NVIDIA GeForce RTX 3050 OEM vs NVIDIA P106-090 Comparison

Head-to-Head Benchmarks

The recorded data contains two direct head-to-head comparisons between the NVIDIA GeForce RTX 3050 OEM and the NVIDIA P106-090, and in both cases the RTX 3050 OEM dominates decisively. The most striking gap appears in Geekbench Vulkan, where the RTX 3050 OEM scores 57103 against the P106-090’s 18596, a delta of 207.1%. That is more than triple the raw score, indicating a fundamental advantage in compute throughput and driver-level Vulkan optimization. The Geekbench OpenCL result is similarly lopsided: the RTX 3050 OEM posts 60740 while the P106-090 manages 21304, a 185.1% difference. These are not marginal wins; they represent entirely different performance tiers.

The broader benchmark database confirms the separation. The RTX 3050 OEM has an average benchmark score of 15199 across all recorded tests, placing it at the 57th percentile among all GPUs. Its nearest rivals include the AMD Radeon RX 7600 (avg score 15171, delta 0.2%), the AMD Radeon 680M (avg score 15270, delta -0.5%), and the NVIDIA GeForce GTX 580 (avg score 15283, delta -0.5%). This clustering shows the RTX 3050 OEM sits in a competitive mid-range band where small percentage differences separate it from peers. The P106-090, by contrast, carries an average benchmark score of 13470, good for the 54th percentile. Its nearest rivals are the NVIDIA GeForce GTX 570 (avg score 13515, delta -0.3%), the AMD Radeon Pro 555 (avg score 13407, delta 0.5%), and the AMD Radeon HD 7770M (avg score 13536, delta -0.5%). The P106-090 is thus a full tier lower, roughly 12% behind the RTX 3050 OEM in average score (15199 vs 13470).

The head-to-head deltas are far larger than the average score gap suggests, primarily because the two direct tests (OpenCL and Vulkan) stress raw compute and modern API paths where the RTX 3050 OEM’s architecture is vastly superior. The P106-090 has no recorded results in DirectX 10, 11, 12, or 9 tests, nor in Passmark G2D, G3D, or GPU compute suites. The RTX 3050 OEM, meanwhile, has a full suite of Passmark scores: DirectX 10 at 61, DirectX 11 at 86, DirectX 12 at 58, DirectX 9 at 137, G2D at 973, G3D at 11857, and GPU compute at 5779. These numbers show a card that excels in older DirectX 9 workloads (137) while still delivering respectable DirectX 11 performance (86), but the G3D score of 11857 is the headline figure for traditional rasterization.

Where Each One Wins

The RTX 3050 OEM wins every recorded head-to-head benchmark, so the use-case split is straightforward: it is the superior choice for any application that relies on OpenCL or Vulkan acceleration. In OpenCL, the 60740 score indicates strong compute throughput for tasks like video encoding, physics simulation, or general-purpose GPU compute. The P106-090’s 21304 score is roughly one-third of that, meaning compute-heavy workloads will see a massive slowdown. In Vulkan, the RTX 3050 OEM’s 57103 versus 18596 highlights its advantage in modern gaming APIs and cross-vendor graphics workloads. Vulkan is increasingly common in game engines and professional rendering tools, so this gap translates directly to real-world performance differences.

The P106-090 has no wins in any recorded benchmark, but its existence as a mining-specific card (generation listed as "Mining GPUs") suggests its strengths lie outside the measured tests. It has no display outputs, meaning it cannot drive a monitor directly, and its PCIe interface is limited to PCIe 1.0 x1, which severely constrains data transfer between the CPU and GPU. The RTX 3050 OEM, by contrast, uses PCIe 4.0 x8, offering up to eight times the theoretical bandwidth of the P106-090’s interface. For gaming, content creation, or any interactive workload, the RTX 3050 OEM is the only viable option of the two.

For users who need a GPU purely for compute tasks that do not require display output and can tolerate a low-bandwidth PCIe link, the P106-090 might still function, but the benchmark data shows it will deliver less than half the performance of the RTX 3050 OEM in the two tests where both have scores. The RTX 3050 OEM also supports DirectX 12 Ultimate (12_2), while the P106-090 only reaches DirectX 12 (12_1). That means the RTX 3050 OEM can handle ray tracing and other DirectX 12 Ultimate features at the API level, whereas the P106-090 lacks those capabilities entirely. The RTX 3050 OEM includes 18 RT cores and 72 tensor cores; the P106-090 has neither. Any workload that leverages ray tracing, DLSS, or tensor-core acceleration will simply not run on the P106-090.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA GeForce RTX 3050 OEM has an average benchmark score of 15199, while the NVIDIA P106-090 averages 13470. That places the RTX 3050 OEM at the 57th percentile of all GPUs, versus the 54th percentile for the P106-090.

Q: How large is the performance gap in Vulkan?

A: The RTX 3050 OEM scores 57103 in Geekbench Vulkan, which is 207.1% higher than the P106-090’s 18596. This is the largest recorded delta between the two cards.

Q: Does the P106-090 support DirectX 12 Ultimate?

A: No. The P106-090 supports DirectX 12 (12_1), while the RTX 3050 OEM supports DirectX 12 Ultimate (12_2). The RTX 3050 OEM also has 18 RT cores and 72 tensor cores, neither of which are present on the P106-090.

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

A: The P106-090 has no display outputs, so it cannot connect to a monitor directly. The RTX 3050 OEM offers 1x HDMI 2.1 and 3x DisplayPort 1.4a outputs, making it suitable for standard display use.

Q: What is the memory configuration difference?

A: The RTX 3050 OEM has 8 GB of GDDR6 memory on a 128-bit bus, yielding 224.0 GB/s of bandwidth. The P106-090 has 3 GB of GDDR5 on a 192-bit bus, yielding 192.2 GB/s. The RTX 3050 OEM has more total memory but a narrower bus; the bandwidth advantage still favors the RTX 3050 OEM.

Q: How do the transistor counts compare?

A: The RTX 3050 OEM uses 12,000 million transistors on a 276 mm² die, while the P106-090 uses 4,400 million transistors on a 200 mm² die. The RTX 3050 OEM’s transistor density is 43.5M per mm² versus 22.0M per mm² for the P106-090.

Specification Differences

The two cards differ across nearly every core specification. The RTX 3050 OEM has 2304 shading units, 72 texture mapping units, and 32 ROPs, while the P106-090 has 768 shading units, 48 TMUs, and 48 ROPs. The RTX 3050 OEM’s shading unit count is triple that of the P106-090, which explains its massive compute advantage. The P106-090 has more ROPs (48 vs 32), but that does not compensate for the far lower shader count. The RTX 3050 OEM includes 18 RT cores and 72 tensor cores; the P106-090 has none. The pixel rate for the RTX 3050 OEM is 56.16 GPixel/s, while the P106-090 achieves 73.49 GPixel/s, a rare specification where the older card is faster. The texture rate, however, favors the RTX 3050 OEM at 126.4 GTexel/s versus 73.49 GTexel/s for the P106-090.

FP32 performance is 8.087 TFLOPS for the RTX 3050 OEM and 2.352 TFLOPS for the P106-090. FP16 performance is also 8.087 TFLOPS (1:1) for the RTX 3050 OEM, but only 36.74 GFLOPS (1:64) for the P106-090. That is a 220x difference in FP16 throughput, reflecting the RTX 3050 OEM’s modern shader design. Memory differs: 8 GB GDDR6 at 14 Gbps effective for the RTX 3050 OEM versus 3 GB GDDR5 at 8 Gbps effective for the P106-090. Bus widths are 128-bit and 192-bit, respectively, but bandwidth is 224.0 GB/s for the RTX 3050 OEM and 192.2 GB/s for the P106-090.

Power requirements also differ substantially. The RTX 3050 OEM has a TDP of 130 W, uses a 1x 8-pin power connector, and suggests a 300 W PSU. The P106-090 has a TDP of 75 W, uses a 1x 6-pin connector, and suggests a 250 W PSU. Both are dual-slot cards, but the RTX 3050 OEM is 242 mm long (9.5 inches) with a height of 112 mm (4.4 inches), while the P106-090 is 250 mm long (9.8 inches) with no recorded height. The RTX 3050 OEM uses a PCIe 4.0 x8 interface; the P106-090 uses PCIe 1.0 x1. Display outputs are entirely absent on the P106-090, while the RTX 3050 OEM has 1x HDMI 2.1 and 3x DisplayPort 1.4a.

Architecture Differences

The RTX 3050 OEM is built on the Ampere architecture using the GA106 chip, fabricated on Samsung’s 8 nm process. The P106-090 uses the Pascal architecture with the GP106 chip, fabricated on TSMC’s 16 nm process. The process node difference is significant: 8 nm versus 16 nm allows the RTX 3050 OEM to pack 12,000 million transistors into a 276 mm² die, while the P106-090 fits 4,400 million transistors into a 200 mm² die. Transistor density is 43.5M per mm² for the RTX 3050 OEM versus 22.0M per mm² for the P106-090, a doubling in density that enables the newer card’s higher core counts and specialized hardware.

The RTX 3050 OEM belongs to the GeForce 30-series generation, with its predecessor in the GeForce 20 series and successor in the GeForce 40 series. The P106-090 is listed under "Mining GPUs" with no predecessor or successor, reflecting its purpose-built mining role. The RTX 3050 OEM supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The P106-090 supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. Both have identical OpenGL and Vulkan API levels, but the DirectX feature level differs, meaning the RTX 3050 OEM can use the full DirectX 12 Ultimate feature set including ray tracing and variable-rate shading.

The RTX 3050 OEM has 18 RT cores for hardware ray tracing and 72 tensor cores for AI acceleration, features entirely absent from the P106-090. The P106-090’s architecture is a stripped-down Pascal design, lacking any dedicated ray tracing or tensor hardware. Its FP16 ratio of 1:64 indicates that half-precision compute is severely deprioritized, whereas the RTX 3050 OEM offers 1:1 FP16 performance, making it far more flexible for mixed-precision workloads. The P106-090’s lack of display outputs and PCIe 1.0 x1 interface further reinforce its mining-only design. In contrast, the RTX 3050 OEM is a complete consumer GPU with display outputs, modern PCIe connectivity, and a full feature set.

The release dates are also far apart: the RTX 3050 OEM launched on 2022-01-03, while the P106-090 launched on 2017-07-30. Both are end-of-life, but the RTX 3050 OEM is a more recent design. The P106-090’s 16 nm process and Pascal architecture predate the RTX 3050 OEM’s 8 nm Ampere by roughly four and a half years. That generational gap explains the benchmark results: the RTX 3050 OEM is not just faster in raw numbers, it is architecturally superior in every measurable way except for pixel rate and ROP count, where the P106-090 holds narrow leads. Those two exceptions do little to offset the massive compute, memory, and API capability advantages of the RTX 3050 OEM.