AMD Radeon 890M vs NVIDIA GeForce GTX 850M Comparison

The Verdict

The benchmark data presents a stark generational divide. The AMD Radeon 890M is the unequivocal performance winner, dominating the NVIDIA GeForce GTX 850M in every head-to-head comparison. In the Geekbench OpenCL test, the 890M scores 37,254 against the 850M’s 9,821, a 73.6% margin. The Vulkan gap is even wider: 40,808 versus 8,782, a 78.5% difference. There is no scenario in this data where the older NVIDIA part comes out ahead.

For a user deciding between these two, the choice hinges on context. The 890M, an active product from 2024, offers roughly four times the raw compute throughput and modern API support. It is the clear choice for anyone running current workloads or games that leverage DirectX 12 Ultimate. The GTX 850M, an end-of-life part from 2014, is strictly a legacy option. Its only advantage is its lower power draw in a specific sense: the 890M’s TDP is 15 W, while the 850M is rated at 45 W, meaning the newer AMD part delivers vastly higher performance at one-third the thermal envelope. The data suggests the 890M is the superior mobile GPU in every measurable way, but the 850M remains a functional entry-level part for very old software. Pick the 890M for any modern use case; pick the 850M only if you are constrained to a system that predates the 890M’s platform.

Architecture Differences

The two GPUs come from different eras and design philosophies. The NVIDIA GeForce GTX 850M is built on the Maxwell architecture using the GM107 chip, fabricated on a 28 nm process at TSMC. It packs 1,870 million transistors into a 148 mm² die, yielding a transistor density of 12.6 million per square millimeter. The AMD Radeon 890M uses the RDNA 3.5 architecture with the Strix Point chip, manufactured on a 4 nm process, also at TSMC. This newer part integrates 34,000 million transistors across a 233 mm² die, achieving a density of 145.9 million per square millimeter—over eleven times denser than the Maxwell chip.

Core configurations differ substantially. The 850M has 640 shading units, 40 texture mapping units, and 16 raster output pipelines. The 890M more than doubles this: 1,024 shading units, 64 TMUs, and 32 ROPs. Critically, the 890M adds 16 ray tracing cores, a feature entirely absent from the Maxwell-based 850M. Neither GPU includes tensor cores. Clock behavior also diverges: the 850M lists no base or boost clocks in the data, while the 890M has a 400 MHz base and a 2,900 MHz boost. Memory architectures are fundamentally different—the 850M uses 2 GB of dedicated DDR3 on a 128-bit bus, while the 890M relies on system-shared memory with bandwidth described as system dependent.

The feature set reflects the generational leap. The 850M supports DirectX 12 (11_0), while the 890M supports DirectX 12 Ultimate (12_2). Both offer OpenGL 4.6 and Vulkan 1.4. The 890M’s pixel rate of 92.80 GPixel/s and texture rate of 185.6 GTexel/s dwarf the 850M’s 14.43 GPixel/s and 36.08 GTexel/s. Floating-point performance tells the same story: the 890M delivers 5.939 TFLOPS in FP32 and FP16 (at a 1:1 ratio), versus the 850M’s 1,154.6 GFLOPS in FP32 with no listed FP16 support. The 890M also uses PCIe 4.0 x8, while the 850M uses PCIe 3.0 x16.

Head-to-Head Benchmarks

The two direct comparisons available in the data both go decisively to the AMD Radeon 890M. In Geekbench OpenCL, the 890M scores 37,254 against the 850M’s 9,821. That is a 73.6% delta in favor of the newer chip, meaning the 890M is roughly 3.8 times faster in this compute workload. The margin is so large that it suggests not just a clock advantage but a fundamental architectural efficiency gain—the 890M achieves this with a 15 W TDP versus the 850M’s 45 W.

The Geekbench Vulkan test shows an even larger relative gap. The 890M posts 40,808, while the 850M manages only 8,782. The delta is 78.5%, representing a 4.6-fold performance difference. This is particularly notable because Vulkan is a lower-level API that can expose hardware capabilities more directly. The 890M’s modern RDNA 3.5 architecture with its 16 ray tracing cores and DirectX 12 Ultimate feature set is clearly better equipped to handle modern rendering paths. The 850M’s Maxwell architecture, designed for DirectX 11-era workloads, struggles in this comparison.

There are no wins for the NVIDIA part in the head-to-head data. The 850M’s average benchmark score across all tests is 9,302, while the 890M’s is 9,210. Interestingly, despite the 890M’s dominant OpenCL and Vulkan wins, its average score is slightly lower—a 1% difference in favor of the 850M. This is because the 890M’s benchmark suite includes PassMark tests where it scores lower (e.g., 33 in DirectX 10, 73 in DirectX 11, 37 in DirectX 12, 97 in DirectX 9), which drags down its average. The 850M only has two Geekbench scores, both relatively high for its generation.

Specification Differences

The table below highlights only the fields where the two GPUs differ, based strictly on the data provided.

| Specification | NVIDIA GeForce GTX 850M | AMD Radeon 890M |

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

| Architecture | Maxwell | RDNA 3.5 |

| Process Node | 28 nm | 4 nm |

| Transistors | 1,870 million | 34,000 million |

| Die Size | 148 mm² | 233 mm² |

| Transistor Density | 12.6M / mm² | 145.9M / mm² |

| Base Clock | None listed | 400 MHz |

| Boost Clock | None listed | 2,900 MHz |

| Memory Size | 2 GB | System Shared |

| Memory Type | DDR3 | System Shared |

| Memory Bus Width | 128 bit | System Shared |

| Memory Bandwidth | 32.03 GB/s | System Dependent |

| Memory Clock | 1001 MHz (2 Gbps effective) | System Shared |

| Shading Units | 640 | 1,024 |

| TMUs | 40 | 64 |

| ROPs | 16 | 32 |

| RT Cores | None | 16 |

| Pixel Rate | 14.43 GPixel/s | 92.80 GPixel/s |

| Texture Rate | 36.08 GTexel/s | 185.6 GTexel/s |

| FP32 Performance | 1,154.6 GFLOPS | 5.939 TFLOPS |

| FP16 Performance | None listed | 5.939 TFLOPS (1:1) |

| TDP | 45 W | 15 W |

| Bus Interface | PCIe 3.0 x16 | PCIe 4.0 x8 |

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

| Production Status | End-of-life | Active |

| Release Date | 2014-03-11 | 2024-07-14 |

| Predecessor | GeForce 700M | Navi II IGP |

| Successor | GeForce 900M | None |

| Avg Benchmark Score | 9,302 | 9,210 |

| Percentile vs All GPUs | 46 | 45 |

FAQ

Q: Which GPU is faster in raw compute performance?

A: The AMD Radeon 890M is decisively faster. In Geekbench OpenCL, it scores 37,254 compared to the GTX 850M’s 9,821, a 73.6% advantage. In Geekbench Vulkan, the 890M scores 40,808 versus 8,782, a 78.5% advantage.

Q: How does the power consumption compare between the two?

A: The AMD Radeon 890M has a TDP of 15 W, while the NVIDIA GeForce GTX 850M is rated at 45 W. This means the 890M delivers its vastly higher performance while drawing one-third the power of the older part.

Q: Do these GPUs support ray tracing?

A: Yes, but only the AMD Radeon 890M. It includes 16 ray tracing cores. The NVIDIA GeForce GTX 850M has no ray tracing cores listed in its specifications.

Q: Which GPU has better memory bandwidth?

A: The GTX 850M has a dedicated 2 GB DDR3 memory pool on a 128-bit bus, providing 32.03 GB/s bandwidth. The 890M uses system-shared memory with bandwidth described as "System Dependent," so its effective bandwidth varies by platform.

Q: What are the DirectX capabilities of each GPU?

A: The AMD Radeon 890M supports DirectX 12 Ultimate (12_2), while the NVIDIA GeForce GTX 850M supports DirectX 12 (11_0). Both support OpenGL 4.6 and Vulkan 1.4.

Q: How do their average benchmark scores compare?

A: The GTX 850M has an average benchmark score of 9,302, while the 890M averages 9,210—a 1% difference in favor of the NVIDIA part. The 890M’s average is pulled down by its PassMark DirectX tests, despite its dominant Geekbench results.