AMD Radeon 890M vs NVIDIA GeForce GTX 870M Comparison

NVIDIA’s GeForce GTX 870M is a mobile Kepler part from 2014, while AMD’s Radeon 890M is a 2024 RDNA 3.5 integrated GPU. The benchmark data shows a decisive generational shift, with the newer AMD chip dominating in raw compute while the older NVIDIA part holds its own in legacy API scenarios. Below is a breakdown based solely on the supplied facts.

Head-to-Head Benchmarks

The only shared benchmark between the two GPUs is Geekbench OpenCL, and the result is lopsided. The AMD Radeon 890M scores 37,254 points, compared to the NVIDIA GeForce GTX 870M’s 12,630 points. That is a deltaPct of -66.1% for the NVIDIA part, meaning the AMD chip delivers roughly 2.95 times the OpenCL performance. This is the single largest win for either product in the head-to-head data, and it reflects the architectural gulf between a 2014 discrete mobile GPU and a 2024 integrated solution.

The GTX 870M’s average benchmark score across all tests is 9,959, while the Radeon 890M averages 9,210. Interestingly, the older NVIDIA card actually has a higher average score, despite losing the OpenCL test so decisively. This is because the GTX 870M’s only other benchmark, Geekbench Metal, yields 7,288 points, and its two-score average is pulled up by the OpenCL result. The Radeon 890M, by contrast, has a wider range of tests: its highest is 40,808 in Geekbench Vulkan, and its lowest is 33 in Passmark DirectX 10, dragging the average down.

In terms of overall percentile ranking, the GTX 870M sits at the 48th percentile of all GPUs, while the Radeon 890M is at the 45th. Neither is a top-tier part by modern standards, but the data shows the AMD chip’s raw compute capability is far higher. The Radeon 890M’s FP32 throughput is 5.939 TFLOPS, more than double the GTX 870M’s 2.599 TFLOPS. The texture rate is similarly higher: 185.6 GTexel/s versus 108.3 GTexel/s, and the pixel rate is 92.80 GPixel/s against 27.08 GPixel/s.

Where Each One Wins

The AMD Radeon 890M wins clearly in modern compute-heavy workloads. Its OpenCL score of 37,254 is its strongest result in the provided data, and it also posts a Geekbench Vulkan score of 40,808, which is the highest single benchmark number for either GPU. The Vulkan result suggests strong performance in modern graphics APIs, and the 890M’s support for DirectX 12 Ultimate (12_2) and Vulkan 1.4 means it is built for current and near-future titles. The 890M also has 16 ray tracing cores, which the GTX 870M lacks entirely, giving it a hardware advantage in ray-traced scenes, provided the software uses those features.

The NVIDIA GeForce GTX 870M, despite being a decade older, has a few areas where it does not fall as far behind as the OpenCL gap implies. Its average benchmark score is higher, which suggests it performs more consistently across the tests it does run. The GTX 870M also has a higher percentile ranking (48 vs 45), meaning it beats a slightly larger share of the GPU population in the database. That said, the GTX 870M’s only other benchmark, Geekbench Metal (7,288), is far lower than any of the Radeon 890M’s modern API scores. The GTX 870M’s advantage is essentially about consistency in a limited test set, not about raw speed.

For legacy DirectX 9 and 10 workloads, the Radeon 890M actually posts very low scores: 97 in Passmark DirectX 9 and 33 in Passmark DirectX 10. These are the two lowest results in either product’s benchmark list. The GTX 870M does not have corresponding Passmark scores in the data, so a direct comparison is impossible, but the low numbers suggest the AMD chip may not be optimized for older APIs. The GTX 870M’s DirectX support is only 12 (11_0), so it cannot run DirectX 12 Ultimate features, but its older architecture may handle legacy code with fewer translation overheads.

Architecture Differences

The two GPUs are separated by more than a decade of silicon evolution. The NVIDIA GeForce GTX 870M uses the GK104 chip on a 28 nm process from TSMC, with 3,540 million transistors on a 294 mm² die. That yields a transistor density of 12.0 million per square millimeter. The AMD Radeon 890M uses the Strix Point chip on a 4 nm process, also from TSMC, packing 34,000 million transistors into a 233 mm² die, for a density of 145.9 million per square millimeter. That is a 12x improvement in density, which is the core reason the 890M can deliver over twice the FP32 throughput while drawing far less power.

The GTX 870M is built on Kepler architecture, with 1,344 shading units, 112 texture mapping units, and 24 ROPs. It has no ray tracing cores and no tensor cores. The Radeon 890M uses RDNA 3.5, with 1,024 shading units, 64 TMUs, and 32 ROPs, plus 16 ray tracing cores. The AMD part has fewer shading units but a much higher boost clock — 2,900 MHz versus 967 MHz for the NVIDIA — which explains the higher FP32 and texture rates. The base clocks are also telling: the GTX 870M runs at 941 MHz base, while the Radeon 890M’s base is just 400 MHz, but its boost is far higher, indicating aggressive power management in the integrated design.

Memory architecture is fundamentally different. The GTX 870M has 3 GB of dedicated GDDR5 on a 192-bit bus, delivering 120.0 GB/s of bandwidth. The Radeon 890M uses system-shared memory, with the type, bus width, and bandwidth all listed as "System Shared" or "System Dependent." This means the AMD part’s memory performance depends entirely on the host system’s RAM, so it cannot match a dedicated VRAM solution in bandwidth consistency. However, the 890M’s pixel rate of 92.80 GPixel/s is more than triple the GTX 870M’s 27.08 GPixel/s, which matters for fill-rate-bound scenarios.

Power consumption is starkly different: the GTX 870M has a 100 W TDP and uses an MXM module form factor, while the Radeon 890M has a 15 W TDP and is an integrated GPU (IGP) on a PCIe 4.0 x8 interface. The GTX 870M is a discrete card that requires a slot, while the 890M is built into a processor. The process node difference (28 nm vs 4 nm) is the primary driver of the efficiency gap.

FAQ

Q: Which GPU has the higher raw compute throughput?

A: The AMD Radeon 890M, with 5.939 TFLOPS FP32, compared to the NVIDIA GeForce GTX 870M’s 2.599 TFLOPS. The Radeon also has a higher FP16 throughput of 5.939 TFLOPS (1:1), while the GTX 870M has no listed FP16 capability.

Q: How do their benchmark averages compare?

A: The GTX 870M has a higher average benchmark score of 9,959, versus the Radeon 890M’s 9,210. However, the Radeon’s average is dragged down by very low Passmark DirectX 9 and 10 scores (97 and 33, respectively), while its modern API scores (OpenCL 37,254, Vulkan 40,808) are far higher.

Q: Does the Radeon 890M support ray tracing?

A: Yes, it has 16 ray tracing cores. The GTX 870M has no ray tracing cores at all.

Q: What is the memory configuration for each GPU?

A: The GTX 870M has 3 GB of dedicated GDDR5 on a 192-bit bus with 120.0 GB/s bandwidth. The Radeon 890M uses system-shared memory, with bandwidth described as "System Dependent."

Q: Which GPU is more power-efficient?

A: The Radeon 890M has a 15 W TDP, while the GTX 870M has a 100 W TDP. The Radeon draws a fraction of the power while delivering higher raw throughput, thanks to its 4 nm process versus 28 nm.

Q: What are the API support differences?

A: The Radeon 890M supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The GTX 870M supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The Radeon is newer and supports more advanced features.

Specification Differences

| Specification | NVIDIA GeForce GTX 870M | AMD Radeon 890M |

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

| Architecture | Kepler | RDNA 3.5 |

| Process Node | 28 nm | 4 nm |

| Transistors | 3,540 million | 34,000 million |

| Die Size | 294 mm² | 233 mm² |

| Transistor Density | 12.0M / mm² | 145.9M / mm² |

| Base Clock | 941 MHz | 400 MHz |

| Boost Clock | 967 MHz | 2900 MHz |

| Memory Size | 3 GB | System Shared |

| Memory Type | GDDR5 | System Shared |

| Memory Bus Width | 192 bit | System Shared |

| Memory Bandwidth | 120.0 GB/s | System Dependent |

| Shading Units | 1344 | 1024 |

| TMUs | 112 | 64 |

| ROPs | 24 | 32 |

| Ray Tracing Cores | None | 16 |

| Pixel Rate | 27.08 GPixel/s | 92.80 GPixel/s |

| Texture Rate | 108.3 GTexel/s | 185.6 GTexel/s |

| FP32 | 2.599 TFLOPS | 5.939 TFLOPS |

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

| TDP | 100 W | 15 W |

| Slot Width | MXM Module | IGP |

| Bus Interface | MXM-B (3.0) | PCIe 4.0 x8 |

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

| Vulkan Support | 1.2.175 | 1.4 |

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

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

| Predecessor | GeForce 700M | Navi II IGP |

| Successor | GeForce 900M | None |