) suggests a lighter board. For systems with limited power budgets, or for compute tasks that do not require display output or high memory bandwidth, the P106-100 could be the lower-friction option. The absence of display outputs makes it unusable for desktop graphics, but for headless compute or mining workloads, the lower power draw is a genuine advantage.

The percentile rankings complicate the picture slightly. The P106-100 sits at the 68th percentile of all GPUs, while the Arc A750 sits at the 66th. This reflects the different benchmark suites each card was subjected to. The P106-100's average of 23249 comes from just three tests, all of which are modern and compute-heavy. The Arc A750's average of 20582 includes ten tests, several of which are legacy DirectX 9 and DirectX 10 workloads that score poorly (181 and 65 respectively). In the tests they share, the Arc A750 wins by wide margins. The percentile gap should not be read as the NVIDIA card being faster overall; it is an artifact of test selection. The head-to-head data is unambiguous.

For a user choosing between these two end-of-life cards, the decision hinges on context. If the workload is modern, graphics-intensive, or compute-heavy, the Arc A750 wins outright. If the workload is power-constrained, headless, and does not need bandwidth or FP16, the P106-100 offers a lighter alternative. There is no scenario in the recorded data where the P106-100 outperforms the Arc A750 in speed. The only wins are in power consumption and physical requirements.