NVIDIA GeForce GTX 850M vs NVIDIA GeForce MX350 Comparison

The Verdict

The recorded benchmark data splits the win count evenly, but the nature of those wins tells a clear story. The NVIDIA GeForce GTX 850M takes the OpenCL test with a score of 9821, beating the MX350's 8689 by 11.5%. The MX350 strikes back decisively in Vulkan, scoring 13077 against the GTX 850M's 8782, a massive 48.9% advantage. For modern workloads that leverage Vulkan, the MX350 is the clear choice. For legacy OpenCL compute tasks, the GTX 850M holds a measurable edge. The MX350 also occupies a higher overall position in the database, sitting at the 49th percentile of all GPUs compared to the GTX 850M's 46th percentile. Its average benchmark score of 10883 also outpaces the older card's 9302. Users prioritizing current API performance and overall standing should pick the MX350; those with specific OpenCL-bound applications from the GTX 850M era may find the older card more suitable.

Architecture Differences

The two GPUs come from different architectural generations and foundries. The MX350 uses the GP107S chip built on Pascal architecture, manufactured by Samsung on a 14 nm process. The GTX 850M relies on the GM107 chip with Maxwell architecture, produced by TSMC on a 28 nm node. This process shrink allows the MX350 to pack 3,300 million transistors into a 132 mm² die, yielding a transistor density of 25.0M per mm². The GTX 850M has 1,870 million transistors across a larger 148 mm² die, giving a density of only 12.6M per mm². The MX350's newer node enables roughly double the transistor density, which contributes to its efficiency advantages.

Core counts show a mixed picture. Both GPUs feature 640 shading units and 16 ROPs, but the MX350 has 32 TMUs while the GTX 850M has 40 TMUs. The MX350 compensates with higher clocks: a base of 1354 MHz and boost of 1468 MHz, whereas the GTX 850M's clock data is not recorded. Memory configurations differ substantially. The MX350 uses 2 GB of GDDR5 on a 64-bit bus, delivering 56.06 GB/s of bandwidth. The GTX 850M also has 2 GB but uses DDR3 on a 128-bit bus, resulting in only 32.03 GB/s. Despite the narrower bus, the MX350's faster GDDR5 memory provides 75% more bandwidth. The MX350 also has a much higher memory clock at 1752 MHz (7 Gbps effective) compared to the GTX 850M's 1001 MHz (2 Gbps effective).

Compute rates favor the MX350 in raw throughput. Its pixel rate is 23.49 GPixel/s versus 14.43 GPixel/s for the GTX 850M. Texture rate stands at 46.98 GTexel/s for the MX350 against 36.08 GTexel/s. FP32 performance reaches 1.879 TFLOPS for the MX350, comfortably above the GTX 850M's 1,154.6 GFLOPS. The MX350 also supports FP16 at 29.36 GFLOPS (1:64 ratio), while the GTX 850M has no recorded FP16 capability. Power consumption tells the efficiency story: the MX350 draws only 20 W versus 45 W for the GTX 850M, meaning the newer card delivers higher performance at less than half the power draw.

API support shows subtle differences. Both cards support DirectX 12, OpenGL 4.6, and Vulkan 1.4. However, the MX350 supports DirectX 12 at feature level 12_1, while the GTX 850M is limited to 12 (11_0). The MX350 also uses a PCIe 3.0 x4 interface, whereas the GTX 850M uses the wider PCIe 3.0 x16 interface. The MX350's newer Pascal architecture clearly benefits from the 14 nm process, delivering higher clock speeds, more bandwidth, and better power efficiency across the board.

FAQ

Q: Which GPU has a higher average benchmark score?

A: The MX350 has an average benchmark score of 10883, which is 17% higher than the GTX 850M's 9302. The MX350 also sits at the 49th percentile of all GPUs, three points above the GTX 850M's 46th percentile.

Q: How do the two cards compare in OpenCL performance?

A: The GTX 850M wins the OpenCL test with a score of 9821, beating the MX350's 8689 by 11.5%. This is the older card's only benchmark victory in the head-to-head data.

Q: What about Vulkan performance?

A: The MX350 is dramatically faster in Vulkan, scoring 13077 versus the GTX 850M's 8782. That is a 48.9% advantage, making it the largest performance gap between the two in any recorded test.

Q: How much memory bandwidth does each card have?

A: The MX350 has 56.06 GB/s of bandwidth from its 2 GB GDDR5 memory on a 64-bit bus. The GTX 850M has 32.03 GB/s from 2 GB DDR3 on a 128-bit bus. Despite the narrower bus, the MX350 offers 75% more bandwidth.

Q: Which GPU is more power efficient?

A: The MX350 has a 20 W TDP, less than half the GTX 850M's 45 W TDP. This efficiency comes from the 14 nm Samsung process versus the 28 nm TSMC node used for the GTX 850M.

Q: Do both GPUs support the same modern APIs?

A: Both support DirectX 12, OpenGL 4.6, and Vulkan 1.4. The MX350 supports DirectX 12 at feature level 12_1, while the GTX 850M is capped at feature level 11_0.

Specification Differences

| Specification | NVIDIA GeForce MX350 | NVIDIA GeForce GTX 850M |

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

| Architecture | Pascal | Maxwell |

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

| Transistors | 3,300 million | 1,870 million |

| Die Size | 132 mm² | 148 mm² |

| Transistor Density | 25.0M / mm² | 12.6M / mm² |

| Base Clock | 1354 MHz | Not recorded |

| Boost Clock | 1468 MHz | Not recorded |

| Memory Clock | 1752 MHz (7 Gbps effective) | 1001 MHz (2 Gbps effective) |

| Memory Type | GDDR5 | DDR3 |

| Memory Bus Width | 64 bit | 128 bit |

| Memory Bandwidth | 56.06 GB/s | 32.03 GB/s |

| TMUs | 32 | 40 |

| Pixel Rate | 23.49 GPixel/s | 14.43 GPixel/s |

| Texture Rate | 46.98 GTexel/s | 36.08 GTexel/s |

| FP32 | 1.879 TFLOPS | 1,154.6 GFLOPS |

| FP16 | 29.36 GFLOPS (1:64) | Not recorded |

| TDP | 20 W | 45 W |

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

| Bus Interface | PCIe 3.0 x4 | PCIe 3.0 x16 |

| Release Date | 2020-02-09 | 2014-03-11 |

| Predecessor | Not recorded | GeForce 700M |

| Successor | Not recorded | GeForce 900M |

Head-to-Head Benchmarks

The two recorded benchmarks show a clear split in strengths. Starting with OpenCL, the GTX 850M posts 9821 points against the MX350's 8689. The 11.5% delta favors the older Maxwell card, which outperforms despite its lower clock speeds and older memory technology. This result suggests that OpenCL workloads in the database respond well to the GTX 850M's wider 128-bit memory bus and higher TMU count, even though its raw bandwidth is lower.

The Vulkan test flips the script entirely. The MX350 scores 13077, while the GTX 850M manages only 8782. The 48.9% delta is the largest margin in any comparison involving these two cards. This massive gap likely stems from the MX350's Pascal architecture, which brings architectural improvements that directly benefit Vulkan's modern rendering pipeline. The MX350's higher pixel rate of 23.49 GPixel/s and texture rate of 46.98 GTexel/s give it substantial raw throughput advantages that Vulkan can exploit.

Contextualizing against their nearest rivals clarifies the standing. The MX350's average score of 10883 places it just 0.7% ahead of the AMD Radeon Pro 450 (10804) and 1.1% ahead of the NVIDIA Quadro K2200 (10761). It trails the NVIDIA GeForce GTX 1650 SUPER (11047) by 1.5% and the AMD Radeon RX 550 (11075) by 1.7%. The GTX 850M's average of 9302 sits nearly level with the NVIDIA GeForce GTX 465 (9294, 0.1% ahead) and the AMD Radeon R7 M380 (9313, 0.1% behind). It is 0.3% ahead of the NVIDIA GeForce GTX 960 (9273) and 0.9% ahead of the AMD Radeon Vega 8 (9221). These tight margins show both cards are competitive within their respective performance tiers, but the MX350 operates in a higher tier overall.

The average benchmark scores reinforce the Vulkan result. The MX350's 10883 average is 17% higher than the GTX 850M's 9302. This aggregate metric weights both tests equally, so the MX350's dominant Vulkan win more than compensates for its OpenCL deficit. The percentile ranking confirms the same story: 49th percentile for the MX350 versus 46th for the GTX 850M.

Where Each One Wins

The MX350 wins decisively in any workload that leverages Vulkan. The 48.9% advantage in the Vulkan benchmark is the single largest performance gap recorded between these two cards. This makes the MX350 the better choice for modern games and applications that use Vulkan as their primary graphics or compute API. Its higher pixel rate (23.49 GPixel/s), texture rate (46.98 GTexel/s), and FP32 throughput (1.879 TFLOPS) provide substantial headroom for graphics-intensive tasks. The MX350 also wins on efficiency, with its 20 W TDP enabling deployment in thinner, lighter laptops without active cooling concerns. Its GDDR5 memory at 56.06 GB/s bandwidth, despite the narrower 64-bit bus, delivers more than enough throughput for its compute capabilities. For users running current software titles, the MX350's Pascal architecture and DirectX 12_1 support make it the forward-looking option.

The GTX 850M wins the OpenCL test outright, beating the MX350 by 11.5%. This indicates that certain compute workloads, particularly those that rely on OpenCL's execution model, still favor the older card. The GTX 850M's 128-bit memory bus and 40 TMUs may contribute to this advantage in memory-latency-sensitive compute tasks. Its higher TDP of 45 W also suggests it was designed for larger chassis with more substantial cooling, allowing sustained performance in compute-heavy scenarios. Users running legacy OpenCL applications, scientific compute tools, or older rendering pipelines may see better results with the GTX 850M. However, this advantage is narrow and specific. The GTX 850M's overall benchmark average is lower, its Vulkan performance is dramatically worse, and its DDR3 memory bandwidth of 32.03 GB/s is a significant bottleneck for modern workloads. The GTX 850M also lacks recorded FP16 support, limiting its utility in machine learning or media applications that use half-precision arithmetic.

In summary, the MX350 is the superior all-round performer for modern use cases, with higher average scores, better Vulkan performance, and dramatically lower power consumption. The GTX 850M retains a niche for OpenCL-bound legacy applications, but its age, higher power draw, and weaker API support make it the less compelling choice for most users. The data clearly points to the MX350 for anyone building or buying a laptop today, while the GTX 850M remains a curiosity for specific compute tasks from its era.