NVIDIA GeForce 940MX vs NVIDIA GeForce RTX 4070 GDDR6 Comparison

Head-to-Head Benchmarks

The database records no direct head-to-head benchmark runs between the NVIDIA GeForce 940MX and the NVIDIA GeForce RTX 4070 GDDR6. Instead, the comparison must be assembled from their individual average benchmark scores and the percentile rankings each card holds within the full GPU database. The 940MX shows an average benchmark score of 4844, while the RTX 4070 GDDR6 records an average of 4335. These numbers cannot be compared directly because they come from different test suites: the 940MX was measured with Geekbench OpenCL (4939) and Geekbench Vulkan (4749), whereas the RTX 4070 GDDR6 was measured with 3DMark Steel Nomad DX12 (4334.5).

The 940MX sits at the 28th percentile among all GPUs in the database. Its nearest rivals include the NVIDIA GeForce GTX 560M (average 4855, 0.2% ahead), the AMD Radeon R6 M255DX (average 4867, 0.5% ahead), and the NVIDIA GeForce GTS 450 (average 4893, 1% ahead). It also sits within 1.1% of the NVIDIA GeForce RTX 3080 12 GB (average 4791), which is a striking placement: the database places a mobile Maxwell part from 2016 in the same performance neighborhood as a high-end desktop Ampere card, at least within this particular metric.

The RTX 4070 GDDR6, by contrast, holds the 25th percentile among all GPUs. Its nearest rivals are the Intel Iris Pro Graphics 5200 (average 4360, 0.6% ahead), the AMD FirePro W2100 (average 4295, 0.9% behind), the NVIDIA GeForce 930M (average 4388, 1.2% ahead), and the NVIDIA GeForce GTX 460M (average 4282, 1.2% behind). This grouping is curious. The RTX 4070 GDDR6 is a desktop Ada Lovelace card with 29.15 TFLOPS of FP32 throughput, yet its single recorded 3DMark Steel Nomad result places it among integrated graphics and old mobile parts in percentile terms. The data suggests that the 3DMark Steel Nomad workload and the Geekbench workloads are not comparable across these two cards.

What the head-to-head comparison lacks in direct measurements, it makes up for in architectural contrast. The 940MX delivers 881.7 GFLOPS of FP32 performance, while the RTX 4070 GDDR6 delivers 29.15 TFLOPS, a difference of roughly 33 times. Pixel rate on the 940MX is 6.888 GPixel/s against 158.4 GPixel/s for the RTX 4070 GDDR6, a factor of about 23. Texture rate is 27.55 GTexel/s versus 455.4 GTexel/s, a factor of about 16.5. Memory bandwidth is 40.10 GB/s versus 480.0 GB/s, a factor of 12. The RTX 4070 GDDR6 also has 5888 shading units versus 512, 184 TMUs versus 32, 64 ROPs versus 8, and 12 GB of GDDR6 memory on a 192-bit bus versus 2 GB of GDDR5 on a 64-bit bus.

The Verdict

The data paints two very different pictures depending on the workload. For compute tasks similar to Geekbench OpenCL or Vulkan, the 940MX is the stronger performer, with an average score of 4844 compared to the RTX 4070 GDDR6's 4335. The 940MX also holds a higher percentile rank (28th versus 25th) despite being an end-of-life mobile chip from the GeForce 900M generation. Within the database's own nearest-rival comparisons, the 940MX is competitive with desktop cards like the GTS 450 and even the RTX 3080 12 GB, while the RTX 4070 GDDR6 is grouped with the Iris Pro 5200 and the 930M.

However, the RTX 4070 GDDR6 is clearly the more capable hardware in raw specification terms. It has 11.5 times more shading units, 5.75 times more TMUs, 8 times more ROPs, 6 times more memory, and 12 times more bandwidth. Its FP32 throughput of 29.15 TFLOPS is more than 33 times that of the 940MX. The RTX 4070 GDDR6 also supports DirectX 12 Ultimate (12_2) and includes 46 ray tracing cores and 184 tensor cores, none of which exist on the 940MX, which tops out at DirectX 12 (11_0).

Who should pick which? From the recorded benchmark data alone, the 940MX wins in the one metric where both cards have comparable scores, but that metric is not representative of the RTX 4070 GDDR6's intended workloads. The RTX 4070 GDDR6 was measured only in 3DMark Steel Nomad DX12, a modern DirectX 12 test, and its percentile rank of 25 suggests that test is far more demanding than the older Geekbench workloads. The 940MX cannot run Steel Nomad, and the RTX 4070 GDDR6 cannot run Geekbench, so the database offers no apples-to-apples verdict. The practical answer is that the RTX 4070 GDDR6 is the only choice for any workload that resembles modern gaming, given its ray tracing cores, tensor cores, 12 GB GDDR6 memory, and PCIe 4.0 x16 interface, none of which the 940MX can offer.

The 940MX is the choice only for legacy or embedded use cases. It is an MXM Module card with no power connectors, a 23 W TDP, and portable device dependent display outputs. It was released in 2016, uses 28 nm Maxwell, and has been marked end-of-life. Its Geekbench scores are respectable for its class, but nothing in the data suggests it can approach the RTX 4070 GDDR6 in throughput, memory capacity, or modern API support.

Architecture Differences

The two cards belong to entirely different eras of NVIDIA's architecture timeline. The 940MX uses the GM107 chip, built on TSMC's 28 nm process, with 1,870 million transistors on a 148 mm² die, yielding a transistor density of 12.6M per mm². The RTX 4070 GDDR6 uses the AD104 chip, built on TSMC's 5 nm process, with 35,800 million transistors on a 294 mm² die, yielding a transistor density of 121.8M per mm². That is roughly 19 times more transistors packed into roughly twice the die area. The density difference alone is staggering: 121.8 million transistors per square millimeter is an order of magnitude beyond Maxwell's 12.6 million.

The 940MX is Maxwell architecture, generation GeForce 900M, released June 2016, succeeding the GeForce 800M and succeeded by GeForce 10 Mobile. The RTX 4070 GDDR6 is Ada Lovelace architecture, generation GeForce 46 ray tracing cores and 184 tensor cores.

The 940MX has no ray tracing cores and no tensor cores, no FP16 hardware path recorded, and DirectX 12 (11_0) support, while the RTX 4070 GDDR6 has 46 ray tracing cores, 184 tensor cores, FP16 at 29.15 TFLOPS at 1:1 rate, DirectX 12 Ultimate (12_2), and FP32 29.15 TFLOPS.

The 940MX uses a PCIe 3.0 x8 bus interface, while the RTX 4070 GDDR6 uses PCIe 4.0 x16. The 940MX is an MXM Module with no power connectors and a 23 W TDP, the RTX 4070 GDDR6 is a dual-slot card with a 1x 16-pin power connector and a 200 W TDP, with a suggested PSU of 550 W. The RTX 4070 GDDR6 measures 240 mm by 110 mm by 40 mm, the 940MX has no recorded dimensions.

Specification Differences

The table below lists only the fields where the two cards differ, drawing directly from the database records.

| Field | 940MX | RTX 4070 GDDR6 |

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

| Architecture | Maxwell | Ada Lovelace |

| Generation | GeForce 900M | GeForce 40 |

| Process node | 28 nm | 5 nm |

| Transistors | 1,870 million | 35,800 million |

| Die size | 148 mm² | 294 mm² |

| Transistor density | 12.6M / mm² | 121.8M / mm² |

| Base clock | 795 MHz | 1920 MHz |

| Boost clock | 861 MHz | 2475 MHz |

| Memory clock | 1253 MHz, 5 Gbps effective | 2500 MHz, 20 Gbps effective |

| Memory size | 2 GB | 12 GB |

| Memory type | GDDR5 | GDDR6 |

| Memory bus | 64 bit | 192 bit |

| Memory bandwidth | 40.10 GB/s | 480.0 GB/s |

| Shading units | 512 | 5888 |

| TMUs | 32 | 184 |

| ROPs | 8 | 64 |

| RT cores | None | 46 |

| Tensor cores | None | 184 |

| Pixel rate | 6.888 GPixel/s | 158.4 GPixel/s |

| Texture rate | 27.55 GTexel/s | 455.4 GTexel/s |

| FP32 | 881.7 GFLOPS | 29.15 TFLOPS |

| FP16 | Not recorded | 29.15 TFLOPS (1:1) |

| TDP | 23 W | 200 W |

| Slot width | MXM Module | Dual-slot |

| Power connectors | None | 1x 16-pin |

| Suggested PSU | Not recorded | 550 W |

| Bus interface | PCIe 3.0 x8 | PCIe 4.0 x16 |

| Display outputs | Portable Device Dependent | 1x HDMI 2.1, 3x DisplayPort 1.4a |

| DirectX | 12 (11_0) | 12 Ultimate (12_2) |

| Release date | 2016-06-27 | 2024-08-19 |

| Predecessor | GeForce 800M | GeForce 30 |

| Successor | GeForce 10 Mobile | GeForce 50 |

Both cards share the same manufacturer (NVIDIA) and the same OpenGL 4.6 and Vulkan 1.4 API support. The launch MSRP of the RTX 4070 GDDR6 is 599 USD; the 940MX has no recorded launch MSRP.

FAQ

Q: Which card has a higher average benchmark score in the database?

A: The 940MX has an average benchmark score of 4844, while the RTX 4070 GDDR6 has an average of 4335. However, these averages come from different test suites, so they are not directly comparable.

Q: What is the memory bandwidth difference between the two cards?

A: The 940MX has 40.10 GB/s of bandwidth from 2 GB of GDDR5 on a 64-bit bus, while the RTX 4070 GDDR6 has 480.0 GB/s from 12 GB of GDDR6 on a 192-bit bus. That is roughly 12 times more bandwidth.

Q: Does the 940MX support ray tracing or tensor cores?

A: No. The 940MX has no ray tracing cores and no tensor cores. The RTX 4070 GDDR6 has 46 ray tracing cores and 184 tensor cores.

Q: What is the FP32 throughput of each card?

A: The 940MX delivers 881.7 GFLOPS of FP32 performance. The RTX 4070 GDDR6 delivers 29.15 TFLOPS, which is more than 33 times higher.

Q: Are there any benchmark tests where both cards were run?

A: No. The database records no head-to-head benchmark runs between the two cards. The 940MX was tested with Geekbench OpenCL and Vulkan, while the RTX 4070 GDDR6 was tested with 3DMark Steel Nomad DX12.

Q: What is the TDP of each card?

A: The 940MX has a TDP of 23 W and requires no power connectors. The RTX 4070 GDDR6 has a TDP of 200 W, uses a 1x 16-pin power connector, and has a suggested PSU of 550 W.

Where Each One Wins

The 940MX wins in the Geekbench OpenCL and Vulkan workloads recorded in the database. Its Geekbench OpenCL score is 4939 and its Vulkan score is 4749, giving it an average of 4844. It holds the 28th percentile among all GPUs, placing it ahead of the RTX 4070 GDDR6's 25th percentile. In the nearest-rival comparisons, the 940MX is within 0.2% of the GTX 560M and within 1.1% of the RTX 3080 12 GB, which suggests the database considers it a mid-pack performer in those legacy compute tests.

The RTX 4070 GDDR6 wins in every raw architectural specification. It has more shading units, more TMUs, more ROPs, more memory, more bandwidth, higher clocks, higher pixel rate, higher texture rate, and higher FP32 throughput. It also has features the 940MX completely lacks: ray tracing cores, tensor cores, FP16 support, and DirectX 12 Ultimate. Its single recorded benchmark, 3DMark Steel Nomad DX12, produced a score of 4334.5, and its nearest rivals are all integrated or older mobile GPUs, which suggests that Steel Nomad is a workload that stresses modern desktop hardware far more than the Geekbench tests stress the 940MX.

The practical use-case split is clear. The 940MX is for legacy systems, embedded MXM applications, or any scenario where 23 W power draw, no external power connectors, and portable device dependent display outputs are requirements. The RTX 4070 GDDR6 is for modern desktop gaming and compute workloads that can use its 46 ray tracing cores, 184 tensor cores, 12 GB of GDDR6 memory, and PCIe 4.0 x16 bandwidth. The database records no scenario where the 940MX outperforms the RTX 4070 GDDR6 in a shared workload, and the RTX 4070 GDDR6 outperforms the 940MX in every specification field where they can be compared. The only reason to prefer the 940MX is the specific benchmark context of Geekbench, where it scores higher, but that context does not extend to the RTX 4070 GDDR6's intended modern workloads.