AMD Instinct MI308X vs Intel Arc A380M Comparison

Head-to-Head Benchmarks

The recorded database contains no head-to-head benchmark entries for the AMD Instinct MI308X and Intel Arc A380M. Both parts hold a 50th percentile position against all GPUs in the database, and each carries an average benchmark score of zero. This means there are no measured performance contests, no synthetic test deltas, and no percentage gaps to report from direct comparison runs. The absence of data is itself informative: these two accelerators occupy entirely different segments, and no shared workload has been submitted to the database for both.

The AMD Instinct MI308X delivers 81.72 TFLOPS of FP32 compute and 81.72 TFLOPS of FP16 compute at a 1:1 ratio. The Intel Arc A380M delivers 4.096 TFLOPS of FP32 and 8.192 TFLOPS of FP16 at a 2:1 ratio. The FP32 gap is exactly 20x in favor of the AMD part. The FP16 gap is roughly 10x in favor of the AMD part, though the Intel architecture processes FP16 at twice the rate of its FP32 path while the AMD architecture processes both at the same rate. Texture rate favors the MI308X at 2,553.6 GTexel/s versus 128.0 GTexel/s for the A380M, a 19.95x difference. Pixel rate reverses the pattern: the A380M posts 64.00 GPixel/s while the MI308X records 0 MPixel/s, because the Instinct part has zero ROPs and no display output path.

Architecture Differences

The MI308X uses the Aqua Vanjaram chip built on CDNA 3.0 architecture, manufactured on a 5 nm TSMC process. The A380M uses the DG2-128 chip built on Xe-HPG architecture, manufactured on a 6 nm TSMC process. Transistor counts differ dramatically: 153,000 million on the MI308X versus 7,200 million on the A380M. Die size follows suit at 1017 mm² versus 157 mm². Transistor density per square millimeter stands at 150.4M for the AMD chip and 45.9M for the Intel chip, indicating the CDNA 3.0 design packs logic more tightly despite the larger overall die.

Memory architecture separates the two completely. The MI308X carries 192 GB of HBM3 on an 8192-bit bus, delivering 5.32 TB/s of bandwidth. The A380M carries 6 GB of GDDR6 on a 96-bit bus, delivering 186.0 GB/s. The bandwidth ratio is roughly 28.6x in favor of the AMD part. Clock behavior differs as well: the MI308X runs a 1000 MHz base and 2100 MHz boost, with memory at 1300 MHz (5.2 Gbps effective). The A380M runs a 1550 MHz base and 2000 MHz boost, with memory at 1937 MHz (15.5 Gbps effective). The Intel part has a higher base clock by 550 MHz, but the AMD part has a 100 MHz higher boost clock.

Compute resources show the scale gap. The MI308X deploys 19,456 shading units and 1,216 TMUs, with zero ROPs. The A380M deploys 1,024 shading units, 64 TMUs, and 32 ROPs. The A380M includes 8 ray tracing cores; the MI308X records no ray tracing core count. API support differs fundamentally. The A380M supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The MI308X records N/A for DirectX, OpenGL, and Vulkan, reflecting its compute-oriented design with no graphics API path.

Power and physical configuration reinforce the divergence. The MI308X draws 750 W TDP with a suggested PSU of 1150 W, uses an OAM Module slot width, and has no power connectors listed. The A380M draws 35 W TDP, uses an MXM Module slot width with an MXM-A (3.1) bus interface, and lists no suggested PSU. The MI308X connects via PCIe 5.0 x16; the A380M uses MXM-A (3.1). Display outputs: the MI308X has none, while the A380M lists "Portable Device Dependent" outputs. Release dates place the A380M earlier at 2023-01-23 and the MI308X at 2023-12-05. Production status for the A380M is Active; the MI308X has no production status recorded.

Where Each One Wins

The MI308X wins in raw compute throughput, memory capacity, memory bandwidth, texture throughput, and transistor scale. Its 81.72 TFLOPS FP32 and FP16 figures position it for dense numerical workloads that operate on very large datasets. The 192 GB HBM3 pool with 5.32 TB/s bandwidth supports models and arrays that would never fit in the A380M's 6 GB GDDR6 frame buffer. The 8192-bit bus width is 85.3x wider than the A380M's 96-bit bus. For any workload that streams data continuously, the MI308X has an overwhelming advantage.

The A380M wins in pixel throughput, ray tracing capability, API compatibility, and power efficiency. Its 64.00 GPixel/s pixel rate is meaningful for graphics output, while the MI308X records zero. The 8 ray tracing cores provide hardware acceleration for ray-traced scenes, a feature the MI308X does not document. API support for DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4 means the A380M can run standard graphics stacks; the MI308X cannot. The 35 W TDP versus 750 W means the Intel part operates at roughly 4.7% of the AMD part's power draw, making it suitable for compact, thermally constrained environments such as portable devices. The A380M also has a higher base clock (1550 MHz versus 1000 MHz), which can benefit latency-sensitive workloads that do not scale with massive parallelism.

FAQ

Q: Which GPU has higher FP32 compute performance?

A: The AMD Instinct MI308X records 81.72 TFLOPS FP32, exactly 20x the Intel Arc A380M's 4.096 TFLOPS FP32.

Q: How do memory capacities compare?

A: The MI308X has 192 GB of HBM3, while the A380M has 6 GB of GDDR6. The MI308X also has an 8192-bit memory bus versus 96-bit, and 5.32 TB/s bandwidth versus 186.0 GB/s.

Q: Does the MI308X support graphics APIs?

A: No. The MI308X records N/A for DirectX, OpenGL, and Vulkan. The A380M supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

Q: What are the thermal design power figures?

A: The MI308X has a 750 W TDP with a suggested PSU of 1150 W. The A380M has a 35 W TDP with no suggested PSU listed.

Q: Which GPU includes ray tracing hardware?

A: The Intel Arc A380M includes 8 ray tracing cores. The AMD Instinct MI308X records no ray tracing core count.

Q: What are the manufacturing process nodes?

A: The MI308X uses a 5 nm TSMC process with 153,000 million transistors on a 1017 mm² die. The A380M uses a 6 nm TSMC process with 7,200 million transistors on a 157 mm² die.

The Verdict

The database indicates two products with no shared benchmark results and no overlapping use cases. The MI308X is a 750 W OAM module with 192 GB of HBM3, 81.72 TFLOPS FP32, no display outputs, and no graphics API support. It is designed for large-scale compute workloads where memory capacity and bandwidth dominate. The A380M is a 35 W MXM module with 6 GB of GDDR6, 4.096 TFLOPS FP32, 64.00 GPixel/s pixel rate, 8 ray tracing cores, and full graphics API support. It is designed for portable or embedded graphics tasks.

The MI308X should be selected when the workload demands maximum FP32 or FP16 throughput, very large memory residency, or extreme memory bandwidth. The 5.32 TB/s bandwidth and 192 GB capacity are the defining characteristics. The A380M should be selected when the workload requires graphics API compatibility, pixel rendering, ray tracing, or operation within a 35 W power envelope. Its active production status and portable-device-dependent display outputs confirm its role in mobile or compact systems.

Neither part can substitute for the other. The MI308X cannot render frames or run DirectX/Vulkan applications. The A380M cannot approach the compute density or memory scale of the MI308X. The data supports a strict segmentation: compute accelerator versus mobile graphics processor.

Specification Differences

| Field | AMD Instinct MI308X | Intel Arc A380M |

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

| Chip | Aqua Vanjaram | DG2-128 |

| Architecture | CDNA 3.0 | Xe-HPG |

| Generation | Instinct (MIx) | Alchemist (Arc 3 Mobile) |

| Process Node | 5 nm | 6 nm |

| Transistors | 153,000 million | 7,200 million |

| Die Size | 1017 mm² | 157 mm² |

| Transistor Density | 150.4M / mm² | 45.9M / mm² |

| Base Clock | 1000 MHz | 1550 MHz |

| Boost Clock | 2100 MHz | 2000 MHz |

| Memory Clock | 1300 MHz, 5.2 Gbps effective | 1937 MHz, 15.5 Gbps effective |

| Memory Size | 192 GB | 6 GB |

| Memory Type | HBM3 | GDDR6 |

| Memory Bus Width | 8192 bit | 96 bit |

| Memory Bandwidth | 5.32 TB/s | 186.0 GB/s |

| Shading Units | 19456 | 1024 |

| TMUs | 1216 | 64 |

| ROPs | 0 | 32 |

| Ray Tracing Cores | Not recorded | 8 |

| Pixel Rate | 0 MPixel/s | 64.00 GPixel/s |

| Texture Rate | 2,553.6 GTexel/s | 128.0 GTexel/s |

| FP32 Compute | 81.72 TFLOPS | 4.096 TFLOPS |

| FP16 Compute | 81.72 TFLOPS (1:1) | 8.192 TFLOPS (2:1) |

| TDP | 750 W | 35 W |

| Slot Width | OAM Module | MXM Module |

| Power Connectors | None | Not recorded |

| Suggested PSU | 1150 W | Not recorded |

| Bus Interface | PCIe 5.0 x16 | MXM-A (3.1) |

| Display Outputs | No outputs | Portable Device Dependent |

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

| OpenGL | N/A | 4.6 |

| Vulkan | N/A | 1.4 |

| Production Status | Not recorded | Active |

| Release Date | 2023-12-05 | 2023-01-23 |

| Predecessor | Radeon Instinct | Not recorded |