AMD Instinct MI350P vs AMD Radeon 760M Comparison

AMD Instinct MI350P and AMD Radeon 760M represent two entirely different branches of AMD’s GPU lineup, one aimed at massive compute installations and the other at integrated graphics for mobile and desktop processors. The MI350P is a dual-slot accelerator built on the CDNA 4.0 architecture with a 3 nm process, while the 760M is an IGP on the RDNA 3.0 architecture using a 4 nm process. Benchmark data for the MI350P is not recorded in the database, but its architectural specifications and the 760M’s measured scores provide a clear basis for comparison.

FAQ

Q: What is the process node difference between the AMD Instinct MI350P and the AMD Radeon 760M?

A: The Instinct MI350P uses a 3 nm process, while the Radeon 760M uses a 4 nm process. Both are fabricated by TSMC.

Q: How much memory does the AMD Instinct MI350P have, and what type is it?

A: The Instinct MI350P has 144 GB of HBM3e memory with an 8192-bit bus width and 8.19 TB/s of bandwidth. The Radeon 760M uses system shared memory with system-dependent bandwidth.

Q: What is the shading unit count for each GPU?

A: The Instinct MI350P has 8192 shading units, while the Radeon 760M has 512 shading units. This is a 16x difference in shading unit count.

Q: What is the thermal design power (TDP) for each?

A: The Instinct MI350P has a TDP of 600 W and requires a 1000 W suggested PSU, while the Radeon 760M has a TDP of 15 W and uses no power connectors as an IGP.

Q: What is the release date for each product?

A: The Instinct MI350P is set for release on May 6, 2026, while the Radeon 760M was released on January 30, 2024.

Q: Does the Radeon 760M support ray tracing?

A: Yes, the Radeon 760M has 8 ray tracing cores. The Instinct MI350P has no listed ray tracing cores.

Q: What is the average benchmark score for the Radeon 760M?

A: The Radeon 760M has an average benchmark score of 6019, placing it at the 35th percentile of all GPUs in the database. The Instinct MI350P has no recorded benchmark scores.

Architecture Differences

The architecture gap between these two GPUs is fundamental. The AMD Instinct MI350P uses CDNA 4.0, a compute-focused architecture designed for data center workloads. It is built on a 3 nm process from TSMC and packs 73,000 million transistors on a 1190 mm² die, resulting in a transistor density of 61.3 million per mm². In contrast, the AMD Radeon 760M uses RDNA 3.0, a graphics-oriented architecture for integrated solutions. It is built on a 4 nm process with 25,390 million transistors on a 178 mm² die, achieving a much higher transistor density of 142.6 million per mm².

The MI350P has 8192 shading units, 512 texture mapping units, and 0 raster operations pipelines. Its pixel rate is listed as 0 MPixel/s, which reflects its non-rendering nature. The 760M has 512 shading units, 32 TMUs, and 16 ROPs, with a pixel rate of 41.58 GPixel/s. The 760M also includes 8 ray tracing cores, a feature absent from the MI350P’s specification sheet.

Memory architecture further separates them. The MI350P uses 144 GB of HBM3e with a massive 8192-bit bus and 8.19 TB/s bandwidth. The 760M relies on system shared memory, which means bandwidth is system dependent. The MI350P’s memory clock is listed at 2000 MHz with 8 Gbps effective, while the 760M has no dedicated memory clock. API support also differs sharply: the MI350P reports N/A for DirectX, OpenGL, and Vulkan, whereas the 760M supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

Head-to-Head Benchmarks

Direct benchmark comparisons are impossible because the Instinct MI350P has no recorded scores in the database. However, the Radeon 760M has a full suite of measured results that can be analyzed independently. Its average benchmark score is 6019, and it sits at the 35th percentile of all GPUs. The nearest rivals in the database are the AMD Radeon RX 6400 with an average score of 6001 (0.3% ahead), the NVIDIA GeForce GTX 770M at 6000 (0.3% ahead), the NVIDIA RTX PRO 6000 Blackwell Server at 5996 (0.4% ahead), and the NVIDIA Quadro P2000 at 6049 (0.5% behind).

Looking at specific tests, the 760M shows its strongest result in Geekbench Vulkan with a score of 30336. In Geekbench OpenCL it scores 20255. The Passmark suite shows more varied results: Passmark G3D scores 5310, Passmark G2D scores 890, and Passmark GPU Compute scores 2840. DirectX tests are lower, with Passmark DirectX 11 at 52, DirectX 12 at 25, and DirectX 10 at 19. Passmark DirectX 9 scores 65. The 3DMark Steel Nomad DX12 test yields a score of 400.

The MI350P’s theoretical specifications suggest a vastly different performance envelope. Its FP32 compute is listed at 36.04 TFLOPS, and its FP16 compute is also 36.04 TFLOPS with a 1:1 ratio. The 760M delivers 5.323 TFLOPS for both FP32 and FP16, also with a 1:1 ratio. The texture rate for the MI350P is 1,126.4 GTexel/s, while the 760M manages 83.17 GTexel/s. Pixel rate is 0 MPixel/s for the MI350P versus 41.58 GPixel/s for the 760M. These numbers indicate the MI350P is built for raw compute throughput, not rasterization, which explains the absence of display outputs and the N/A API listings.

The Verdict

The data indicates that the AMD Instinct MI350P and AMD Radeon 760M serve completely different purposes with no overlap in intended use cases. The MI350P is a data center accelerator with 144 GB of HBM3e memory, 8192 shading units, and a 600 W TDP. It has no display outputs and no consumer API support, which means it is not designed for gaming or desktop graphics. The 760M, by contrast, is an integrated GPU with 512 shading units, 8 ray tracing cores, a 15 W TDP, and full DirectX 12 Ultimate support, making it suitable for lightweight gaming and general graphical tasks in laptops or compact desktops.

For compute-heavy workloads such as AI training or large-scale simulations, the MI350P’s specifications indicate a clear advantage in memory capacity and bandwidth, as well as raw FP32 and FP16 throughput. The 760M’s 5.323 TFLOPS is a fraction of the MI350P’s 36.04 TFLOPS, and its system shared memory cannot match the 8.19 TB/s of dedicated HBM3e. The MI350P’s 8192-bit bus width is 16 times wider than what the 760M effectively uses, though the 760M’s bus width is listed as system shared.

For consumer applications like gaming or desktop productivity, the 760M is the only viable option because the MI350P lacks the APIs and display outputs required for such tasks. The 760M’s benchmark scores show it competes closely with entry-level discrete GPUs like the Radeon RX 6400 and the GeForce GTX 770M, with deltas of 0.3% either way. Its ray tracing capability, despite being modest, is something the MI350P does not list at all.

The production status also differs: the 760M is marked as Active, while the MI350P has no status listed. The MI350P’s release date of May 6, 2026, is over two years after the 760M’s January 30, 2024 launch. The MI350P is a dual-slot card measuring 267 mm in length, 111 mm in height, and 40 mm in width, with a 1x 16-pin power connector. The 760M is an IGP with no dimensions listed and no power connectors.

Specification Differences

The following table outlines the key specification differences between the two GPUs, based solely on the recorded data.

| Specification | AMD Instinct MI350P | AMD Radeon 760M |

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

| Architecture | CDNA 4.0 | RDNA 3.0 |

| Process Node | 3 nm | 4 nm |

| Transistors | 73,000 million | 25,390 million |

| Die Size | 1190 mm² | 178 mm² |

| Transistor Density | 61.3M / mm² | 142.6M / mm² |

| Base Clock | 1000 MHz | 800 MHz |

| Boost Clock | 2200 MHz | 2599 MHz |

| Memory Size | 144 GB | System Shared |

| Memory Type | HBM3e | System Shared |

| Memory Bus Width | 8192 bit | System Shared |

| Memory Bandwidth | 8.19 TB/s | System Dependent |

| Shading Units | 8192 | 512 |

| TMUs | 512 | 32 |

| ROPs | 0 | 16 |

| Ray Tracing Cores | None listed | 8 |

| Pixel Rate | 0 MPixel/s | 41.58 GPixel/s |

| Texture Rate | 1,126.4 GTexel/s | 83.17 GTexel/s |

| FP32 Performance | 36.04 TFLOPS | 5.323 TFLOPS |

| FP16 Performance | 36.04 TFLOPS (1:1) | 5.323 TFLOPS (1:1) |

| TDP | 600 W | 15 W |

| Slot Width | Dual-slot | IGP |

| Power Connectors | 1x 16-pin | None |

| Suggested PSU | 1000 W | None listed |

| Bus Interface | PCIe 5.0 x16 | PCIe 4.0 x8 |

| Display Outputs | No outputs | Motherboard Dependent |

| DirectX Support | N/A | 12 Ultimate (12_2) |

| OpenGL Support | N/A | 4.6 |

| Vulkan Support | N/A | 1.4 |

| Release Date | 2026-05-06 | 2024-01-30 |

| Production Status | Not listed | Active |

| Predecessor | Radeon Instinct | Navi II IGP |