AMD Radeon R9 M275X vs NVIDIA GeForce GTX 870M Comparison

Head-to-Head Benchmarks

The benchmark data contains a single head-to-head comparison between the AMD Radeon R9 M275X and the NVIDIA GeForce GTX 870M, and it decisively favors the NVIDIA part. In the Geekbench OpenCL test, the GTX 870M scores 12,630 against the R9 M275X's 11,200. That is a deltaPct of -11.3% from the AMD card's perspective, meaning the NVIDIA GPU is roughly eleven percent faster in this specific workload. The margin is not enormous, but it is consistent and measurable. For a mobile GPU pairing, an 11.3% swing in compute performance can translate into visibly smoother frame pacing in GPU-bound tasks.

Looking beyond the direct head-to-head, the average benchmark scores reinforce the same ordering, though with a twist. The R9 M275X averages 10,582 across its benchmark results, while the GTX 870M averages 9,959. That puts the AMD card ahead by about 6% in overall average score, despite losing the single OpenCL test. How can that be? The R9 M275X has two benchmark entries in its FACT PACK: a Geekbench OpenCL score of 11,200 and a Geekbench Vulkan score of 9,964. The GTX 870M has a Geekbench Metal score of 7,288 and a Geekbench OpenCL score of 12,630. Averaging those out yields 10,582 for AMD and 9,959 for NVIDIA. So the NVIDIA GPU wins the OpenCL contest handily, but the AMD part posts a much stronger secondary score in Vulkan than the NVIDIA part does in Metal.

The percentile rankings are close: the R9 M275X sits at the 49th percentile of all GPUs, while the GTX 870M sits at the 48th. Both are mid-pack mobile parts, not flagships. The nearest rivals for the R9 M275X include the AMD Radeon RX 6600S (average score 10,629, deltaPct -0.4%), the AMD Radeon RX 550X (10,481, deltaPct +1%), the NVIDIA GeForce GTX 560 Ti (10,690, deltaPct -1%), and the NVIDIA Quadro K2200 (10,761, deltaPct -1.7%). For the GTX 870M, the nearest rivals are the AMD Radeon Pro 5300M (10,013, deltaPct -0.5%), the NVIDIA Quadro K5100M (10,043, deltaPct -0.8%), the AMD Radeon R9 M375 (10,070, deltaPct -1.1%), and the NVIDIA Quadro 6000 (9,846, deltaPct +1.1%). In both cases, the deltas to nearest rivals are within roughly two percent, indicating that neither card is an outlier in its performance class.

Architecture Differences

The architectural divide between these two mobile GPUs is stark. The AMD Radeon R9 M275X uses the Venus chip built on GCN 1.0 architecture, fabricated on a 28 nm TSMC process. The NVIDIA GeForce GTX 870M uses the GK104 chip with Kepler architecture, also on a 28 nm TSMC process. Both are end-of-life products, but their internal designs could not be more different in scale.

The transistor counts tell the story of their respective design philosophies. The NVIDIA chip packs 3,540 million transistors onto a 294 mm² die, yielding a transistor density of 12.0 million per square millimeter. The AMD chip is far smaller: 1,500 million transistors on a 123 mm² die, with a density of 12.2 million per square millimeter. The density figures are nearly identical, which makes sense given the shared 28 nm node. But the NVIDIA die is more than twice the physical size and holds more than twice the transistors. That larger budget funds a substantially wider execution engine: the GTX 870M has 1,344 shading units, 112 texture mapping units, and 24 ROPs, versus the R9 M275X's 640 shaders, 40 TMUs, and 16 ROPs.

Clock speeds are modest on both sides, with the NVIDIA part running slightly higher: a 941 MHz base and 967 MHz boost for the GTX 870M, against 900 MHz base and 925 MHz boost for the R9 M275X. The memory subsystems differ as well. The AMD card uses 2 GB of GDDR5 on a 128-bit bus, with a 1,125 MHz memory clock (4.5 Gbps effective) and 72.00 GB/s of bandwidth. The NVIDIA card has 3 GB of GDDR5 on a 192-bit bus, with a 1,250 MHz memory clock (5 Gbps effective) and 120.0 GB/s of bandwidth. That 67% bandwidth advantage for NVIDIA is significant for texture-heavy workloads.

The compute throughput figures reflect the raw shader count differences. The GTX 870M delivers 2.599 TFLOPS of FP32 performance, while the R9 M275X delivers 1,184.0 GFLOPS. The NVIDIA card also leads in pixel rate (27.08 GPixel/s versus 14.80 GPixel/s) and texture rate (108.3 GTexel/s versus 37.00 GTexel/s). API support is broadly similar: both support DirectX 12 (with AMD at 11_1 and NVIDIA at 11_0), OpenGL 4.6, and Vulkan (1.2.170 for AMD, 1.2.175 for NVIDIA). The NVIDIA card also has a Metal benchmark score, while the AMD card does not list Metal support.

The Verdict

The data points to a clear winner for raw compute and memory throughput: the NVIDIA GeForce GTX 870M. Its OpenCL score of 12,630 versus 11,200 for the AMD Radeon R9 M275X is a concrete, 11.3% advantage. The NVIDIA card also has more than double the shaders, nearly triple the texture units, 50% more ROPs, and 67% more memory bandwidth. For any workload that scales with those resources, the GTX 870M is the better choice.

However, the average benchmark score tells a more nuanced story. The R9 M275X averages 10,582 across its tests, which is higher than the GTX 870M's 9,959 average. That is driven by the AMD card's respectable Vulkan score of 9,964, compared to the NVIDIA card's weak Metal score of 7,288. If your software stack favors Vulkan over OpenCL or Metal, the AMD GPU may actually deliver better real-world performance than its single OpenCL loss would suggest.

For a buyer choosing between these two end-of-life mobile GPUs, the decision hinges on workload. If the priority is OpenCL compute, high-bandwidth memory access, or maximum shader throughput, the GTX 870M is the stronger pick. If the priority is Vulkan performance, the R9 M275X holds its own and even edges ahead in aggregate. The percentile rankings (49th for AMD, 48th for NVIDIA) are too close to separate them. The GTX 870M wins the head-to-head, but the R9 M275X is not a pushover.

Specification Differences

| Specification | AMD Radeon R9 M275X | NVIDIA GeForce GTX 870M |

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

| Chip | Venus | GK104 |

| Architecture | GCN 1.0 | Kepler |

| Generation | Gem System (R9 M200) | GeForce 800M |

| Transistors | 1,500 million | 3,540 million |

| Die Size | 123 mm² | 294 mm² |

| Transistor Density | 12.2M / mm² | 12.0M / mm² |

| Base Clock | 900 MHz | 941 MHz |

| Boost Clock | 925 MHz | 967 MHz |

| Memory Clock | 1125 MHz (4.5 Gbps effective) | 1250 MHz (5 Gbps effective) |

| Memory Size | 2 GB | 3 GB |

| Memory Bus Width | 128 bit | 192 bit |

| Memory Bandwidth | 72.00 GB/s | 120.0 GB/s |

| Shading Units | 640 | 1344 |

| TMUs | 40 | 112 |

| ROPs | 16 | 24 |

| Pixel Rate | 14.80 GPixel/s | 27.08 GPixel/s |

| Texture Rate | 37.00 GTexel/s | 108.3 GTexel/s |

| FP32 | 1,184.0 GFLOPS | 2.599 TFLOPS |

| TDP | Not specified | 100 W |

| Slot Width | Not specified | MXM Module |

| Power Connectors | Not specified | None |

| Bus Interface | PCIe 3.0 x16 | MXM-B (3.0) |

| Display Outputs | Not specified | Portable Device Dependent |

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

| OpenGL | 4.6 | 4.6 |

| Vulkan | 1.2.170 | 1.2.175 |

| Release Date | 2014-01-27 | 2014-03-11 |

| Predecessor | Solar System | GeForce 700M |

| Successor | Polaris Mobile | GeForce 900M |

The two cards share a 28 nm TSMC process and both are end-of-life, but nearly every other specification diverges. The NVIDIA GPU is a bigger, wider, faster chip in terms of raw resources, but it also draws more power (100 W TDP is specified for NVIDIA; no TDP is listed for AMD). The bus interface differs as well: AMD uses a standard PCIe 3.0 x16 slot, while NVIDIA uses an MXM-B (3.0) module form factor.

FAQ

Q: Which GPU has the higher OpenCL benchmark score?

A: The NVIDIA GeForce GTX 870M scores 12,630 in Geekbench OpenCL, while the AMD Radeon R9 M275X scores 11,200. The NVIDIA card wins by 11.3%.

Q: How do the average benchmark scores compare?

A: The AMD Radeon R9 M275X has an average benchmark score of 10,582, which is higher than the NVIDIA GeForce GTX 870M's average of 9,959.

Q: What is the memory bandwidth difference?

A: The NVIDIA GeForce GTX 870M has 120.0 GB/s of memory bandwidth, while the AMD Radeon R9 M275X has 72.00 GB/s. The NVIDIA card offers roughly 67% more bandwidth.

Q: Which GPU has more shading units?

A: The NVIDIA GeForce GTX 870M has 1,344 shading units, compared to 640 on the AMD Radeon R9 M275X. That is more than double the shader count.

Q: What are the transistor counts for each chip?

A: The NVIDIA GeForce GTX 870M uses a chip with 3,540 million transistors, while the AMD Radeon R9 M275X uses a chip with 1,500 million transistors. The NVIDIA die is also larger at 294 mm² versus 123 mm².

Q: Do both GPUs support DirectX 12?

A: Yes, both support DirectX 12. The AMD Radeon R9 M275X supports DirectX 12 (11_1), and the NVIDIA GeForce GTX 870M supports DirectX 12 (11_0).

Where Each One Wins

The NVIDIA GeForce GTX 870M wins in every direct performance comparison in the FACT PACK. It has the higher OpenCL score (12,630 vs 11,200), the higher FP32 throughput (2.599 TFLOPS vs 1,184.0 GFLOPS), the higher pixel rate (27.08 GPixel/s vs 14.80 GPixel/s), the higher texture rate (108.3 GTexel/s vs 37.00 GTexel/s), and the higher memory bandwidth (120.0 GB/s vs 72.00 GB/s). For any workload that stresses raw compute, texture filtering, or memory bandwidth, the GTX 870M is the correct choice.

The AMD Radeon R9 M275X wins in the aggregate benchmark average (10,582 vs 9,959) and in the Vulkan benchmark (9,964 vs no Vulkan score listed for NVIDIA). If your application uses the Vulkan API, the AMD card is the stronger performer. It also has a slightly higher transistor density (12.2M / mm² vs 12.0M / mm²), though that is a minor distinction. The AMD card's smaller die (123 mm² vs 294 mm²) and lower transistor count suggest it may be more efficient per square millimeter, but no TDP data is available for the AMD card to confirm a power advantage.

For gaming specifically, the GTX 870M's superior memory bandwidth and shader count make it the better bet for high-resolution textures and complex shaders. The R9 M275X's Vulkan score suggests it could hold its own in Vulkan-based titles, but the single head-to-head OpenCL test favors NVIDIA. The GTX 870M also has a Metal score of 7,288, which is lower than its OpenCL score but still provides a data point for macOS or Metal-based workloads. The R9 M275X has no Metal score listed.

In practical terms, the GTX 870M is the GPU for compute-heavy tasks and high-bandwidth workloads. The R9 M275X is the GPU for Vulkan-centric software stacks and offers a better average benchmark score despite losing the head-to-head. Neither card is a clear all-around winner; the choice depends entirely on which API and workload matters most to the user.