GPU Comparison

The data places the NVIDIA GeForce 920M and the NVIDIA Quadro P1000 in a clear hierarchy, despite both sharing the same 20th percentile ranking among all GPUs. The Quadro P1000 is decisively faster in every head-to-head benchmark, leading by 72.6% in Geekbench OpenCL and 63.2% in Geekbench Vulkan. However, the 920M’s average benchmark score of 3287 actually exceeds the P1000’s 3163 average, a counterintuitive result driven by the P1000’s weaker Passmark DirectX 10 and 12 scores. The 920M is an end-of-life mobile integrated part from 2015, while the P1000 is a 2017 single-slot workstation card, so the choice depends on whether the user prioritizes the P1000’s raw compute dominance or the 920M’s more consistent average performance across a broader test suite. For professional workloads leveraging OpenCL or Vulkan, the Quadro P1000 is the only rational pick; for legacy compatibility or scenarios where the 920M’s specific driver profile is required, the older chip remains a functional, if far slower, option.

Architecture Differences

The two GPUs represent fundamentally different design philosophies separated by two process generations. The GeForce 920M uses the GK208B chip built on Kepler 2.0 architecture, manufactured by TSMC on a 28 nm process. It packs 1,020 million transistors into an 87 mm² die, yielding a transistor density of 11.7M / mm². In contrast, the Quadro P1000 employs the GP107 chip with Pascal architecture, fabricated by Samsung on a 14 nm node. This newer process allows 3,300 million transistors in a 132 mm² die, achieving 25.0M / mm² density, more than double the 920M’s density per square millimeter.

The shading resources diverge sharply. The 920M carries 384 shading units, 32 TMUs, and just 8 ROPs. The P1000 more than doubles the ROP count with 32 ROPs, while increasing shading units to 640 and TMUs to 40. Neither GPU includes ray tracing cores or tensor cores, as both predate those features, but the architectural gap manifests in raw throughput: the 920M delivers 732.7 GFLOPS of FP32 compute versus the P1000’s 1.894 TFLOPS. The P1000 also exposes a token FP16 capability at 29.60 GFLOPS (1:64), while the 920M lists no FP16 support at all.

Memory subsystems are equally divergent. The 920M uses 2 GB of DDR3 on a 64-bit bus, producing 14.40 GB/s of bandwidth. The P1000 uses 4 GB of GDDR5 on a 128-bit bus, achieving 80.19 GB/s, a 5.6x bandwidth advantage. Clock behavior also differs: the 920M runs at a flat 954 MHz base and boost, while the P1000 boosts from 1266 MHz to 1480 MHz. The P1000’s memory clock is 1253 MHz (5 Gbps effective) versus the 920M’s 900 MHz (1800 Mbps effective).

API support reveals the generational leap. The 920M supports DirectX 12 (11_0) and Vulkan 1.2.175, while the P1000 supports DirectX 12 (12_1) and Vulkan 1.4. Both share OpenGL 4.6. The P1000 also benefits from a PCIe 3.0 x16 interface versus the 920M’s PCIe 3.0 x8, and it provides 4x mini-DisplayPort 1.4a outputs, whereas the 920M’s display outputs are “Portable Device Dependent.”

Head-to-Head Benchmarks

Only two shared benchmarks appear in the data, and the Quadro P1000 wins both by overwhelming margins. In Geekbench OpenCL, the P1000 scores 13584 against the 920M’s 3725, a delta of -72.6% from the perspective of the 920M. This 3.6x performance gap aligns with the raw compute difference: the P1000’s 1.894 TFLOPS versus 732.7 GFLOPS suggests the OpenCL workload scales nearly linearly with FP32 throughput. The P1000’s larger memory bus and GDDR5 bandwidth likely prevent any memory-bound stalls that could otherwise let the 920M close the gap.

In Geekbench Vulkan, the P1000 scores 7739 versus 2849 for the 920M, a -63.2% delta. The Vulkan margin is slightly narrower than OpenCL, possibly because Vulkan’s lower overhead allows the 920M’s Kepler architecture to utilize its limited resources more efficiently. Still, the P1000 maintains a 2.7x lead. Notably, the 920M’s Vulkan score of 2849 is only 23% below its OpenCL score of 3725, while the P1000’s Vulkan score is 43% below its OpenCL result, suggesting the P1000’s Vulkan drivers are either less mature or the workload favors the 920M’s simpler architecture disproportionately.

The wins stand at 2 for the P1000, 0 for the 920M. But the broader benchmark picture complicates the story. The P1000’s Passmark scores reveal severe weaknesses: DirectX 10 at 21, DirectX 11 at 31, DirectX 12 at 19, and DirectX 9 at 79. These are anomalously low relative to its G3D score of 4512 and G2D score of 589. The 920M has no Passmark scores in the data, so a direct comparison there is impossible, but the P1000’s average score of 3163, dragged down by those single-digit-to-low-double-digit DirectX results, actually sits below the 920M’s average of 3287.

Specification Differences

| Specification | NVIDIA GeForce 920M | NVIDIA Quadro P1000 |

|---|---|---|

| Chip | GK208B | GP107 |

| Architecture | Kepler 2.0 | Pascal |

| Process Node | 28 nm (TSMC) | 14 nm (Samsung) |

| Transistors | 1,020 million | 3,300 million |

| Die Size | 87 mm² | 132 mm² |

| Transistor Density | 11.7M / mm² | 25.0M / mm² |

| Base Clock | 954 MHz | 1266 MHz |

| Boost Clock | 954 MHz | 1480 MHz |

| Memory Size | 2 GB | 4 GB |

| Memory Type | DDR3 | GDDR5 |

| Memory Bus | 64 bit | 128 bit |

| Memory Bandwidth | 14.40 GB/s | 80.19 GB/s |

| Shading Units | 384 | 640 |

| TMUs | 32 | 40 |

| ROPs | 8 | 32 |

| Pixel Rate | 7.632 GPixel/s | 47.36 GPixel/s |

| Texture Rate | 30.53 GTexel/s | 59.20 GTexel/s |

| FP32 | 732.7 GFLOPS | 1.894 TFLOPS |

| FP16 |, | 29.60 GFLOPS (1:64) |

| TDP | 33 W | 47 W |

| Slot Width | IGP | Single-slot |

| Bus Interface | PCIe 3.0 x8 | PCIe 3.0 x16 |

| Display Outputs | Portable Device Dependent | 4x mini-DisplayPort 1.4a |

| DirectX | 12 (11_0) | 12 (12_1) |

| Vulkan | 1.2.175 | 1.4 |

| Release Date | 2015-03-12 | 2017-02-06 |

| Predecessor | GeForce 800M | Quadro Maxwell |

| Successor | GeForce 10 Mobile | Quadro Volta |

FAQ

Q: Which GPU has higher raw compute performance?

A: The Quadro P1000 delivers 1.894 TFLOPS of FP32 compute versus the GeForce 920M’s 732.7 GFLOPS, a 2.6x advantage. This is reflected in the Geekbench OpenCL score of 13584 for the P1000 against 3725 for the 920M.

Q: Why does the 920M have a higher average benchmark score than the P1000?

A: The 920M’s average is 3287, while the P1000’s is 3163. The P1000’s average is pulled down by its extremely low Passmark DirectX scores, 21 for DirectX 10, 31 for DirectX 11, and 19 for DirectX 12, which are not present in the 920M’s benchmark list.

Q: Is the memory bandwidth difference significant?

A: Yes. The P1000’s 80.19 GB/s bandwidth from GDDR5 on a 128-bit bus is over 5.5 times the 920M’s 14.40 GB/s from DDR3 on a 64-bit bus. This likely explains the P1000’s larger lead in OpenCL versus Vulkan, as memory-bound workloads benefit more from the bandwidth increase.

Q: Which GPU supports newer graphics APIs?

A: The Quadro P1000 supports DirectX 12 (12_1) and Vulkan 1.4, while the GeForce 920M supports DirectX 12 (11_0) and Vulkan 1.2.175. Both support OpenGL 4.6.

Q: Are these GPUs still in production?

A: No. Both are listed as End-of-life production status. The 920M was released on 2015-03-12, and the P1000 on 2017-02-06.

Q: How do they compare to their nearest rivals in the data?

A: The 920M is 0.9% slower than the GeForce GT 730M, 1.4% slower than Intel HD Graphics 530, 2.4% faster than the GeForce GT 640, and 3% slower than Intel HD Graphics P4600. The P1000 is 0.6% slower than Intel Arc Pro B60, 1.5% slower than the GeForce GT 640, 2.9% faster than the GeForce RTX 5080 SUPER, and 3.8% slower than the GeForce 920M.

Where Each One Wins

The Quadro P1000 wins decisively in compute-heavy and modern API workloads. Its Geekbench OpenCL score of 13584 is nearly four times the 920M’s 3725, and its Vulkan score of 7739 is 2.7 times the 920M’s 2849. The P1000 also wins on every raw specification that matters for throughput: 640 shading units versus 384, 32 ROPs versus 8, 80.19 GB/s bandwidth versus 14.40 GB/s, and a PCIe 3.0 x16 interface versus x8. Its 4 GB GDDR5 frame buffer doubles the 920M’s 2 GB DDR3, and its 47.36 GPixel/s pixel rate is 6.2 times higher. For professional graphics, CAD, or any OpenCL/Vulkan compute task, the P1000 is the clear choice, the data shows no contest in these shared benchmarks.

The GeForce 920M wins on the aggregate benchmark average, scoring 3287 against the P1000’s 3163. This is a narrow margin, but it stems from the P1000’s catastrophic Passmark DirectX results, scores of 21 (DX10), 31 (DX11), and 19 (DX12) suggest severe driver or architectural issues in legacy DirectX paths that the 920M does not exhibit (though the 920M lacks these test scores entirely). The 920M also consumes less power at 33 W versus 47 W, and its IGP slot width means it requires no physical expansion slot, making it suitable for ultra-thin portable devices where the P1000’s single-slot 150 mm length and 69 mm height would not fit. For users running older DirectX 9/10/11 applications where the P1000’s Passmark numbers indicate potential problems, the 920M’s more consistent profile might be preferable, but this is a niche scenario, as the P1000’s modern API support and 2.6x FP32 advantage dominate any forward-looking use case.