AMD Radeon RX 9070 XT vs NVIDIA GeForce GTX 965M Comparison

The NVIDIA GeForce GTX 965M and the AMD Radeon RX 9070 XT represent two vastly different eras of GPU design, separated by a decade of architectural evolution. The data shows a decisive performance gulf, with the AMD Radeon RX 9070 XT winning both shared benchmark tests by substantial margins. However, a closer look at the benchmark results reveals a more complex picture regarding their standing among all GPUs, as the newer card's average score is actually lower than its predecessor's, despite its individual test victories.

Head-to-Head Benchmarks

The two cards share only two direct benchmark comparisons: Geekbench OpenCL and Geekbench Vulkan. In the OpenCL test, the AMD Radeon RX 9070 XT scores 17,428 points against the GTX 965M's 14,509 points. This represents a 16.7% performance advantage for the AMD card, a clear but not overwhelming lead. The Vulkan test tells a far more dramatic story, with the RX 9070 XT achieving 65,879 points compared to the GTX 965M's 14,299 points, a staggering 78.3% delta in favor of the newer architecture.

While the RX 9070 XT wins both head-to-head tests, its performance profile is not uniformly dominant. The GTX 965M, despite its age, demonstrates respectable raw compute capability in the OpenCL test, staying within 16.7% of a card that is over a decade newer. This suggests that the Maxwell 2.0 architecture's compute efficiency remains competitive in certain workloads. The Vulkan result, however, exposes the generational gap in modern API support and execution efficiency, where the RDNA 4.0 architecture's 64 ray accelerators and 1:1 FP16 throughput provide a massive advantage.

It is noteworthy that the RX 9070 XT's average benchmark score across all its tests is 13,543, which is actually lower than the GTX 965M's average of 14,404. This occurs because the RX 9070 XT's benchmark suite includes several Passmark tests with very low scores (e.g., Passmark DirectX 9 at 353, DirectX 11 at 292, and DirectX 12 at 78), which drag down its arithmetic mean. The GTX 965M, having only two benchmark entries, benefits from a higher average despite losing both head-to-head matchups.

The Verdict

From the data alone, the AMD Radeon RX 9070 XT is the clear performance winner, taking both head-to-head benchmarks with deltas of 16.7% and 78.3%. Anyone requiring maximum performance in Vulkan-based applications or modern DirectX 12 Ultimate workloads should choose the RX 9070 XT without hesitation. Its 48.66 TFLOPS FP32 throughput and 16 GB of GDDR6 memory dwarf the GTX 965M's 1.946 TFLOPS and 2 GB GDDR5 allocation.

However, the benchmark data presents an unusual caveat. The RX 9070 XT holds a 55th percentile ranking among all GPUs, while the GTX 965M ranks at the 56th percentile. This means the older card sits slightly higher relative to its contemporaries than the newer card does. Additionally, the RX 9070 XT's nearest rivals include the AMD Radeon HD 7770M (avg score 13,536, delta 0.1%), NVIDIA GeForce GTX 570 (avg score 13,515, delta 0.2%), and NVIDIA P106-090 (avg score 13,470, delta 0.5%). This clustering suggests the RX 9070 XT's average score is not representative of its peak capability, which is heavily skewed by its exceptional Vulkan result.

For users prioritizing raw compute in OpenCL, the GTX 965M remains a surprisingly viable option, trailing by only 16.7%. But for any modern gaming or compute workload leveraging Vulkan, DirectX 12 Ultimate, or ray tracing, the RX 9070 XT is the only rational choice from this dataset. The GTX 965M is end-of-life, while the RX 9070 XT is active production, which further solidifies its position as the forward-looking option.

Architecture Differences

The architectural chasm between these two GPUs is profound. The NVIDIA GeForce GTX 965M utilizes the GM204 chip built on TSMC's 28 nm process, featuring 5,200 million transistors on a 398 mm² die. This translates to a transistor density of 13.1 million per square millimeter. In contrast, the AMD Radeon RX 9070 XT uses the Navi 48 chip on TSMC's 4 nm process, packing 53,900 million transistors into a smaller 357 mm² die, achieving a density of 151.0 million per square millimeter—over 11 times denser.

The GTX 965M is based on Maxwell 2.0 architecture from the GeForce 900M generation, supporting DirectX 12 (12_1), while the RX 9070 XT uses RDNA 4.0 from the Navi IV (RX 9000) generation, supporting DirectX 12 Ultimate (12_2). Both support OpenGL 4.6 and Vulkan 1.4. The RX 9070 XT includes 64 dedicated ray accelerators, a feature entirely absent from the GTX 965M, which has no RT cores at all.

Core configurations differ massively: the GTX 965M has 1,024 shading units, 64 TMUs, and 32 ROPs, while the RX 9070 XT features 4,096 shading units, 256 TMUs, and 128 ROPs. The RX 9070 XT's FP32 throughput of 48.66 TFLOPS is exactly 25 times the GTX 965M's 1.946 TFLOPS. The AMD card also offers FP16 compute at 48.66 TFLOPS (1:1 ratio), a capability not listed for the NVIDIA card.

FAQ

Q: Which card has higher raw compute performance?

A: The AMD Radeon RX 9070 XT delivers 48.66 TFLOPS FP32, which is 25 times the GTX 965M's 1.946 TFLOPS. In Geekbench OpenCL, the RX 9070 XT scores 17,428 versus 14,509 for the GTX 965M, a 16.7% advantage.

Q: How do they compare in Vulkan performance?

A: The RX 9070 XT achieves 65,879 in Geekbench Vulkan, while the GTX 965M scores 14,299. This represents a 78.3% delta in favor of the AMD card, the largest performance gap in any shared benchmark.

Q: What are the memory specifications?

A: The GTX 965M has 2 GB of GDDR5 memory on a 128-bit bus with 80.19 GB/s bandwidth. The RX 9070 XT has 16 GB of GDDR6 memory on a 256-bit bus with 644.6 GB/s bandwidth.

Q: Which card has a higher percentile ranking?

A: The GTX 965M ranks at the 56th percentile among all GPUs, while the RX 9070 XT ranks at the 55th percentile. Despite losing both head-to-head tests, the GTX 965M holds a slightly higher relative standing.

Q: What is the production status of each card?

A: The GTX 965M is end-of-life, having been released on 2015-01-08. The RX 9070 XT is active production, released on 2025-03-05, with a successor listed as null.

Q: What are the power requirements?

A: The RX 9070 XT has a TDP of 304 W, requires 2x 8-pin power connectors, and suggests a 700 W PSU. The GTX 965M has no TDP listed, uses no power connectors, and is an MXM module.

Where Each One Wins

The AMD Radeon RX 9070 XT wins decisively in all modern, high-intensity workloads. Its Vulkan score of 65,879 is the standout result, making it the preferred choice for any application leveraging this API. The 16 GB GDDR6 frame buffer with 644.6 GB/s bandwidth supports high-resolution textures and large datasets that would instantly exhaust the GTX 965M's 2 GB allocation. The 64 ray accelerators enable hardware-accelerated ray tracing, a feature the GTX 965M cannot perform. For DirectX 12 Ultimate content, the RX 9070 XT's 12_2 support provides feature-level parity with modern consoles and games.

The NVIDIA GeForce GTX 965M's wins are more subtle. In the OpenCL test, it trails by only 16.7%, showing competitive compute performance for a 2015-era mobile chip. Its 28 nm process and 398 mm² die size make it physically larger than the RX 9070 XT's 357 mm² die, despite having 48,700 million fewer transistors. The GTX 965M also holds a higher percentile ranking (56th vs 55th) among all GPUs, indicating that relative to its release-era peers, it was a stronger performer than the RX 9070 XT is among its contemporaries. For legacy applications that do not benefit from modern API features or large memory pools, the GTX 965M remains serviceable.

Specification Differences

| Specification | NVIDIA GeForce GTX 965M | AMD Radeon RX 9070 XT |

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

| Chip | GM204 | Navi 48 |

| Architecture | Maxwell 2.0 | RDNA 4.0 |

| Process Node | 28 nm | 4 nm |

| Transistors | 5,200 million | 53,900 million |

| Die Size | 398 mm² | 357 mm² |

| Transistor Density | 13.1M / mm² | 151.0M / mm² |

| Base Clock | 924 MHz | 1660 MHz |

| Boost Clock | 950 MHz | 2970 MHz |

| Memory Size | 2 GB | 16 GB |

| Memory Type | GDDR5 | GDDR6 |

| Memory Bus | 128 bit | 256 bit |

| Memory Bandwidth | 80.19 GB/s | 644.6 GB/s |

| Shading Units | 1024 | 4096 |

| TMUs | 64 | 256 |

| ROPs | 32 | 128 |

| RT Cores | None | 64 |

| Pixel Rate | 30.40 GPixel/s | 380.2 GPixel/s |

| Texture Rate | 60.80 GTexel/s | 760.3 GTexel/s |

| FP32 | 1.946 TFLOPS | 48.66 TFLOPS |

| FP16 | Not listed | 48.66 TFLOPS (1:1) |

| TDP | Not listed | 304 W |

| Slot Width | MXM Module | Dual-slot |

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

| Suggested PSU | Not listed | 700 W |

| Bus Interface | MXM-B (3.0) | PCIe 5.0 x16 |

| Display Outputs | Portable Device Dependent | 1x HDMI 2.1b, 3x DisplayPort 2.1a |

| DirectX | 12 (12_1) | 12 Ultimate (12_2) |

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

| Release Date | 2015-01-08 | 2025-03-05 |

| Launch MSRP | Not listed | 599 USD (stated once) |