AMD Radeon HD 6850 X2 vs NVIDIA GeForce 940MX Comparison

# The Verdict

The data positions these two mobile graphics processors as direct contemporaries, but the recorded benchmarks show a lopsided outcome. The AMD Radeon HD 6850 X2 delivers a Geekbench OpenCL score of 3977, while the NVIDIA GeForce 940MX reaches 4939. That is a 19.5% margin in favor of the NVIDIA part, and it is the only head-to-head benchmark in the database where NVIDIA comes out ahead. For users whose primary workload is OpenCL compute, the NVIDIA GeForce 940MX is the clear choice from this pairing. The AMD card sits in the 24th percentile of all GPUs in the database, the NVIDIA sits in the 28th percentile; both are below the median, so neither is a top-tier part. However, when the only available comparison is between these two, the data points to the NVIDIA as the faster part. The AMD Radeon HD 6850 X2, with its dual-chip Barts design, does have more shading units (960 versus 512) and more render output units (32 versus 8), but in the recorded OpenCL workload it fails to translate that raw width into a higher score. The 940MX's Maxwell architecture proves more efficient in this specific test, and the 28 nm process node versus the 40 nm node of the AMD part may be part of the explanation.

FAQ

Q: Which GPU has the higher OpenCL benchmark score?

A: The NVIDIA GeForce 940MX scores 4939, the AMD Radeon HD 6850 X2 scores 3977. The delta is 19.5% in favor of the NVIDIA.

Q: How do these parts compare in transistor density?

A: The AMD Radeon HD 6850 X2 packs 1,700 million transistors into a 255 mm² die, for 6.7 million transistors per mm², on TSMC's 40 nm node. The NVIDIA GeForce 940MX has 1,870 million transistors on a 148 mm² die, which is 12.6 million per mm². The NVIDIA part is built on a 28 nm process. That higher density does not produce a higher benchmark score in this CPU because the AMD part uses a dual-chip configuration.

Q3: What is the memory bandwidth difference?

A: The AMD Radeon HD 6850 X2 has a 256 bit memory bus with GDDR5 running at 134.4 GB/s. The NVIDIA GeForce 940MX has a 64 bit bus with GDDR5 and 40.10 GB/s. The AMD part has more than three times the bandwidth (134.4 against 40.10 GB/s), but the NVIDIA part still wins the OpenCL test, which shows that memory bandwidth is not the limiting factor.

Q4: Which part is more power hungry?

A: The AMD Radeon HD 6850 X2 is rated at 254 W TDP, the NVIDIA GeForce 940MX at 23 W. The AMD part is a dual-slot card with two 8-pin connectors and a 600 W PSU suggested; the NVIDIA part is an MXM module without external power connectors and no PSU suggestion recorded. The data records the AMD as a much higher power part.

Q5: Which one supports more modern APIs?

A: The AMD Radeon HD 6850 X2 supports DirectX 11.2 (11_0) and OpenGL 4.4, but no Vulkan. The NVIDIA GeForce 940MX supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.4. NVIDIA is the only one with a Vulkan driver on record.

Q6: What are the release dates?

A: The AMD Radeon HD 6850 X2 came out on 2011-09-18 and is end-of-life, with the Evergreen as its predecessor and Southern Islands as its successor. The NVIDIA GeForce 940MX was released on 2016-06-27, is end-of-life, with GeForce 800M before it and GeForce 10 Mobile after it.

Architecture Differences

The two chips come from completely different design families. The AMD Radeon HD 6850 X2 is based on the Barts chip, which implements the TeraScale 2 architecture and is built on TSMC's 40 nm process. This is the Northern Islands generation, the HD 6800 series. The NVIDIA GeForce 940MX uses the GM107 chip, the Maxwell architecture, at TSMC 28 nm. It belongs to the GeForce 900M generation.

The internal resources reflect the different design objectives. AMD's Barts packs 960 shading units, 48 texture units, and 32 ROPs. The NVIDIA GM107 has 512 shading units, 32 TMUs, and 8 ROPs. The AMD chip is substantially larger in die area (255 mm² versus 148 mm²), and it carries more transistors (1,700 million versus 1,870 million). The transistor density figures are interesting: AMD has 6.7 million per mm² against NVIDIA's 12.6 million per mm². The NVIDIA part is built on a denser node with a smaller die.

The memory subsystem also differs. The AMD card pairs a 256-bit GDDR5 interface with 2 GB of memory. The NVIDIA card has a 64-bit GDDR5 bus with the same 2 GB frame buffer. The bandwidth difference is substantial: 134.4 GB/s on the AMD versus 40.10 GB/s on NVIDIA. Yet, in the recorded OpenCL benchmark, the narrower-bus card posts the higher score, which indicates that the bottleneck lies elsewhere.

Clock behavior is also distinct. The AMD Radeon runs a base clock without listed boost or game clocks, and the memory clock is quoted at 1050 MHz, with 4.2 Gbps effective. The NVIDIA GPU has a base clock of 795 MHz and a boost of 861 MHz; memory is at 1253 MHz, 5 Gbps effective. The NVIDIA part is able to reach higher memory clock, but again the benchmark result favors it.

Specification Differences

| Field | AMD Radeon HD 6850 X2 | NVIDIA GeForce 940MX |

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

| Chip | Barts | GM107 |

| Architecture | TeraScale 2 | Maxwell |

| Generation | Northern Islands (HD 6800) | GeForce 900M |

| Process | 40 nm | 28 nm |

| Foundry | TSMC | TSMC |

| Transistors | 1,700 million | 1,870 million |

| Die size | 255 mm² | 148 mm² |

| Transistor density | 6.7M / mm² | 12.6M / mm² |

| Shading units | 960 | 512 |

| Texture units | 48 | 32 |

| ROPs | 32 | 8 |

| Memory | 2 GB GDDR5 | 2 GB GDDR5 |

| Memory bus | 256-bit | 64-bit |

| Memory bandwidth | 134.4 GB/s | 40.10 GB/s |

| Base clock | not listed | 795 MHz |

| Boost clock | not listed | 861 MHz |

| Memory clock | 1050 MHz, 4.2 Gbps effective | 1253 MHz, 5 Gbps effective |

| Bus interface | PCIe 2.0 x16 | PCIe 3.0 x8 |

| Power connectors | 2x 8-pin | None |

| Slot width | Dual-slot | MXM Module |

| TDP | 254 W | 23 W |

| Suggested PSU | 600 W | Not listed |

| Outputs | 2x DVI, 1x HDMI 1.3a, 1x DisplayPort 1.1 | Portable, device dependent |

| APIs | DirectX 11.2 (11_0), OpenGL 4.4 | DirectX 12 (11_0), OpenGL 4.6, Vulkan 1.4 |

| Status | End-of-life | End-of-life |

| Release date | 2011-09-18 | 2016-06-27 |

| Predecessor | Evergreen | GeForce 800M |

| Successor | Southern Islands | GeForce 10 Mobile |

The table shows two deliberately different design philosophies. The AMD Radeon HD 6850 X2 swallows a huge power budget at 254 W TDP, it is a dual-slot card with dual 8-pin power connectors, and it asks for a 650 W PSU. The NVIDIA GeForce 940MX, by contrast, is a 23 W part, an MXM module, no power connector, and no PSU suggestion. The AMD part belongs to an era when desktop cards were power hungry, the NVIDIA part is a mobile part designed for low power.

Head-to-Head Benchmarks

The only recorded head-to-head benchmark is the Geekbench OpenCL test. The AMD Radeon HD 6850 X2 scores 3977, the NVIDIA GeForce 940MX scores 4939. The delta is 19.5%, and the winner is the NVIDIA GeForce 940MX.

This is the only head-to-head test in the database, and it is the NVIDIA part that wins. The win margin is significant: 19.5% is not a small gap, it is a clear victory for the NVIDIA part. The AMD part has twice the shading units, 1.5 times the memory bandwidth, and 3 times the memory bandwidth, and still it is behind. The benchmark result suggests that the AMD's dual-chip setup is not being used effectively in this OpenCL workload. The record also shows the AMD part is in the 24th percentile across all GPUs, whereas the NVIDIA part is in the 28th percentile. Both are in the lower half overall, so neither is an elite part, but within the comparison, the NVIDIA part has a clear advantage.

The nearest rivals in the database for the AMD card are the NVIDIA Quadro K2000 (score of 3964, 0.3% slower), the NVIDIA GeForce 830M (3957, 0.5% faster), the AMD Radeon R5 M420 (3956, 0.5% slower), and the NVIDIA GeForce GT 745M (3953, 0.6% slower). Those four scores are within 0.6% of the AMD score, so they are a close cluster.

The nearest rivals for the NVIDIA GeForce 940MX: the NVIDIA GeForce GTX 560M (4855, 0.2% lower), the AMD Radeon R7 M260X (4867, 0.5% lower), the NVIDIA GeForce GTS 450 (4893, 1% lower), and the NVIDIA GeForce RTX 3080 12 GB (4791, 1.1% higher). The immediate group is very closely packed around the 940MX, with scores within about 1% of each other.

Where each one wins

If the workload is OpenCL compute, the NVIDIA GeForce 940MX is the part to pick. It wins the only recorded head-to-head test by 19.5%. There is no recorded benchmark test where the AMD Radeon HD 6850 X2 comes out ahead. This is a one-sided comparison in the database.

The AMD Radeon HD 6850 X2, however, has its own area of superiority in the specifications. Its memory subsystem is far larger: 256-bit bus, 134.4 GB/s bandwidth, 2 GB GDDR5. The NVIDIA part only has a 64-bit bus and 40.10 GB/s. For workloads that rely heavily on memory and memory bandwidth, the AMD card has a theoretical advantage, but the recorded OpenCL scores do not reflect this. The AMD part also has many more shaders (960 vs 512), texture units (48 vs 32), and ROPs (32 vs 8). The AMD card is a dual-slot desktop part with two 8-pin connectors; the NVIDIA part is a low-power MXM module, with a TDP of 23 W vs 254 W, it is designed for a different environment.

Choose the NVIDIA GeForce 940MX for the highest recorded OpenCL performance, and if low power consumption matters. Choose the AMD Radeon HD 6850 X2 only if the system can support a 254 W, dual-slot, dual-8-pin card and if the massive memory interface and raw shading resources matter more than the benchmark suite.