AMD Radeon Pro 5600M vs NVIDIA GeForce GTX 965M Comparison

Where Each One Wins

The benchmark split between these two mobile graphics solutions is completely one-sided. Across the two shared tests in the database, the AMD Radeon Pro 5600M claims both victories, leaving the NVIDIA GeForce GTX 965M without a single head-to-head win.

The AMD part dominates compute-oriented workloads. In Geekbench OpenCL, the Radeon Pro 5600M records 47,783 points against 14,509 for the GTX 965M, a 229.3% advantage. The Vulkan test tells a similar story: 46,425 versus 14,299, a 224.7% gap. These are not marginal differences; the AMD solution more than triples the NVIDIA part's output in both APIs.

The GTX 965M's strengths, such as they are, lie outside the shared benchmark set. The database lists no DirectX or Passmark results for the NVIDIA part, so its legacy DirectX 9, 10, 11, and 12 performance cannot be directly compared. The AMD card, by contrast, has a full suite of Passmark scores: 118 in DirectX 9, 62 in DirectX 10, 59 in DirectX 11, and 39 in DirectX 12, alongside a Passmark G3D score of 9,279 and a GPU compute score of 4,031. Without matching NVIDIA numbers, the database cannot confirm where the older Maxwell chip might claw back ground in older APIs.

The percentile ranking reinforces the gap. The Radeon Pro 5600M sits at the 59th percentile among all GPUs, while the GTX 965M rests at the 56th percentile. The average benchmark score tells a clearer story: 16,351 for AMD versus 14,404 for NVIDIA. The AMD part's nearest rivals cluster just above it, with the RX 5700 XT at 16,361 (0.1% higher), the RTX PRO 6000 Blackwell at 16,408 (0.3% higher), and the Radeon PRO W7500 at 16,415 (0.4% higher). The GTX 965M's nearest rivals, meanwhile, sit slightly below it: the RX Vega 11 at 14,385 (0.1% lower), the GTX TITAN at 14,373 (0.2% lower), and the Radeon Vega 11 at 14,352 (0.4% lower).

Architecture Differences

The two GPUs come from different manufacturing generations and design philosophies. The AMD Radeon Pro 5600M uses the Navi 12 chip built on RDNA 1.0 architecture, fabricated on TSMC's 7 nm process. The NVIDIA GeForce GTX 965M uses the GM204 chip on Maxwell 2.0 architecture, fabricated on TSMC's 28 nm process. That process gap is substantial: 7 nm versus 28 nm represents a significant generational leap in transistor density and power efficiency.

The database lists transistor counts for the GTX 965M: 5,200 million transistors on a 398 mm² die, yielding a density of 13.1 million transistors per square millimeter. No transistor count or die size is recorded for the Radeon Pro 5600M, so a direct density comparison is not possible from the data.

Memory configurations differ fundamentally. The AMD part ships with 8 GB of HBM2 across a 2048-bit bus, delivering 394.2 GB/s of bandwidth. The NVIDIA part has 2 GB of GDDR5 on a 128-bit bus, providing 80.19 GB/s. The bandwidth advantage for AMD is roughly fivefold, which directly explains the large compute benchmark gaps. Memory clocks also differ: the Radeon Pro 5600M runs at 770 MHz (1,540 Mbps effective), while the GTX 965M runs at 1,253 MHz (5 Gbps effective).

Compute resources are heavily skewed toward AMD. The Radeon Pro 5600M has 2,560 shading units, 160 texture mapping units, and 64 ROPs. The GTX 965M has 1,024 shading units, 64 TMUs, and 32 ROPs. The AMD part doubles the TMU and ROP counts and has 2.5 times the shading units. Pixel rate is 73.22 GPixel/s for AMD versus 30.40 GPixel/s for NVIDIA. Texture rate is 183.0 GTexel/s versus 60.80 GTexel/s. FP32 throughput is 5.857 TFLOPS versus 1.946 TFLOPS. The AMD part also supports FP16 at 11.71 TFLOPS (2:1), while no FP16 figure is listed for the NVIDIA part.

Clock speeds tell a different story. The GTX 965M runs higher: 924 MHz base and 950 MHz boost, versus 822 MHz base and 1,144 MHz boost for the Radeon Pro 5600M. The AMD part has a lower base clock but a meaningfully higher boost ceiling, and the architectural advantages more than compensate.

The form factors differ as well. The Radeon Pro 5600M is listed as an IGP (integrated graphics processor) with no power connectors, while the GTX 965M is an MXM module with no power connectors. The database records a 50 W TDP for the AMD part but no TDP for the NVIDIA part. The AMD card uses PCIe 4.0 x16, while the NVIDIA card uses MXM-B (3.0). Both are end-of-life products. The AMD part launched on 2020-06-14, while the GTX 965M launched on 2015-01-08.

API support is broadly similar. Both support DirectX 12 (12_1) and OpenGL 4.6. The AMD part supports Vulkan 1.3, while the NVIDIA part supports Vulkan 1.4. Neither has ray tracing or tensor cores listed.

Head-to-Head Benchmarks

The database records exactly two shared benchmark tests between these GPUs, and both go decisively to the AMD Radeon Pro 5600M.

In Geekbench OpenCL, the AMD part scores 47,783 against 14,509 for the GTX 965M. The delta is 229.3% in favor of AMD. This is the larger of the two gaps. OpenCL is a general-purpose compute API, and the Radeon Pro 5600M's combination of 2,560 shading units, HBM2 bandwidth, and RDNA 1.0 architecture produces an overwhelming advantage. The GTX 965M's 1,024 shading units and 80.19 GB/s of GDDR5 bandwidth simply cannot keep pace.

In Geekbench Vulkan, the AMD part scores 46,425 against 14,299 for the GTX 965M, a 224.7% delta. Vulkan is a low-level graphics and compute API, and the result tracks closely with the OpenCL gap. The slightly smaller margin suggests the NVIDIA part is marginally less disadvantaged in graphics-oriented workloads than in pure compute, but the difference is negligible: 229.3% versus 224.7%.

The average benchmark scores in the database reinforce these results. The Radeon Pro 5600M averages 16,351 across all recorded tests, while the GTX 965M averages 14,404. The AMD part sits 13.5% higher on average. Its nearest rival, the RX 5700 XT, averages 16,361, just 0.1% above the Radeon Pro 5600M. The GTX 965M's nearest rival, the RX Vega 11, averages 14,385, just 0.1% below it.

The AMD part's additional benchmark entries, which have no NVIDIA counterparts, show a mixed profile across older APIs. Passmark DirectX 9 scores 118, DirectX 10 scores 62, DirectX 11 scores 59, and DirectX 12 scores 39. The G2D score is 679, and the G3D score is 9,279. The GPU compute score is 4,031. These numbers suggest the RDNA 1.0 architecture is strongest in compute and modern APIs, with lower relative scores in older DirectX versions. Without GTX 965M numbers for these tests, no direct comparison is possible.

The Verdict

The data points to one conclusion: the AMD Radeon Pro 5600M is the substantially faster GPU in every measurable shared workload. The OpenCL and Vulkan results show a two-to-threefold performance advantage, and the average benchmark score, percentile rank, memory bandwidth, and compute resources all favor AMD.

For compute-intensive tasks such as OpenCL and Vulkan workloads, the Radeon Pro 5600M is the clear choice. The 229.3% OpenCL lead and 224.7% Vulkan lead are not close to being competitive. The HBM2 memory subsystem, with 394.2 GB/s of bandwidth, gives the AMD part a massive throughput advantage that no amount of clock speed on the GTX 965M can offset.

For users constrained to the GTX 965M's generation, the Maxwell 2.0 architecture still holds up in the database's percentile ranking at 56th overall, and its nearest rivals are all within 0.6% of its average score. But against the Radeon Pro 5600M, it is outclassed in every shared metric.

The GTX 965M is an older product from 2015, built on 28 nm, with 2 GB of GDDR5. The Radeon Pro 5600M is from 2020, built on 7 nm, with 8 GB of HBM2. The generational gap is apparent in every benchmark result. Neither product has a launch MSRP listed in the database, so no pricing analysis is possible.

The practical verdict: pick the Radeon Pro 5600M for any workload that uses OpenCL or Vulkan. The GTX 965M has no recorded advantage in any shared test. The only reason to choose the NVIDIA part would be platform compatibility or a specific need for its MXM form factor, neither of which the benchmark data can quantify.

FAQ

Q: How much faster is the AMD Radeon Pro 5600M in OpenCL?

A: The Radeon Pro 5600M scores 47,783 in Geekbench OpenCL, while the GTX 965M scores 14,509. That is a 229.3% advantage for the AMD part.

Q: What about Vulkan performance?

A: In Geekbench Vulkan, the Radeon Pro 5600M scores 46,425 versus 14,299 for the GTX 965M, a 224.7% lead for AMD.

Q: Which GPU has more memory bandwidth?

A: The Radeon Pro 5600M has 8 GB of HBM2 on a 2048-bit bus, delivering 394.2 GB/s. The GTX 965M has 2 GB of GDDR5 on a 128-bit bus, delivering 80.19 GB/s.

Q: What are the architecture differences?

A: The Radeon Pro 5600M uses Navi 12 on RDNA 1.0 architecture at 7 nm. The GTX 965M uses GM204 on Maxwell 2.0 architecture at 28 nm. The GTX 965M has 5,200 million transistors on a 398 mm² die.

Q: Do both GPUs support the same APIs?

A: Both support DirectX 12 (12_1) and OpenGL 4.6. The Radeon Pro 5600M supports Vulkan 1.3, while the GTX 965M supports Vulkan 1.4.

Q: How do their average benchmark scores compare?

A: The Radeon Pro 5600M averages 16,351 across all recorded tests, while the GTX 965M averages 14,404. The AMD part sits at the 59th percentile of all GPUs, the NVIDIA part at the 56th.

Specification Differences

| Specification | AMD Radeon Pro 5600M | NVIDIA GeForce GTX 965M |

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

| Chip | Navi 12 | GM204 |

| Architecture | RDNA 1.0 | Maxwell 2.0 |

| Generation | Radeon Pro Mac (Navi Mobile) | GeForce 900M |

| Process Node | 7 nm | 28 nm |

| Foundry | TSMC | TSMC |

| Transistors | Not listed | 5,200 million |

| Die Size | Not listed | 398 mm² |

| Transistor Density | Not listed | 13.1M / mm² |

| Base Clock | 822 MHz | 924 MHz |

| Boost Clock | 1,144 MHz | 950 MHz |

| Memory Clock | 770 MHz (1,540 Mbps effective) | 1,253 MHz (5 Gbps effective) |

| Memory Size | 8 GB | 2 GB |

| Memory Type | HBM2 | GDDR5 |

| Memory Bus Width | 2048 bit | 128 bit |

| Memory Bandwidth | 394.2 GB/s | 80.19 GB/s |

| Shading Units | 2,560 | 1,024 |

| TMUs | 160 | 64 |

| ROPs | 64 | 32 |

| Pixel Rate | 73.22 GPixel/s | 30.40 GPixel/s |

| Texture Rate | 183.0 GTexel/s | 60.80 GTexel/s |

| FP32 | 5.857 TFLOPS | 1.946 TFLOPS |

| FP16 | 11.71 TFLOPS (2:1) | Not listed |

| TDP | 50 W | Not listed |

| Slot Width | IGP | MXM Module |

| Power Connectors | None | None |

| Bus Interface | PCIe 4.0 x16 | MXM-B (3.0) |

| Vulkan Support | 1.3 | 1.4 |

| Release Date | 2020-06-14 | 2015-01-08 |

| Predecessor | Not listed | GeForce 800M |

| Successor | Not listed | GeForce 10 Mobile |

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