AMD Instinct MI355X vs Intel Arc A380E Comparison

Head-to-Head Benchmarks

The recorded database contains no synthetic or application-level benchmark scores for either the AMD Instinct MI355X or the Intel Arc A380E. Both entries return zero benchmark results and zero wins in head-to-head comparisons. The percentile fields place both parts at the 50th percentile against all GPUs, though this value appears to be a neutral default rather than a derived ranking, given the absence of any measured workload data. Consequently, a direct numerical performance comparison cannot be constructed from the available measurements. What the data does provide is a set of architectural and specification deltas that allow for an informed qualitative projection of relative capability, particularly in compute throughput, memory bandwidth, and feature support.

The most striking numerical difference appears in raw compute throughput. The MI355X lists FP32 performance at 78.64 TFLOPS, while the A380E lists 4.096 TFLOPS. That represents a 19.2x gap in single-precision floating-point throughput. For FP16, the MI355X again lists 78.64 TFLOPS with a 1:1 ratio, meaning it does not gain any advantage by switching to reduced precision. The A380E lists 8.192 TFLOPS FP16 with a 2:1 ratio, meaning it doubles its throughput relative to FP32. Even with that efficiency advantage, the A380E reaches only about 10.4% of the MI355X's FP16 throughput. Texture rate differences are similarly lopsided: the MI355X lists 2,457.6 GTexel/s versus 128.0 GTexel/s for the A380E, a 19.2x gap consistent with the FP32 ratio. Pixel rate is an outlier: the MI355X lists 0 MPixel/s due to having no ROPs, while the A380E lists 64.00 GPixel/s. This reflects their fundamentally different design targets, rather than a meaningful comparison of rasterization ability.

Memory bandwidth shows another enormous divide. The MI355X lists 8.19 TB/s across an 8192-bit HBM3e interface, while the A380E lists 186.0 GB/s across a 96-bit GDDR6 bus. The MI355X therefore provides approximately 44.0x the memory bandwidth of the A380E. Capacity differs by a factor of 48: 288 GB versus 6 GB. These are not incremental differences; they represent different classes of hardware entirely.

Architecture Differences

The two parts come from opposite ends of the GPU spectrum, and their architecture choices reflect that. The MI355X uses CDNA 4.0, AMD's compute-optimized architecture, built on a TSMC 3 nm process with 185,000 million transistors on a 2380 mm² die. Transistor density reaches 77.7M per mm². The chip is designated MI350 256CU, indicating 256 compute units. It has 16,384 shading units and 1,024 texture mapping units, but zero ROPs. That absence of ROPs, combined with no display outputs, confirms a pure compute accelerator design with no rasterization pipeline. The API support list shows N/A for DirectX, OpenGL, and Vulkan, reinforcing that this part is not intended for graphics workloads.

The A380E uses Intel's Xe-HPG architecture, specifically the DG2-128 chip, on a TSMC 6 nm process with 7,200 million transistors on a 157 mm² die. Transistor density is 45.9M per mm². It has 1,024 shading units, 64 TMUs, 32 ROPs, and 8 ray tracing cores. It supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, and drives 4x DisplayPort 2.0 outputs. This is a full graphics card with ray tracing hardware, intended for rendering and display. The MI355X has no ray tracing cores listed, no tensor cores listed, and no graphics API support. The A380E has no tensor cores listed either, but its Xe-HPG architecture includes dedicated Xe cores that handle vector and matrix operations; the database does not enumerate them separately.

Clock behavior also differs. The MI355X has a 1000 MHz base clock and 2400 MHz boost clock, a 2.4x boost range. The A380E runs at a flat 2000 MHz for both base and boost, with no clock headroom. Memory clocks differ as well: the MI355X runs at 2000 MHz with 8 Gbps effective data rate, while the A380E runs at 1937 MHz with 15.5 Gbps effective. The A380E's smaller bus width (96 bit versus 8192 bit) is the limiting factor, resulting in the much lower total bandwidth.

The power envelope tells the same story. The MI355X lists a TDP of 1400 W with a suggested PSU of 1800 W, while the A380E lists 75 W TDP with a suggested PSU of 250 W. Neither requires external power connectors. The MI355X is an OAM module, 102 mm long and 165 mm wide, while the A380E is a single-slot card, 254 mm long, 127 mm tall, and 20 mm wide. The MI355X has no display outputs; the A380E has four. The MI355X uses PCIe 5.0 x16; the A380E uses PCIe 4.0 x8.

FAQ

Q: Which GPU has higher FP32 compute throughput?

A: The AMD Instinct MI355X lists 78.64 TFLOPS FP32, which is 19.2x the Intel Arc A380E's 4.096 TFLOPS. This is the largest single-metric gap in the database comparison.

Q: Does the Intel Arc A380E support ray tracing?

A: Yes. The A380E lists 8 ray tracing cores and supports DirectX 12 Ultimate (12_2), which includes ray tracing APIs. The MI355X lists no ray tracing cores and has N/A for all graphics APIs.

Q: What is the memory capacity difference?

A: The MI355X has 288 GB of HBM3e memory, while the A380E has 6 GB of GDDR6. The MI355X provides 48x the capacity and approximately 44x the bandwidth (8.19 TB/s versus 186.0 GB/s).

Q: Can either GPU be used for display output?

A: Only the A380E. It provides 4x DisplayPort 2.0 outputs. The MI355X has no display outputs and lists 0 MPixel/s pixel rate due to having no ROPs.

Q: What process nodes do the two chips use?

A: The MI355X uses TSMC 3 nm with 185,000 million transistors on a 2380 mm² die. The A380E uses TSMC 6 nm with 7,200 million transistors on a 157 mm² die. The MI355X has a transistor density of 77.7M per mm² versus 45.9M per mm² for the A380E.

Q: Which GPU has a higher boost clock?

A: The A380E runs at a flat 2000 MHz for both base and boost. The MI355X has a 1000 MHz base and 2400 MHz boost, so its boost clock is 400 MHz higher than the A380E's sustained clock.

Specification Differences

| Field | AMD Instinct MI355X | Intel Arc A380E |

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

| Architecture | CDNA 4.0 | Xe-HPG |

| Process Node | 3 nm | 6 nm |

| Transistors | 185,000 million | 7,200 million |

| Die Size | 2380 mm² | 157 mm² |

| Transistor Density | 77.7M / mm² | 45.9M / mm² |

| Base Clock | 1000 MHz | 2000 MHz |

| Boost Clock | 2400 MHz | 2000 MHz |

| Memory Clock | 2000 MHz 8 Gbps effective | 1937 MHz 15.5 Gbps effective |

| Memory Size | 288 GB | 6 GB |

| Memory Type | HBM3e | GDDR6 |

| Memory Bus Width | 8192 bit | 96 bit |

| Memory Bandwidth | 8.19 TB/s | 186.0 GB/s |

| Shading Units | 16,384 | 1,024 |

| TMUs | 1,024 | 64 |

| ROPs | 0 | 32 |

| Ray Tracing Cores | None listed | 8 |

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

| Texture Rate | 2,457.6 GTexel/s | 128.0 GTexel/s |

| FP32 | 78.64 TFLOPS | 4.096 TFLOPS |

| FP16 | 78.64 TFLOPS (1:1) | 8.192 TFLOPS (2:1) |

| TDP | 1400 W | 75 W |

| Suggested PSU | 1800 W | 250 W |

| Slot Width | OAM Module | Single-slot |

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

| Display Outputs | No outputs | 4x DisplayPort 2.0 |

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

| OpenGL Support | N/A | 4.6 |

| Vulkan Support | N/A | 1.4 |

| Dimensions | 102 mm x 165 mm | 254 mm x 127 mm x 20 mm |

| Release Date | 2025-06-11 | 2024-03-31 |

| Production Status | Not listed | End-of-life |

| Successor | None listed | Battlemage |

Where Each One Wins

The AMD Instinct MI355X wins decisively in every compute-oriented metric where the database records numbers. FP32 throughput is 19.2x higher. FP16 throughput is 9.6x higher even without accounting for the A380E's 2:1 ratio advantage. Texture rate is 19.2x higher. Memory bandwidth is approximately 44x higher. Memory capacity is 48x higher. The MI355X also uses a more advanced process node (3 nm versus 6 nm), has a higher transistor density, and supports PCIe 5.0 x16 versus PCIe 4.0 x8. Its boost clock of 2400 MHz exceeds the A380E's 2000 MHz. The MI355X has 16x the shading units and 16x the TMUs. It has no ROPs, which means it cannot rasterize, but in pure compute terms it holds every recorded advantage.

The Intel Arc A380E wins in every graphics-oriented metric. It has 32 ROPs versus zero, enabling a pixel rate of 64.00 GPixel/s. It has 8 ray tracing cores. It supports DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4. It provides 4x DisplayPort 2.0 outputs. Its base clock of 2000 MHz is double the MI355X's base clock of 1000 MHz, though the MI355X's boost clock overtakes it. The A380E also has a much lower power requirement: 75 W TDP versus 1400 W, with a suggested PSU of 250 W versus 1800 W. It is a standard single-slot card with conventional dimensions, whereas the MI355X is an OAM module. The A380E has a flat clock curve, which may indicate more predictable sustained performance in its intended workloads. Its FP16 ratio of 2:1 gives it a relative efficiency advantage in half-precision tasks, even though its absolute FP16 output remains far below the MI355X.

The Verdict

The database presents two parts with no overlapping design goals. The AMD Instinct MI355X is a compute accelerator: no display outputs, no ROPs, no graphics API support, 288 GB of HBM3e, 8.19 TB/s bandwidth, and 78.64 TFLOPS FP32. It targets workloads where memory capacity and floating-point throughput dominate, such as large-scale inference or scientific simulation. Its 1400 W TDP and OAM form factor indicate a server or data center installation, not a desktop. The Intel Arc A380E is a graphics card: 4x DisplayPort 2.0, ray tracing cores, DirectX 12 Ultimate support, 6 GB GDDR6, and a 75 W TDP. It targets rendering, display output, and general-purpose graphics workloads in a conventional single-slot form factor.

For a user selecting between these two, the decision hinges entirely on workload type. If the task requires rasterization, ray tracing, or driving displays, only the A380E is viable; the MI355X cannot output video and has no graphics pipeline. If the task requires massive memory capacity or extreme FP32 throughput, only the MI355X is viable; the A380E's 6 GB capacity and 4.096 TFLOPS are insufficient for large compute workloads. The MI355X also requires a substantially larger power budget and cooling solution, given the 1400 W TDP and 1800 W suggested PSU. The A380E, with its 75 W TDP and 250 W suggested PSU, fits into far more modest systems.

Release timing suggests the MI355X is the newer part, with a recorded release date of 2025-06-11, while the A380E released on 2024-03-31 and is marked end-of-life with a successor named Battlemage. The MI355X has no listed successor or production status. Neither part has a recorded launch MSRP in the database. The MI355X's predecessor is listed as Radeon Instinct, while the A380E's predecessor is Xe Graphics. These lineage entries reinforce their separate trajectories: one from AMD's compute-oriented Instinct line, the other from Intel's graphics-oriented Xe line. Given the data, the MI355X is the compute leader by every recorded metric, and the A380E is the graphics leader by every recorded metric. There is no scenario in the recorded data where one outperforms the other outside its own domain.