GPU Comparison

The NVIDIA GeForce 920MX and NVIDIA GeForce 920M are two mobile graphics solutions from the same GeForce 900M generation, yet benchmark results show a clear performance hierarchy between them. While both parts are end-of-life products aimed at entry-level laptops, the data reveals that the 920MX consistently outperforms the 920M in every recorded test, though the margins vary significantly by workload.

Head-to-Head Benchmarks

The head-to-head data is unambiguous: the GeForce 920MX wins both benchmark comparisons, taking 2 wins to 0 for the 920M. In the Geekbench OpenCL test, the 920MX scores 4068 against the 920M's 3725, a delta of 9.2%. This is the larger of the two margins and indicates a meaningful advantage in general-purpose compute workloads that leverage the OpenCL API. The Vulkan test shows a narrower gap: the 920MX scores 2987 while the 920M scores 2849, a 4.8% delta. While the 920MX remains ahead in Vulkan, the reduced margin suggests that graphics-oriented tasks may not translate the full compute advantage into frame-rate gains.

Looking at average benchmark scores across all recorded tests, the 920MX posts 3528, which is 7.3% higher than the 920M's 3287. This aggregate figure confirms that the 920MX's lead is consistent rather than isolated to a single test. The percentile rankings reinforce this: the 920MX sits at the 21st percentile of all GPUs, while the 920M sits at the 20th percentile. Both are entry-level performers, but the 920MX edges out its sibling in overall standing.

Context from nearest rivals helps calibrate these scores. The 920MX's average score of 3528 places it just 1.8% behind the NVIDIA GeForce GT 545 (avg score 3594) and 1.9% behind the NVIDIA RTX 5000 Mobile Ada Generation (avg score 3596). It is 2.4% behind the GeForce GT 735M (avg score 3616) but 2.7% ahead of the Quadro 2000M (avg score 3434). For the 920M, its average of 3287 is 0.9% behind the GeForce GT 730M (avg score 3316) and 1.4% behind Intel HD Graphics 530 (avg score 3332), while it is 2.4% ahead of the GeForce GT 640 (avg score 3210) and 3% behind Intel HD Graphics P4600 (avg score 3389). Notably, the 920MX's closest rival cluster includes the RTX 5000 Mobile Ada Generation, a modern professional card, which highlights how benchmark scores can cluster despite vast architectural differences.

Architecture Differences

The two GPUs come from different architectures despite sharing the GeForce 900M generation label. The 920MX is built on the Maxwell architecture using the GM108S chip, while the 920M uses the older Kepler 2.0 architecture with the GK208B chip. Both are fabricated by TSMC on a 28 nm process node, and both pack the same 1,020 million transistors. The die sizes differ: the 920MX measures 77 mm², while the 920M is larger at 87 mm². This yields a transistor density of 13.2M per mm² for the 920MX versus 11.7M per mm² for the 920M, reflecting Maxwell's more efficient use of silicon.

The core configurations tell a more complex story. The 920MX has 256 shading units, 24 texture mapping units, and 8 ROPs. The 920M, despite its older architecture, has a higher count of shading units at 384, 32 TMUs, and the same 8 ROPs. This is a classic architectural trade-off: Kepler's design relies on more, simpler cores, while Maxwell achieves higher per-core efficiency. The clock speeds are similar, with the 920MX running at 965 MHz base and 993 MHz boost, while the 920M runs at a flat 954 MHz for both base and boost. The 920MX's slightly higher boost clock, combined with its architecture, allows it to reach a pixel rate of 7.944 GPixel/s versus 7.632 GPixel/s for the 920M.

Texture and compute throughput reveal the 920M's theoretical advantage. The 920M posts a texture rate of 30.53 GTexel/s, which is substantially higher than the 920MX's 23.83 GTexel/s. Similarly, FP32 performance is 732.7 GFLOPS for the 920M versus 508.4 GFLOPS for the 920MX. These raw numbers suggest the 920M should be faster, yet the benchmark results show the opposite. This discrepancy points to architectural efficiency, driver optimization, and real-world workload behavior that raw specifications do not capture. Memory configurations are identical: both use 2 GB of DDR3 on a 64-bit bus, delivering 14.40 GB/s of bandwidth at 900 MHz (1800 Mbps effective). The 920MX consumes 16 W TDP, while the 920M draws 33 W, making the 920MX significantly more power-efficient despite its lower theoretical throughput.

Where Each One Wins

The benchmark data shows the 920MX winning in both tested scenarios, but the nature of those wins differs. In OpenCL compute, the 920MX leads by 9.2%, a substantial margin that suggests it is the better choice for general-purpose GPU computing tasks such as data processing, video encoding, or any workload that can offload to the GPU via OpenCL. The Vulkan test shows a smaller 4.8% lead, indicating that graphics rendering and gaming workloads, which increasingly rely on Vulkan, still favor the 920MX but by a narrower margin.

The 920M, despite losing both head-to-head tests, has theoretical advantages in texture fill rate (30.53 GTexel/s vs 23.83 GTexel/s) and FP32 compute (732.7 GFLOPS vs 508.4 GFLOPS). In scenarios where raw shader throughput is the limiting factor and where Kepler's architecture is well-optimized, the 920M could potentially perform relatively better, though the benchmark data does not confirm this. The 920M's higher TDP of 33 W suggests it may sustain clocks better under sustained load in a chassis with adequate cooling, but no benchmark data supports this claim directly.

For users, the practical distinction is clear: the 920MX is the faster part in all measured workloads. The 920M's only advantages are its lower power draw in the case of the 920MX (which is actually an advantage for the 920MX) and its higher theoretical compute throughput, which does not translate to benchmark wins. The 920M's larger die size and higher transistor count per area suggest it may be cheaper to produce or integrate, but that is not a performance consideration.

The Verdict

The data supports a straightforward conclusion: the GeForce 920MX is the superior GPU for any user prioritizing performance. It wins both head-to-head benchmarks, delivers a higher average benchmark score (3528 vs 3287), and achieves this with a much lower TDP of 16 W compared to the 920M's 33 W. The 920MX also carries a higher percentile ranking at 21 versus 20, and its nearest rival cluster includes the RTX 5000 Mobile Ada Generation, whereas the 920M's rivals include integrated graphics like Intel HD Graphics 530.

The 920M is not without merit. Its higher shading unit count (384 vs 256), TMU count (32 vs 24), and FP32 throughput (732.7 GFLOPS vs 508.4 GFLOPS) indicate that it may have been designed for different workload characteristics. However, in the absence of any benchmark where the 920M wins, these specifications remain theoretical. For gaming, general compute, or any application relying on OpenCL or Vulkan, the 920MX is the pick.

Users who already own a laptop with the 920M should not expect parity with the 920MX, as the data shows a consistent 4.8% to 9.2% performance gap. Those choosing between the two should select the 920MX without hesitation, assuming the laptop chassis supports the MXM module form factor, as the 920MX uses an MXM Module slot while the 920M is an IGP. The 920MX's lower power draw also makes it more suitable for thinner, lighter laptops where thermal headroom is limited.

FAQ

Q: Which GPU wins in OpenCL performance?

A: The GeForce 920MX wins with a score of 4068 versus the 920M's 3725, a 9.2% advantage.

Q: Is the GeForce 920M ever faster than the 920MX?

A: No. The head-to-head data shows the 920MX winning both the OpenCL and Vulkan tests, with 2 wins for the 920MX and 0 for the 920M.

Q: What are the power consumption differences?

A: The GeForce 920MX has a TDP of 16 W, while the GeForce 920M has a TDP of 33 W, making the 920MX more power-efficient.

Q: Do both GPUs have the same memory configuration?

A: Yes. Both have 2 GB of DDR3 memory on a 64-bit bus, with 14.40 GB/s bandwidth and 1800 Mbps effective memory speed.

Q: How does the 920MX compare to its nearest rival, the GeForce GT 545?

A: The 920MX's average benchmark score of 3528 is 1.8% behind the GT 545's 3594, a negligible difference.

Q: What architecture does each GPU use?

A: The 920MX uses the Maxwell architecture with the GM108S chip, while the 920M uses Kepler 2.0 with the GK208B chip. Both are on TSMC's 28 nm process.

Specification Differences

| Specification | NVIDIA GeForce 920MX | NVIDIA GeForce 920M |

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

| Chip | GM108S | GK208B |

| Architecture | Maxwell | Kepler 2.0 |

| Process Node | 28 nm | 28 nm |

| Die Size | 77 mm² | 87 mm² |

| Transistor Density | 13.2M / mm² | 11.7M / mm² |

| Base Clock | 965 MHz | 954 MHz |

| Boost Clock | 993 MHz | 954 MHz |

| Shading Units | 256 | 384 |

| TMUs | 24 | 32 |

| Pixel Rate | 7.944 GPixel/s | 7.632 GPixel/s |

| Texture Rate | 23.83 GTexel/s | 30.53 GTexel/s |

| FP32 Performance | 508.4 GFLOPS | 732.7 GFLOPS |

| TDP | 16 W | 33 W |

| Slot Width | MXM Module | IGP |

| Vulkan API | 1.4 | 1.2.175 |

| Release Date | 2016-03-24 | 2015-03-12 |

| Average Benchmark Score | 3528 | 3287 |

| Percentile vs All GPUs | 21 | 20 |