AMD Instinct MI350X vs AMD Radeon 8065S Comparison

Head-to-Head Benchmarks

The recorded database contains no head-to-head benchmark results for the AMD Instinct MI350X versus the AMD Radeon 8065S. Both parts show an average benchmark score of zero, and neither has any nearest rivals listed. The wins count is zero for each, meaning there is no measured performance comparison to draw upon from our measurements. Consequently, any evaluation of these two accelerators must be derived entirely from their architectural specifications and the theoretical throughput figures recorded in the database.

What the data does show is a stark contrast in raw compute capacity. The Instinct MI350X delivers 72.09 TFLOPS of FP32 throughput, while the Radeon 8065S delivers 15.36 TFLOPS. That places the MI350X at roughly 4.69 times the FP32 throughput of the 8065S, a significant margin that reflects their entirely different design targets. In FP16, both parts maintain a 1:1 ratio with their FP32 figures, so the MI350X again shows 72.09 TFLOPS versus 15.36 TFLOPS for the 8065S.

Texture throughput tells a similar story. The MI350X achieves 2,252.8 GTexel/s, while the 8065S achieves 480.0 GTexel/s. The MI350X is about 4.69 times faster here as well, consistent with the FP32 ratio, since texture rate is closely tied to shader count and clock speed. Pixel rate is where the comparison inverts: the MI350X is recorded at 0 MPixel/s with zero ROPs, whereas the 8065S delivers 192.0 GPixel/s from its 64 ROPs. The MI350X is not designed for rasterized graphics output at all, so this is not a weakness in its intended role, but it does mean the 8065S is the only one of the two that can drive a display pipeline.

FAQ

Q: Which GPU has higher FP32 compute throughput?

A: The AMD Instinct MI350X, with 72.09 TFLOPS, compared to 15.36 TFLOPS for the AMD Radeon 8065S.

Q: What is the memory configuration of each part?

A: The MI350X uses 288 GB of HBM3e memory on an 8192-bit bus, with 8.19 TB/s of bandwidth. The 8065S uses System Shared memory, with a System Shared bus width and System Dependent bandwidth.

Q: Are these GPUs the same physical size?

A: No. The MI350X is an OAM Module with dimensions of 102 mm in length and 165 mm in width. The 8065S is an IGP with no recorded dimensions.

Q: Which GPU has a higher boost clock?

A: The Radeon 8065S boosts to 3000 MHz, while the Instinct MI350X boosts to 2200 MHz.

Q: What is the process node for each chip?

A: The MI350X uses a 3 nm process, and the 8065S uses a 4 nm process. Both are fabricated by TSMC.

Q: Does either GPU support PCIe 5.0?

A: Yes, both the MI350X and the 8065S use a PCIe 5.0 x16 bus interface.

Architecture Differences

The two accelerators come from entirely different architectural families within AMD. The Instinct MI350X is built on CDNA 4.0, a compute-focused architecture designed for data center workloads, while the Radeon 8065S uses RDNA 3.5, a graphics-oriented architecture for mobile devices. Their generations reflect this split: the MI350X belongs to the Instinct (MIx) generation, and the 8065S belongs to the Navi Mobile (RX 8000M) generation.

The silicon itself diverges sharply. The MI350X uses the MI350 256CU chip on a 3 nm TSMC process, with 185,000 million transistors on a 2380 mm² die, yielding a transistor density of 77.7M per mm². The 8065S uses the Gorgon Halo chip on a 4 nm TSMC process, with a 308 mm² die and unknown transistor count. The MI350X is a massive multi-chip module style accelerator, while the 8065S is a compact integrated graphics processor.

Compute resources follow the same pattern. The MI350X contains 16,384 shading units and 1,024 TMUs, with zero ROPs. The 8065S contains 2,560 shading units, 160 TMUs, and 64 ROPs. The MI350X has no recorded ray tracing cores, while the 8065S has 40 RT cores. Neither part lists tensor cores. The MI350X has no display outputs, while the 8065S has outputs described as Portable Device Dependent.

Memory architecture is fundamentally different as well. The MI350X carries 288 GB of HBM3e memory with an 8192-bit bus and 8.19 TB/s of bandwidth, clocked at 2000 MHz with 8 Gbps effective. The 8065S uses System Shared memory, meaning it relies on the host system's main memory rather than dedicated VRAM. Its memory clock is recorded as System Shared, and bandwidth is System Dependent.

Power and physical design also differentiate them. The MI350X has a TDP of 1000 W, requires a suggested PSU of 1400 W, uses an OAM Module slot width, and has no power connectors of its own. The 8065S has a TDP of 55 W, uses an IGP slot width, has no power connectors, and no suggested PSU is recorded. API support is another clear divider: the MI350X records N/A for DirectX, OpenGL, and Vulkan, while the 8065S supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

The Verdict

The data indicates two products with no overlap in purpose. The AMD Instinct MI350X is a 1000 W accelerator built for compute density, with 72.09 TFLOPS of FP32, 288 GB of HBM3e memory, and no graphics output capability. The AMD Radeon 8065S is a 55 W integrated GPU for mobile systems, with 15.36 TFLOPS of FP32, 64 ROPs, ray tracing cores, and full graphics API support. Based strictly on the recorded specifications, the MI350X is for server-side compute tasks that demand massive throughput and memory capacity, while the 8065S is for client devices that need graphics rendering and display output at low power. Neither part can substitute for the other.

The percentile ranking for both is 50%, but this reflects the absence of benchmark data rather than any measured performance equivalence. The MI350X has no launch MSRP recorded, and neither does the 8065S.

Specification Differences

| Specification | AMD Instinct MI350X | AMD Radeon 8065S |

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

| Architecture | CDNA 4.0 | RDNA 3.5 |

| Generation | Instinct (MIx) | Navi Mobile (RX 8000M) |

| Process Node | 3 nm | 4 nm |

| Foundry | TSMC | TSMC |

| Transistors | 185,000 million | Unknown |

| Die Size | 2380 mm² | 308 mm² |

| Base Clock | 1000 MHz | 1295 MHz |

| Boost Clock | 2200 MHz | 3000 MHz |

| Memory Size | 288 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 | 16384 | 2560 |

| TMUs | 1024 | 160 |

| ROPs | 0 | 64 |

| RT Cores | Null | 40 |

| Pixel Rate | 0 MPixel/s | 192.0 GPixel/s |

| Texture Rate | 2,252.8 GTexel/s | 480.0 GTexel/s |

| FP32 | 72.09 TFLOPS | 15.36 TFLOPS |

| FP16 | 72.09 TFLOPS (1:1) | 15.36 TFLOPS (1:1) |

| TDP | 1000 W | 55 W |

| Slot Width | OAM Module | IGP |

| Suggested PSU | 1400 W | Null |

| Bus Interface | PCIe 5.0 x16 | PCIe 5.0 x16 |

| Display Outputs | No outputs | Portable Device Dependent |

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

| OpenGL | N/A | 4.6 |

| Vulkan | N/A | 1.4 |

| Release Date | 2025-06-11 | 2025-12-31 |

| Predecessor | Radeon Instinct | Polaris Mobile |

Where Each One Wins

The AMD Instinct MI350X wins decisively in raw compute throughput. Its 72.09 TFLOPS of FP32 and FP16 is 4.69 times that of the Radeon 8065S. Its texture rate of 2,252.8 GTexel/s similarly dwarfs the 8065S's 480.0 GTexel/s. The MI350X also provides 288 GB of dedicated HBM3e memory with 8.19 TB/s of bandwidth, a feature set aimed at large-scale compute workloads such as AI training and scientific simulation. Its 1000 W TDP and OAM Module form factor indicate it is meant for data center racks, not desktops or laptops.

The AMD Radeon 8065S wins in graphics and mobile applicability. It delivers 192.0 GPixel/s from its 64 ROPs, while the MI350X delivers 0 MPixel/s. The 8065S includes 40 ray tracing cores, support for DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4, and has display outputs. Its 55 W TDP and IGP form factor make it suitable for portable devices, and its 3000 MHz boost clock is higher than the MI350X's 2200 MHz. The 8065S also uses system shared memory, which eliminates the need for dedicated VRAM and simplifies mobile system design.

For workloads that require FP32 or FP16 throughput, the MI350X is the clear choice based on the data. For workloads that require rasterization, ray tracing, or any display output, the 8065S is the only option of the two. The MI350X has no graphics API support recorded, so it cannot render frames. The 8065S has no dedicated memory, so it cannot match the MI350X's bandwidth or capacity. The database records show two accelerators that are complementary, not competitive.